Docstoc

Thesis

Document Sample
Thesis Powered By Docstoc
					                                                       ABSTRACT .
This thesis uses a literature survey to evaluate the assumptions made in Froot, Scharfstein and Stein's framework published in the
Journal of Finance in 1993. Froot, Scharfstein and Stein argued that because external capital is expensive firms s hould use derivatives
to redistribute the proceeds from assets in place so that the amount of internal capital would correlate with the investment and
financing opportunities.

The main question in this thesis was whether Froot, Scharfstein and Stein's rather technical framework should be implemented by firms.
The answer to that question is no for most firms.

This thesis shows that external capital is not expensive for all firms. More important is the fact that although the derivatives market is
huge, the amount of underlying assets is limited. If the firm can not trade derivatives on the risky assets that determines the inves tment
opportunities Froot, Scharfstein and Stein's framework can not be used. This is true for most firms. Empirical work does howe ver
support many of Froot, Scharfstein and Stein's arguments.
                                                                      TABLE OF CONTENTS.
              Abstract. …………………………………………………………...……………………………..1
              Table of Contents. ………………………………………………...…………...…………………2
              List of Figures. …………………………………………………...………………………………7
              List of Tables. ………………………………………………...…...……………………………..7
              Symbols. ………………………………….……………………………………………………....8
1. INT RODUCT ION...................................................................................................................................................... 7
1.1    This Thesis. ..................................................................................................................................................... 7
1.2    Research Design ............................................................................................................................................. 8
1.2.1      Research Problems. ................................................................................................................................. 8
1.2.2      Choice of Research Method. ................................................................................................................. 8
1.2.3      Research Limitations............................................................................................................................... 8
1.2.4      Overview of this Thesis. ........................................................................................................................ 8
1.2.5      The Consequences of Limiting the Research Material...................................................................... 9
1.3    Why Study This?............................................................................................................................................ 9
2. OVERVIEW OF RELEVANT FINANCIAL THEORY. ............................................................................................. 9
2.1   The Goal of the Firm. ...................................................................................................................................... 9
2.2   Risk.................................................................................................................................................................... 9
2.3   Hedging. Definition and Objectives........................................................................................................... 11
2.4   Other Hedging Theories. ............................................................................................................................. 11
3. THE PROBLEMS. ................................................................................................................................................... 12
3.1     Problem I. Are the Costs of External Capital Necessarily High?............................................................ 12
3.2     Problem II. What Are the Costs of Froot, Scharfstein and Stein's Framework?.................................. 12
3.3     Problem III. Will Stakeholders Approve?.................................................................................................. 12
3.4     Problem IV. What Makes the Firm Invest? ............................................................................................... 12
3.5     Problem V. What Instruments Should Be Used? ..................................................................................... 12
3.5.1       Do the Right Derivatives Exist? .......................................................................................................... 12
3.5.1.1         A Perfect Hedge............................................................................................................................. 13
3.5.1.2         The Derivatives' Time to Maturity. ............................................................................................. 13
3.5.2       Can Other Derivatives Be Used? ........................................................................................................ 13
3.6     Problem VI. Does Empirical Evidence Support Froot, Scharfstein and Stein's Framework?.............. 13
4. A BST RACT OF FROOT , SCHARFST EIN AND ST EIN'S PAPER, WIT H EXT ENSIONS. .................................... 13
4.1     The Origin. ..................................................................................................................................................... 14
4.2     The Framework. ............................................................................................................................................. 14
4.3     A Two-Period Model.................................................................................................................................... 15
4.4     Hedging in Froot, Scharfstein and Stein's Framework. ........................................................................... 16
4.4.1       The Simple Model. ................................................................................................................................ 16
4.4.1.1          Adverse Protection in the Simple Model. .................................................................................. 17
4.4.1.2          No Protection in the Simple Model, Out of Boundary Exposure............................................ 17
4.4.2       Changing Investment Opportunities.................................................................................................. 18
4.4.2.1          The No-Hedging Model. .............................................................................................................. 18
4.4.2.2          Adverse Protection in the No-Hedging Model......................................................................... 19
4.4.2.3          The Speculation Model. ............................................................................................................... 19
4.4.2.4          The Uncorrelated Model. ............................................................................................................. 19
4.4.2.5          The Reversed Simple Model. ....................................................................................................... 20
4.5     The Hedge Ratio. .......................................................................................................................................... 20
4.6     Changing Financing Opportunities............................................................................................................ 21
4.7     Multinationals. .............................................................................................................................................. 21
4.8     Further Extensions. ....................................................................................................................................... 21
4.9     Implications of Lognormal Distribution of the Risky Asset................................................................... 22
5. A RE T HE COST S OF EXT ERNAL CAPIT AL NECESSARILY HIGH? ................................................................. 23
5.1    Research on Capital Structure..................................................................................................................... 23
5.2    Why Use Outside Financing at All? .......................................................................................................... 24
5.3    Debt Financing. ............................................................................................................................................. 24
5.3.1            The Use of Debt. ................................................................................................................................... 24
5.3.2            Why Risky Debt Matters. .................................................................................................................... 24
5.3.3            The Costs of Bankruptcy and Financial Distress............................................................................. 25
5.3.3.1                 Bankruptcy. .................................................................................................................................... 25
5.3.3.2                 Financial Distress. ......................................................................................................................... 26
5.3.4            Reducing the Costs of Bankruptcy and Financial Distress. ........................................................... 26
5.3.5            Other Disadvantages of Debt.............................................................................................................. 27
5.3.6            Agency Costs of Debt.......................................................................................................................... 27
5.3.7            Reducing the Cost of Debt. ................................................................................................................. 27
5.3.7.1                 Low Risk Projects. ......................................................................................................................... 28
5.3.7.2                 Contracts......................................................................................................................................... 28
5.3.7.3                 Dividend Restrictions. .................................................................................................................. 28
5.3.7.4                 The Manager. ................................................................................................................................. 29
5.3.7.5                 Monitoring...................................................................................................................................... 29
5.3.7.6                 Reputation. ..................................................................................................................................... 29
5.3.7.7                 Growth Options VS Collateral. ..................................................................................................... 29
5.3.7.8                 The Debt Structure. ....................................................................................................................... 29
5.3.7.9                 Bank Debt. ...................................................................................................................................... 30
5.3.7.10                  Convertible Debt. ........................................................................................................................ 30
5.3.7.11                  Repurchasing Stocks. ................................................................................................................. 30
5.3.8            Summary. ................................................................................................................................................ 30
5.4          Equity Financing. .......................................................................................................................................... 31
5.4.1            The Manager Acts in Self-Interest. .................................................................................................... 31
5.4.2            Reducing the Agency Costs of Managers Who Act in Self-Interest. ......................................... 31
5.4.2.1                 Growth Options.............................................................................................................................. 32
5.4.2.2                 Debt. ................................................................................................................................................ 32
5.4.2.3                 Convertible Bonds......................................................................................................................... 32
5.4.2.4                 Monitoring...................................................................................................................................... 33
5.4.2.5                 The Board. ...................................................................................................................................... 33
5.4.2.6                 The Manager's Ownership. .......................................................................................................... 33
5.4.2.7                 The Manager's Duties................................................................................................................... 34
5.4.2.8                 Take-over Threat. .......................................................................................................................... 34
5.4.3            Summary. ................................................................................................................................................ 34
5.4.4            The Manager Acts in the Interest of the Existing Shareholders.................................................... 34
5.4.4.1                 Firm Quality. ................................................................................................................................... 34
5.4.4.2                 Timing.............................................................................................................................................. 35
5.4.5            Reducing Agency Costs Between Existing VS New Shareholders. .............................................. 36
5.4.5.1                 The Pecking Order. ........................................................................................................................ 36
5.4.5.2                 The Pecking Order, Empirical Tests. ........................................................................................... 37
5.4.5.3                 Timing.............................................................................................................................................. 37
5.4.5.4                 Contracts......................................................................................................................................... 38
5.4.5.5                 The Manager's Ownership. .......................................................................................................... 38
5.4.5.6                 Capital Structure. ........................................................................................................................... 38
5.4.6            Summary. ................................................................................................................................................ 38
5.5          IPOs................................................................................................................................................................. 38
5.5.1            Firm Size and Age.................................................................................................................................. 39
5.5.2            Why IPOs Underprice. ......................................................................................................................... 39
5.5.3            Who Carries the Costs? ....................................................................................................................... 39
5.5.4            Underwriters........................................................................................................................................... 39
5.5.5            The IPO and SEOs................................................................................................................................. 40
5.5.6            Privately Held Firms. ............................................................................................................................. 41
5.6          Transaction Costs......................................................................................................................................... 41
5.7          Conclusion. .................................................................................................................................................... 41
6. W HAT A RE T HE COST S OF FROOT , SCHARFST EIN AND ST EIN'S FRAMEWORK?..................................... 42
6.1   The Direct Costs of Hedging. ..................................................................................................................... 42
6.1.1     Transaction Costs. ................................................................................................................................ 42
6.1.2     Opportunity Cost. ................................................................................................................................. 42
6.1.3     Liquidity.................................................................................................................................................. 42
6.1.4     Managing the Derivatives Portfolio. .................................................................................................. 42
6.2   Conclusion. .................................................................................................................................................... 43
7. W ILL ST AKEHOLDERS A PPROVE ? ................................................................................................................... 43
7.1    The Manager. ................................................................................................................................................ 43
7.1.1            The Manager's Risk Aversion............................................................................................................. 44
7.1.2            Does the Firm Become a Take-over Target? ..................................................................................... 44
7.1.3            Is Froot, Scharfstein and Stein's framework a Better Alternative Than to Go Public? ............... 45
7.2          The Shareholders. ......................................................................................................................................... 46
7.3          Will Other Stakeholders Approve? ............................................................................................................ 46
7.4          Concision. ...................................................................................................................................................... 46
8. W HAT M AKES T HE FIRM INVEST ?................................................................................................................... 47
8.1   Valuation of Investments............................................................................................................................. 47
8.1.1     The Discounted Cashflow Model and Why it May Not Be an Optimal Decision Rule. ............ 47
8.1.2     Simple Methods..................................................................................................................................... 48
8.1.3     Real Options........................................................................................................................................... 48
8.2   About the Firm's Investments?................................................................................................................... 48
8.2.1     Strategy................................................................................................................................................... 48
8.2.2     Strategic Behaviour............................................................................................................................... 49
8.2.3     Replacement Investments. ................................................................................................................... 49
8.2.4     Technology. ........................................................................................................................................... 49
8.3   What Make Firms Invest? ........................................................................................................................... 49
8.3.1     General Findings.................................................................................................................................... 50
8.3.2     The Interests rate. ................................................................................................................................. 50
8.3.3     The Exchange Rate................................................................................................................................ 50
8.3.4     The Business Cycle. ............................................................................................................................. 51
8.3.5     Leading Indicators. ............................................................................................................................... 51
8.4   Comments....................................................................................................................................................... 52
9. W HAT INST RUMENT S SHOULD BE USED? ...................................................................................................... 52
9.1      How Derivatives Are Used.......................................................................................................................... 52
9.2      Futures............................................................................................................................................................ 52
9.2.1        About Futures. ...................................................................................................................................... 52
9.2.2        Advantages of Futures. ....................................................................................................................... 53
9.2.3        Disadvantages of Futures.................................................................................................................... 53
9.3      Long-term Contracts..................................................................................................................................... 53
9.3.1        About Long-term Contracts. ............................................................................................................... 53
9.3.2        Advantages of Long-term Contracts. ................................................................................................ 53
9.3.3        Disadvantages of Long-term Contracts............................................................................................. 54
9.4      Forwards......................................................................................................................................................... 54
9.4.1        About Forwards. ................................................................................................................................... 54
9.4.2        Advantages of Forwards. .................................................................................................................... 54
9.4.3        Disadvantages of Forwards................................................................................................................. 54
9.4.4        Special Types of Forwards. ................................................................................................................. 54
9.4.4.1           Forwards on Commodities............................................................................................................ 54
9.4.4.2           Forwards with Optionlike Characteristics. ................................................................................. 55
9.5      FRAs, IRGs and SAFEs. .............................................................................................................................. 55
9.6      Options. .......................................................................................................................................................... 55
9.6.1        About Options....................................................................................................................................... 55
9.6.2        Advantages of Options........................................................................................................................ 56
9.6.3        Disadvantages of Options. .................................................................................................................. 56
9.6.4        Special types of options....................................................................................................................... 56
9.6.4.1           Macro Options. .............................................................................................................................. 56
9.6.4.2           Exotic Options. ............................................................................................................................... 56
9.6.4.3           Digital Options ............................................................................................................................... 56
9.6.4.4           Barrier Options. .............................................................................................................................. 56
9.6.4.5           Non-linear Options. ....................................................................................................................... 57
9.6.4.6           Pay-Later Options.......................................................................................................................... 57
9.6.4.7           Chooser Options............................................................................................................................ 57
9.6.4.8           Asian Options. ............................................................................................................................... 57
9.6.4.9           Options That Are Dependent of More Than One Asset. ....................................................... 58
9.6.4.10            Compound Options..................................................................................................................... 58
9.6.5        Different Types of Option Strategies. ................................................................................................ 58
9.6.6        Synthetic futures. .................................................................................................................................. 59
9.7      Caps and Floors. ........................................................................................................................................... 60
9.7.1        Caps......................................................................................................................................................... 60
9.7.2        Floor. ....................................................................................................................................................... 60
9.8      Swaps.............................................................................................................................................................. 60
9.8.1        About Swaps. ........................................................................................................................................ 60
9.8.2        Advantages of Swaps. ......................................................................................................................... 61
9.8.3        Disadvantages of Swaps...................................................................................................................... 61
9.8.4        Different Types of Swaps. ................................................................................................................... 61
9.8.4.1           Swaps with options like characteristics. .................................................................................... 61
9.8.4.2           Macro Swaps.................................................................................................................................. 62
9.9      Conclusion. .................................................................................................................................................... 62
9.9.1        Which Instruments Are Suited. .......................................................................................................... 62
9.9.2        Advanced Hedging Strategies. ........................................................................................................... 62
9.10      Can the Investment Opportunities Be Hedged? ..................................................................................... 62
9.10.1         Contracts Written Between Suppliers and Customers.................................................................. 62
9.10.2         Marketable Securities. ........................................................................................................................ 63
9.10.3         The Risky Assets that Can Be Hedged. .......................................................................................... 63
9.10.3.1           Investment Triggers and Derivatives....................................................................................... 64
9.10.3.2           Strategic VS Tactical Investments. ........................................................................................... 64
9.10.3.3           Replacement Investments ......................................................................................................... 64
9.11      Conclusion.................................................................................................................................................... 64
10. DOES EMPIRICAL EVIDENCE SUPPORT FROOT , SCHARFST EIN AND ST EIN'S FRAMEWORK?.............. 65
10.1  Research on the Topic. ............................................................................................................................... 65
10.2  Do Firms Rely on Internal Capital? ........................................................................................................... 67
10.3  Do Variations in the Cashflow Cause Investments to Fluctuate?........................................................ 67
10.4  Conclusion.................................................................................................................................................... 67
11. CONCLUSION ...................................................................................................................................................... 68
11.1  Conclusion.................................................................................................................................................... 68
11.2  Are the Costs of External Capital Necessarily High? ............................................................................. 68
11.3  What Are the Costs of Froot, Scharfstein and Stein's Framework? .................................................... 68
11.4  Will Stakeholders Approve?...................................................................................................................... 68
11.5  What Makes the Firm Invest? ................................................................................................................... 68
11.6  What Instruments Should Be Used? ........................................................................................................ 69
11.7  Does Empirical Evidence Support Froot, Scharfstein and Stein's Framework?.................................. 69
11.8  Does the Framework Management Risk?................................................................................................. 69
12. A PPENDIX .......................................................................................................................................................... 69
12.1     Value-at-Risk. ............................................................................................................................................... 69
12.2     Capital Structure Models............................................................................................................................ 70
12.2.1        Modigliani and Miller. ........................................................................................................................ 70
12.2.1.1           No taxes. ....................................................................................................................................... 70
12.2.1.2           Corporate Taxes........................................................................................................................... 71
12.2.1.3           Personal taxes. ............................................................................................................................. 72




                                                                                      List of Figures.
Figure 1 Risk profile..................................................................................................................................................... 11
Figure 2 Situation 1 and 4. .......................................................................................................................................... 16
Figure 3 Situation 2 and 3. .......................................................................................................................................... 16
Figure 4 Stochastic distribution of the cashflow. ................................................................................................... 16
Figure 5 The Simple model using long-term contracts........................................................................................... 17
Figure 6 The Simple model using futures................................................................................................................. 17
Figure 7 Linear hedging with added adverse protection. ...................................................................................... 17
Figure 8 Out of boundary exposure. ......................................................................................................................... 18
Figure 9 Correlated investment opportunities and cashflows. ............................................................................. 18
Figure 10 The No-hedging model with added adverse protection....................................................................... 19
Figure 11 Highly sensitive demand and less sensitive cashflow. The Speculation model. ............................. 19
Figure 12 The Reversed simple model. ..................................................................................................................... 20
Figure 13 Optimal capital structure with risky debt. ............................................................................................... 25
Figure 14 The impact of bankruptcy costs on the firm value................................................................................ 25
Figure 15 Efficient derivatives. .................................................................................................................................. 43
Figure 16 Digital option. ............................................................................................................................................. 56
Figure 17 Up-and-in barrier call option..................................................................................................................... 57
Figure 18 Pay later option........................................................................................................................................... 57
Figure 19 Single option. Adverse protection in the simple model, the no -hedging model and the uncorrelated model.                                                           58
Figure 20 Bullspread. Out of boundary exposure in the simple model, and the reversed simple model........ 59
Figure 21 Speculation model, pure form. .................................................................................................................. 59
Figure 22 Straddle. Speculation model with adverse protection. ......................................................................... 59
Figure 23 One-sided protection by a cap. ................................................................................................................ 60
Figure 24 The cost of capital in a world without taxes, M&M 1958. ................................................................... 71
Figure 25 The cost of capital in a world with corporate taxes, M&M 1963. ....................................................... 71




                                                                                 List of Tables.
Table 1 Symbols used in this thesis, in chronological order. ................................................................................. 7
Table 2 Froot, Scharfstein and Stein’s framework. ................................................................................................. 15
Table 3 Sources of correlation between the risky asset and the cashflow. ........................................................ 16
Table 4 The risky asset and the investment opportunities. .................................................................................. 19
Table 5 The investment opportunities and the w in the no -correlation model. ................................................. 20
Table 6 Avoiding the agency problems associated with debt financing............................................................ 31
Table 7 Avoid the agency problems of overinvestments. .................................................................................... 34
Table 8 Negative returns associated with security offerings. .............................................................................. 35
Table 9 Explanations for the stock repurchasing.................................................................................................... 38
Table 10 Avoiding the agency problems of overinvestments.............................................................................. 38
Table 11 Direct Costs of Offerings............................................................................................................................ 41
Table 12 Pros and cons of different systems solutions......................................................................................... 43
Table 13 Different kinds of swaps............................................................................................................................. 61
Table 14 Which derivatives fit the different investment triggers......................................................................... 64
Table 15 Internal vs. external capital, 1981-94. ........................................................................................................ 67


                                                                                        Symbols.
P                                Profit function. MAX [F(I) - C(e)] = MAX [f(I) - I - C(e)].
I                                Investments.
w                                The cashflows from assets in place.
Pw                               FOC of the profitfunction with respect to w.
e                                External capital, equity and debt.
C(e)                             Dead-weight costs of external financing.
Ce                               FOC of the dead-weight costs with respect to the external capital.
F(I)                             The net present value.
FI                               FOC of the net present value with respect to investments. fI - 1- Ce
f(I)                             The present value.
I                                Investments. w + e
                                The value of the risky asset.  is normally distributed with a mean equal to .
E()                             The expected value of .
 - E()                         Deviations from the expected value.
R                               The risky asset when changes in the risky asset changes the firm's unit revenue.
C                               The risky asset when changes in the risky asset changes the firm's unit cost.
w                               The value of the risky asset when it is of secondary importance if changes in the
                                 risky asset changes the unit revenue or the unit cost, but the main importance is that
                                 the amount of internal capital changes. This means that w = R if the risky asset that
                                 changes the cashflows from assets in place does this through increased unit
                                 revenues. The same can of cause be the case for w = C. If two different risky assets
                                 changes both the revenues and the costs w = C + R.
CORR (w, w)                     The correlation between the amount of internal capital and the risky asset that
                                 changes the amount of internal capital.
wR                               The redistributed amount of internal capital, redistributed through derivatives
                                 hedging.
D                               The value of the risky asset that changes the demand for the firm's goods.
h                                The optimal hedge ratio.
w0                      The expected cashflows from assets in place, w0 = E(w).
X                      The value of the risky asset that changes either the demand for the firm's goods, the
                        unit revenue or the unit costs. The risky asset is thus either D, C or R. This risky
                        asset changes the amount of internal capital either through the prices the firm can
                        charge, the costs or the sales.
                       Parameter between zero and one of the sensitivity between foreign revenues and the
                        risk, the exchange-rate.
 = ( - E()) + 1
fH                      The present value in the home market.
IH                      Investments in the home market.
fA                      The present value abroad.
IH                      Investments abroad.
                       A parameter between zero and one of the sensitivity between foreign investment
                        costs and the risk, the exchange rate.
= ( - E()) + 1.
Table 1 Symbols used in this thesis, in chronological order.



                                                              PART I.
The first part of this thesis covers the introduction, an overview of relevant financial theory, an outline of the questions raised in the
thesis and an abstract of Froot, Scharfstein and Stein's framework


                                              1.          INTRODUCTION.
1.1           This Thesis.
During the second part of 1997 most of the East Asian economies faced financial difficulties. Hong Kong saw a crisis in the r eal estate
industry. The Indonesian economy collapsed and the country ended up with demonstrations and the people demanded President
Suharto's resignation. The difficulties in other East Asian countries turned Japan's ten year long stagnation into a recessio n that
disclosed a banking industry in serious difficulties. Stock markets that had been generous money -makers for investors for years turned
into roller coasters and fundamental analysis was substituted with gambling. In addition to the East Asian economies, other e merging
markets ran into difficulties as well. Russia was hit the hardest, but South American stock markets saw large drops in sharev alues too.
The decline in the huge East Asian markets and the fear of a world recession resulted in two major declines in the Weste rn
stockmarkets in only one year as firms started to report profit warnings.

Risk management on the corporate level, as opposed to investors' risk management , again filled the pages in economic and financial
literature. However, while there are numerous theories that tries to explain the how's of corporate risk management, there is no clear cut
consensus of the why's and even if firms should manage the risks they face.

This thesis takes a critical look at the assumptions and conclusions of one of the theories that tries to explain not only the how's, but
also the why's and the if's in risk management. In the December 1993 issue of The Journal of Finance Froot, Scharfstein and Stein
published a paper called «Risk Management: Co-ordinating Corporate Investment and Financing Policies».
Froot, Scharfstein and Stein tried to answer the following three questions:
 What sorts of risks should be hedged?
 Should they be hedged fully or partly?
 What kinds of instruments will best accomplish the hedging objectives?

Their paper used a different approach to the risk management puzzle than former theories had used. Only a couple of years after it was
                                                                                                   1
published it was already regarded as a breakthrough in the understanding of risk management. Froot, Scharfstein and Stein assumed
that changes in some risky asset changes the business environment in such a way that it is optimal for the firm to change its invests
and financing. The exchange rate, the interests rate and commodity prices are examples of such risky assets. They then argued that if
external capital is more expensive than internal capital, this is a rational for hedging. External capital is equity and/or debt raised by
outside investors, while internal capital is revenues withheld by the firm and not paid out as dividends. The firm should not hedge to
reduce the variation in income per se; it should h edge to redistribute the cashflows so that the cashflow fits the firm's investment and
financing opportunities. Froot, Scharfstein and Stein limited their framework to cover risk management with derivatives even though
risk can be managed in other ways too, such as diversification, foreign denoted debt, operations abroad, serving a wide range of
markets etc.




1
    A good understanding of Froot, Scharfstein and Stein's theory is required in order to understand this thesis. A resume of Fro ot,
    Scharfstein and Stein’s paper, which also includes extensions made by the autho r of this thesis, is found in section 4, «Abstract of
    Froot, Scharfstein and Stein's Paper, with Extensions.», page 13.
The main question of this thesis is whether firms should use hedging to avoid acquiring costly external capital when investme nt
opportunities change.

1.2           Research Design
1.2.1          Research Problems.
It is difficult to do empirical research on risk management and hedging. Froot, Scharfstein and Stein said that since most hedging
operations are off balance sheet, they are not included in datab ases used by researchers, such as the COMPUSTAT.

1.2.2          Choice of Research Method.
Because of the problem mentioned above a literature study has been done to evaluate the usefulness of Froot, Scharfstein and Stein's
framework.
The choice of literature naturally had to be rather selective. The literature that was chosen is mainly written after Froot, Scharfstein and
Stein wrote their paper. The most important reason for this is the fact that if there had been serious theoretical or empiric al evidences
against their framework at the time their paper was written, one must assume that they would have commented those things. Thi s
limitation also reduces the amount of literature down to a manageable amount. All issues of Journal of Finance, Journal of Financial
Economics, Journal of Financial and Quantitative Analysis and Journal of Corporate Finance from 1993 through the summer of 19 98
have been surveyed. This has been supplemented with numerous other papers; both resent papers from other journals and several
older, «classical» papers. In addition to this over thirty books, covering the different subjects in this thesis have been us ed.

1.2.3          Research Limitations.
In order to reduce the size and increase the focus of this thesis only the firms' investment de cisions is covered, not the financing
decisions. This has no consequences in the case of a rejection of the null hypothesis that the framework can and should be us ed by
firms. It does however mean that if this thesis is not able to reject the hypothesis, a similar study of the usefulness of Froot, Scharfstein
and Stein's framework that includes the firms' financing policies, may be able to reject the framework. This reduces the robustness of
the conclusion in this thesis. The fact that the scope of this thesis has been narrowed means that this thesis may ending up supporting
Froot, Scharfstein and Stein's framework even though it would have been rejected if the financing part had been covered as we ll.

Froot, Scharfstein and Stein's assumptions to their second question, should the risk be hedged fully or partly, seems fair and the
conclusions to that question takes up a large portion of their paper and is indeed the most technical part. They came up with the rather
controversial conclusion that the firm should not necessarily hedge all the risk. Non of the assumptions or conclusions concerning
Froot, Scharfstein and Stein's second question have been questioned. Many of the underlying assumptions behind the first and the
third question are however questioned in this thesis.

1.2.4          Overview of this Thesis.
The different theories and findings are presented in a chronological order. This way it is easier to follow the line of reaso ning that the
different authors have used.

The first part of this thesis first goes through relevant financial theory.2 This is followed by an outline of the questions that this thesis
                                                                                                             4
raises and answers.3 The first part ends with an abstract of Froot, Scharfstein and Stein's framework. This is however not only an
abstract. It is quite easy to see that from the basic set-up of Froot, Scharfstein and Stein's framework one can make more models than
what Froot, Scharfstein and Stein did. These models, as well as a couple of other extensions of the framework, have been adde d.
The second part of this theses deals with the more underlying, implicit assumptions that Froot, Scharfstein and Stein made. It
                                                                         5                    6
addresses the questions whether external capital always is expensive, is hedging cheaper and will the stakeholders want to implement
                            7
such a hedging strategy?
The third part covers Froot, Scharfstein and Stein's first and third questions. Froot, Scharfstein and Stein assumed that cha nges in
some risky asset/assets causes the investment and financing opportunities to change. The first question in part three asks what
                             8
actually makes a firm invest. Froot, Scharfstein and Stein just assumed that there are derivatives written on those investment triggers .
                                                                                                    9
Different derivatives and hedging strategies are reviewed in the second se ction on part three. Only when the investment triggers and
available derivatives are identified can one tell what kinds of risks can be hedges and with what instruments.




2
    See section 2, «Overview of Relevant Financial Theory.», page 9.
3
    See section 3, «The Problems.», page 12.
4
    See section 4, «Abstract of Froot, Scharfstein and Stein's Paper, with Extensions.», page 13.
5
    See section 5, «Are the Costs of External Capital Necessarily High?», page 23.
6
    See section 6, «What Are the Costs of Froot, Scharfstein and Stein's Framework?», page 42.
7
    See section 7, «Will Stakeholders Approve?», page 44.
8
    See section 8, «What Makes the Firm Invest?», page 47.
9
    See section 9, «What Instruments Should Be Used?», page 52.
                                                                                                      10
The forth part gives a summary of empirical evidence of Froot, Scharfstein and Stein's framework           as well as the conclusion of this
         11                       12
thesis        and the appendix.

1.2.5            The Consequences of Limiting the Research Material.
One should keep in mind that most of the articles and books covered in this thesis concern firms and trading in the US. This means that
the results presented may not necessarily be applicable for other countries. This is especially true when it comes to the age ncy
problems and problems concerning asymmetric information . Managers may be disciplined by other things in the US than in other
countries and investors may be motivated by other things. There are many, small banks in the US and the market for corporate control
is large and active. The opposite is true in for example Japan and Germany. Some claim that while the former encourage short-term
profits and thus short-term strategies, the latter do the opposite. (Black and Gilson, 1998)

1.3              Why Study This?
In the concluding remarks of their paper Froot, Scharfstein and Stein said that they did not know if the ir framework should be
interpreted solely as a description of how firms do hedge, or as a theory of how firms should hedge.

The findings that have been made in this thesis will help to answer this question. This is important for academics as well as for
corporate managers. Froot, Scharfstein and Stein's framework took a very different standpoint from previous academic work when it
came to the objective of the hedging. Instead of hedging to avoid variations in the cashflow, they argue that the firm should hedge to
optimise variations in the cashflow with respect to investment and financing opportunities.
If Froot, Scharfstein and Stein assumptions and conclusions were correct, research on risk management and project evaluation should
change direction. This means that more emphasis should be put into understanding the corporate investment process, which up u ntil
now basically has relied on evaluation of discounted cashflows (DCF).


                2.         OVERVIEW OF RELEVANT FINANCIAL THEORY.
2.1              The Goal of the Firm.
It is common in financial theory to assume that the goal of a firm is to maximise shareholders' wealth. This means that the f irm's goal is
to give the shareholders a maximum return given the risk level and this return must be at least as good as the return the investors
would get from another investment with the same risk.
Maximising shareholders wealth is implicitly the goal of the firm in Froot, Scharfstein and Stein's framework. There is howev er a logical
failure in Froot, Scharfstein and Stein's framework. The whole idea behind their framework is costly external capital. Some of the
reasons why external capital is costly is the asymmetrical information between the manager and the shareholders and the fact that
managers act in self-interest. This is inconsistent with the assumption that the manager will maximise shareholders' wealth. Agency
problems are essential in the set-up of a model. One must choose one of two different worlds to model. If there are agency problems in
the model, it makes no sense assuming that the manager will do what is best for the shareholder. He will maximise his own bene fits with
respect to giving the shareholders just enough so that he does not lose his job. An alternative way to set up a model is to assume that
there are no agency problems. The problem with such a model is that external capital would have to be priced fairly and the c hoice
between internal and external capital would be less relevant.

It seems that Jensen and Meckling's (1976) definition of a firm as a set of contracts between the stakeholders is better suited for Froot,
Scharfstein and Stein hedging strategy. In such a framework, all stakeholders will maximise their own b enefits. This opens for agency
problems such as moral hazard and adverse selection, which reflect both the real world as well as the implicit assumptions that Froot,
Scharfstein and Stein took. Viewing the firm this way certainly puts restrictions on Froot, Scharfstein and Stein's framework. Their
model will only work to the extent that the stakeholders who can influence the hedging policy will accept it. This means the risk
management strategy must benefit decision-makers, as well as the shareholders. The question whether stakeholders will approve thus
becomes an important part of this thesis.

Rao (1994) said that it is the board's job to supervise the manager on behalf of the shareholders. Many boards fail however to do this.

2.2              Risk.
Rao (1994) made the following distinction between risk and uncertainty.
 Uncertainty: Know all possible outcomes, but not the probabilities.
 Risk: Know both the possible outcomes and the probabilities. This is quantifiable uncertainty.
Jurion (1997) gave the following explanation of risk. Risk is deviations from the expected value and is measured by the stand ard
deviation of the outcomes, which is called volatility. In the fixed-income market, the exposure to the interests rate is called duration.
The stock-market equivalent is systematic risk or beta-risk (). In derivatives, the exposure to movements in the value of the underlying
asset is called delta-risk ().
Jurion gave this list of risk categories.



10
     See section 10, «Does Empirical Evidence Support Froot, Scharfstein and Stein's Framework?», page 65.
11
     See section 11, «Conclusion», page 68.
12
     See section 12, «Appendix», page 70.
     Business risk. This is the risk that firms are there to take; this is what increases shareholders wealth. This risk includes changes in
      technology, design, marketing, management skills, strikes etc. Reynolds (1995) said that these risks should be on the firm's core
      competence.
                                                                      13
     Strategic risk is the risk that the entire market may change.
     Political risk is the risk of political instability, which may alter the competitive environment or in the worst case lead to expropriation
      of plants and equipment.
     Financial risk relates to risk in the financial market due to movements in financial variables.

All these four influences the firm's investment and financing opportunities and are thus relevant for a framework such as Fro ot,
Scharfstein and Stein's.

Financial risk, the risk that Froot, Scharfstein and Stein discussed, can come from several sou rces:
 Credit risk arises when counterparts are unable or unwilling to fulfil their obligations. It does not include loosing the entire notional,
   only the gain from the trade. Galitz (1994) said that the use of an intermediator eliminates this risk.
 Liquidity risk can be due to insufficient market activity or the inability to meet cashflow obligations.
                                                                                           14
 Operational risk is due to management failures, faulty controls, fraud or human error.
 Legal risk arises when the counterpart does not have the legal authority to engage in the transaction.
 Market risk comes from changes in prices of financial assets and liabilities.

Galitz (1994) separated market risk into the following groups:
 Currency risk is exposure to exchange rate changes. Transactions risk affects the proceeds form day to day, while translation risk is
    purely an accounting matter.
 Interests rate risk changes the cost of debt.
 Commodity risk is the risk that commodity prices change.

Although risk management is done to reduce risk, such activities carry with it risks too.
Basis risk occurs when the relationship between the products in the hedge changes. This happens when there is less than perfe ct
correlation between the price of the hedging instrument and the price of the hedged asset. This is the risk that short-term prices deviate
from long-term prices. This can be due to two things:

    The maturity of the financial instrument used in the hedge does not match with the maturity of the exposure. This is delivery basis
     risk.
 There are no instruments that are perfectly correlated with the risk. In this case, the firm must construct a cross hedge. A cross
     hedge is a combination of derivatives constructed to mimic a derivative on the underlying asset, so that it becomes a (close to)
     perfect hedge.
In addition to basis risk, hedging also brings with it model risk, the risk that the mathematical model that is the basis for the hedge is
wrong. This is the fact with the simple Black-Scholes option pricing model published in 1973. The models assumptions do not hold in
               15
the real world.

According to Campbell and Kracaw (1993) the need for risk management started with the break-up of the Bretton Woods exchange rate
agreement in the early 1970's. This lead to increased volatility in interests rate and commodity prices. In October 1979, the US federal
reserve announced that it would let the interests rate be unmanaged. This resulted in even gre ater interests rate volatility.
These risks may have both direct and indirect implications for the firm.
 Smith, Smithson and Wilford (1988) said that while interests rate risk will directly influence on debt payments and on the income for
    financial institutions, it will have an indirect effect of for example a house producer. High interests rates will result in decreased
                               16
    demand for new houses.
 Galitz (1994) gave a similar example for commodity prices. An oil producer is exposed to the price of oil. However, an airline
    company's profitability will also be exposed to the price of oil, only with the opposite sign.
Smith, Smithson and Wilford (1988) said that a risk profile could be made to visualise the risk exposure.




                                          Change in value of the firm.

13
     This happened for the defence industry after the collapse of the Soviet Union.
14
     All of these where present in the Barings Bank bankruptcy.
15
    Black, Scholes and Merton, who developed the model, tried to use it to earn money through derivatives trading, but ended up
  loosing money.
16
   This means that even if the firm has no debt it can be exposed to interests rate risk. In Froot, Scharfstein and Stein's framework it is
  assumed that the firm uses little debt (as well as external equity), but this shows that such a policy does not eliminate int erests rate
  risk.
                                                        Change in the value of the risky asset.



Figure 1 Risk profile.

2.3           Hedging. Definition and Objectives.
Smith and Stulz (1985) defined hedging this way: «Firm a hedges more with respect to state variable i than firm b if the absolute value
of the covariance of the value of firm a with state variable i is less than or equal to that of firm b. Therefor, hedging reduces the
dependency of the value on changes in the state variable.»

Galitz (1994) listed these as hedging objectives:
 Protection against any movement in prices. In this case FRAs , SAFEs, futures, forwards or swaps can be used
 Partly protection against risk. In this case, options or option like derivatives, like for example caps or floors, can be used.

Linear risk management instruments, such as futures , are often referred to as hedging instruments, while non-linear instruments, such
as options, are often called risk insurance. In this thesis, all such instruments are however referred to as hedging instruments.

Smith, Smithson and Wilford (1988) said that financial price risk is a necessary, but no a sufficient condition for hedging if the firm's
object is to maximise its expected value. A sufficient condition for hedging is that it will increase expected firm value. Smith, Smithson
and Wilford also said that if a cashflow model is used, the benefit of hedging could not lie in the discount rate . The discount rate must
                                                                                               17
be irrelevant, since the risk is diversifiable. According to Modigliani and Miller (M&M) the effect must therefor lie in one of the
following three.
 Taxes.
 Transaction costs
 Investment decisions.

Smith, Smithson and Wilford (1988) explained in more detail:
 With linear tax code there is no effect of hedging. With convex tax code the benefit of hedging will depend on the follow ing
    things.
     The progressively of the tax code.
     The tax loss carryforwards.
     The investment tax credit.
     Minimum taxes.

      Transaction costs influence expected costs of financial distress . The default risk depends on the volatility of the income and the
       size of fixed claims. The costs of financial distress depend on the cost of reorganisation, liquidation and the legal costs. In the
       M&M framework, there were fixed investment policies, the firm should accept all positive NPV projects. The underinvestment
       problem means however that the firm does not undertake positive NPV projects if the debt/equity ratio is to high, because the gain
       goes to the debtholders.

      In Froot, Scharfstein and Stein's framework, the investment decision is the rational for hedging. Gèczy, Minton and Schrand (1997)
       agreed with Froot, Scharfstein and Stein when they claimed that imperfect capital markets create incentives to hedge.

Froot, Scharfstein and Stein's framework distinguished itself from other hedging theories. While Froot, Scharfstein and Stein 's
framework focused on optimising future investments, other theories are aimed at stabilising the cashflow. A very good example of the
                                                                                                                     18
widespread attention toward the latter objective is the recent popularity of the Value -at-Risk, VaR, measurement. This way of
quantifying risk focuses only on variations in the cashflow, not the optimal use of those cashflows. This stabilisation -view on risk
management is accepted among academics as well as practitioners.

2.4           Other Hedging Theories.
Several theories have been developed that claim that the firm should hedge all variations in the cashflow, not optimise it as Froot,
Scharfstein and Stein's framework suggested. The following is a list of the most important of these theories.
 Smith and Stulz (1985) made a theory based on risk adverse managers who owned large numbers of shares in their firms. It is more
   costly for these managers to hedge themselves so in order to avoid volatility in their stock portfolio they hedge. However, Gèczy,
   Minton and Schrand (1997) found no empirical evidence for this theory.
 Smith and Stulz (1985) also argued that shareholders can benefit from hedging if the taxes are progressive, there are tax loss carry-
   forwards or minimum taxes. Nance, Smith and Smithson (1993) found evidence for this theory.
 Smith and Stulz (1985) also said that debtholders favour hedging, since it reduces the probability of bankruptcy. In addition to the
   direct and indirect costs of bankruptcy, Shapiro and Titman (1986) said that a bankruptcy means that the firm looses long -term

17
     About Modigliani and Miller, see section 12.2.1, «Modigliani and Miller.», page 71.
18
     For a summary of the VaR concept, see section 12.1, «Value-at-Risk.», page 70.
    relations with suppliers and customers. These relations have a value and a loss of these relations is therefor costly. Tufano (1996)
    found however no evidence that firms hedge to avoid bankruptcy.
   DeMarzo and Duffie (1991) used another argument for why the shareholders can benefit from hedging. The y said that the firm
    should hedge if the managers have private information of unobservable risks. Gèczy, Minton and Schrand (1997) found no empirical
    evidence for this theory.
   Campbell and Kracaw (1993) said that all stakeholders benefit from risk management. According to investment theory, investors
    should diversify their investments to avoid unsystematic risk. Other stakeholders, like employees, managers and suppliers do not
    have that opportunity and they will benefit from risk management. Gèczy, Minton and Schrand (1997) found no empirical evidence
    for this theory either.

The empirical evidence show that the financial theories that pres cribe full hedging do not hold when tested empirically. This means
that one should look for theories with an alternative viewpoint. With this in mind one should approach Froot, Scharfstein and Stein's
framework with an open mind, ready to absorb new insights into the field of financial theory.


                                            3.          THE PROBLEMS.
In order to evaluate the models that Froot, Scharfstein and Stein came up with in their paper, six main problems are answered .
   Problem I:        Are the Costs of External Capital Necessarily High?
   Problem II:       What Are the Costs of Froot, Scharfstein and Stein's Framework?
   Problem III:      Will Stakeholders Approve?
   Problem IV:       What Makes the Firm Invest?
   Problem V:        What Instruments Should Be Used?
   Problem IV:       Does Empirical Evidence Support Froot, Scharfstein and Stein's Framework?

These questions are answered with both theories and empirical findings.

3.1         Problem I. Are the Costs of External Capital Necessarily High?
Earlier research on the cost of capital concluded that external capital is more expensive that internal capital. This formed the core of
Froot, Scharfstein and Stein's framework. They assumed that external capital is more expensive than the sum of the costs of internal
capital plus the costs of derivatives hedging. In order to evaluate Froot, Scharfst ein and Stein's hedging strategy this assumption is
questioned. Are there situations where a firm might just as well issue new capital? Resent research has cast a new light over this issue.

3.2  Problem II. What Are the Costs of Froot, Scharfstein and Stein's
Framework?
It is beyond the scope of this thesis to quantify the costs of hedging versus the cost of issuing equity or debt. This proble m is
restricted to only cast light over the issue and on that ground make a qualified comment of whether the cost of hedging is a problem for
Froot, Scharfstein and Stein's framework.

3.3         Problem III. Will Stakeholders Approve?
In the introduction it was stated that all stakeholders would try to maximise their own utility. Before the manager will approve to Froot,
Scharfstein and Stein's framework he must be convinced that he will not come out worse with the strategy implemented than he would
if the firm was to rely on «old-fashion» capital acquisitions to invest. Other stakeho lders' views of the hedging strategy are also of
importance. In its strongest form, this question actually asks whether Froot, Scharfstein and Stein's framework is a Pareto o ptimal
financial policy.

3.4         Problem IV. What Makes the Firm Invest?
Froot, Scharfstein and Stein assumed that changes in some risky assets cause the firms' business environment to change in suc h a way
that it is profitable for the firm to change its investment level and that it is possible to trade derivatives on these assets. First, it is
identify what causes the firms' business environment to change in such a way that the firms want to invest. Is it really the case that
changes in one or more risky assets is the most important determinant for changes in the firms ' investments?

3.5         Problem V. What Instruments Should Be Used?
Froot, Scharfstein and Stein just took for granted that there exist derivatives with the right properties to fit their framew ork, although
they did not identify the risky assets that the derivatives should be written on. These assets were identified in Problem IV. This
assumption is questioned in two ways.

3.5.1        Do the Right Derivatives Exist?
Are there derivatives written on the risky assets identified in Problem IV? Not only must there exist derivatives on those risky assets,
those derivatives must also have certain properties when it comes to liquidity, price and lifetime.
3.5.1.1         A Perfect Hedge.
                         19
As will be shown later, derivatives markets have become very large in recent years. W ith such a large market, is there any reason to
believe that a firm will have any difficulty finding the right derivatives? If the firm can not find the perfect derivatives, it can always find
a close substitute, one might assume. A close substitute may ho wever not necessarily be good enough. Campbell and Kracaw (1993)
said that in order to eliminate all risk, the hedge has to be perfect. A perfect hedge is one where the value of the derivative moves
exactly similar or opposite of the underlying asset, all the time. Although Froot, Scharfstein and Stein did not want to eliminate all risks
in all states, the perfect hedge is still of interest. This is because the degree of hedging should be controlled through the hedge ratio,
not the degree of perfectness in the hedge. If the optimal hedging ratio is 0.5for example, the firm wants to remove 50% of the variation
in the cashflow from assets in place. With a less then perfect hedge, the firm may end up removing 40% or 60% of the variatio n in the
cashflow. Froot, Scharfstein and Stein used a considerable part of their paper to show how one should calculate the hedging ratio.
Numerous variables were used, several being first and second order derivatives, to ensure that the hedge ratio is found with great
accuracy. It makes little sense to calculate the hedge ratio with such great accuracy if one do not demand the same degree of accuracy
from the derivatives.
Abken (1994) said that a dynamic hedge uses existing futures and options to synthetically create the desired hedge. The composition
of the derivative basket must be adjusted over time. This type of hedging is complex and risky. Because of this, Abken says, most
traders prefer over-the-counter (OTC) instruments to dynamic hedging.

3.5.1.2         The Derivatives' Time to Maturity.
The lifetimes of the derivatives are of great importance because history has shown that using derivatives with the wrong matu rity can
be risky. Mork (1992) said that business cycle recessions typically lasts for a half to two years and booms can last for up to 10 years. In
addition, DeGrauwe (1996) said that exchange rate cycles are often even longer than business cycles . If the risk that correlates with the
cashflow from assets in place is a macroeconomics variable, the firm must either find long lasting derivatives or roll over s hort lived
                                           20
derivatives, which can be very dangerous.

3.5.2         Can Other Derivatives Be Used?
Froot, Scharfstein and Stein limited their discussion to straight futures , forwards and options . It is however likely that other
instruments and more complex hedging strategies will further increase the usefulness of Froot, Scharfstein and Stein's framew ork. This
is likely since the use of financial engineering has increased enormously during the last 10-15 years, as has the range of instruments.
This is especially important if one extends the framework to an intertemporal one. Tufano (1996) found that even the gold mining
industry, with its fairly simple product, use a wide variety if risk management instruments, such as options, warrants, bullion loans,
forwards, spot deferred contracts and gold-linked equity.

3.6    Problem VI. Does Empirical Evidence Support Froot, Scharfstein and
Stein's Framework?
The questions raised above evaluates the usefulness of Froot, Scharfstein and Stein's framework in an indirect way and is answered
with the use of financial theories and empirical findings in fields of finance that covers the assumptions behind Froot, Scharfstein and
Stein's framework, not their framework per se. There are a number of resent papers that covers Froot, Scharfstein and Stein's framework
specifically, together with other risk management theories. After all, it is what works that is interesting, not what is supposed to work.


        4.         ABSTRACT OF FROOT, SCHARFSTEIN AND STEIN'S
                         PAPER, WITH EXTENSIONS.
This section includes an abstract of Froot, Scharfstein and Stein's framework. A couple of things have however been added. Th ere are
variations of the framework that follow directly from the set-up of the framework that Froot, Scharfstein and Stein did not comment on.
These new models have been added in this section. In addition, one of the assumptions that they made regarding the distributi on of
risky assets holds only in the short term. If one view Froot, Scharfstein and Stein's framework in a long er perspective the conclusions
that they made about the optimal choice of derivatives change. This too has been added.

Froot, Scharfstein and Stein focused a lot on the technical parts of their framework, in particular the calculations of the h edge ratios.
The hedge ratios are to a great extent dependent of how the cashflows from assets in place and the investment opportunities correlate
with each other. To simplify the extensions of fssfin most cases the extreme assumption s that the correlation coefficients are either 1, 0
or –1 have been made. This makes it easier to understand the impact of the different correlation coefficients. This simplificatio n does
not however influence the conclusions; they only make it easier to u nderstand the models. Other hedge ratios will naturally be found in
the real world but this will not change the conclusions.



19
     See section 9.1, «How Derivatives Are Used.», page 52.
20
     One of the most famous derivative disasters is the $1.3 bill loss that almost brought Metallgesellschaft, Germany’s 14th -largest
     industrial group with 58.000 employees, to bankruptcy. Metallgesellschaft had ten years contracts with its customers of fixed-price
     oil delivery. To hedge against oil price declines, the firm used three month contracts and rolled them over. In time these sh ort-term
     contracts would equal a ten year forward. The firm was however exposed to basis risk. When the oil prices fell, the margin calls
     exceeded one billion dollars. The firm had serious problems with meeting the cash requirements.
4.1           The Origin.
Froot, Scharfstein and Stein claimed that even though financial theory has done a good job explaining the how's of risk management,
there is less understanding of the why's. Most other risk management theories come with the extreme conclusion that the firms should
hedge fully, removing all risks from the cashflows. In many cases this is however not optimal because such an approach does not take
into account the changes in the environment in which the company operates.

The conclusion about how much a firm ought to hedge comes from Froot, Scharfstein and Stein's mathematical equations. The ide as
that formed the foundation of the framework were actually borrowed from another important financial theory, that is that of Myers and
                                                                                  21
Majluf (1984) which formed the basis for Myers' (1984) Pecking order theory. Myers said that there is a pecking order that ranks the
different sources of capital that the firms can use. External capital is more expensive than internal capital because of such costs as
transaction costs, agency problems and as ymmetric information. Because of this, the firms will first rely on internal capital, which is the
cheapest form of capital for the firm. Debt is the second cheapest and outside equity is the most expensive form of capital. Myers said
that the firm would only rely on external equity when it can not use any of the other two. Even this was nothing new when Mye rs
published his theory. Bethel, Liebeskind and Opler (1998) said that during the 60's and 70's firms diversified to crea te self-sustainable
sources of financing to avoid having to rely on what were considered to be relatively costly and inefficient external sources of capital.
Myers said that «If the dividend payoffs and a reduced profitability mean that the firm does not h ave enough cash from operations for
its investments, it will draw on its cash balance or marketable securities portfolio.» Actually, Froot, Scharfstein and Stein 's contribution
to financial theory was to say that the firm should not just hold any securities. The sum of the firms' cashflow from ongoing operations
and from the derivatives should be optimised with respect to the investment and financing opportunities. Whenever the cashflo ws
from assets in place are not optimally distributed with respect to the investment and financing opportunities the «marketable
securities», i.e. the derivatives, should be used to fill the gap. This means that the sum of the value of the internally generated
cashflows and the derivatives must correlate with the investment and financing decisions. This sentence sums up Froot, Scharfstein
and Stein's approach.

4.2           The Framework.
Froot, Scharfstein and Stein assumed that market imperfections make external capital more costly than internal capital. If there are
variations in the cashflows from assets in place and the firm does not hedge according to Froot, Scharfstein and Stein's framework,
there will be variations in the cheapest form of capital that the firm has at its disposal. Variations in the cashflows from assets in place
will result in one of two things.
 Variations in the investment, if the firm only uses internal funds. Froot, Scharfstein and Stein assumed that the different p arts of the
    corporation are not stand-alones. This means that the management has the ability to allocate funds from one part of the firm to
    another. In this case, the different segments of the firm will be interdependent.
 Variations in the amount of capital raised externally.
Below it is shown that neither of these options are optimal policies.

Froot, Scharfstein and Stein made the not too controversial assumption that the profit function, P, is convex with respect to the
investment amount, I. If the firms choos to use only internal capital from the cashflows from assets in place, w, which fluctuates, the
investment amount will fluctuate too. The profit will thus not be maximised. The first derivative of the profit function with respect to
internal capital, Pw, is negative and the first alternative mentioned above is therefor undesired.
Froot, Scharfstein and Stein also assumed that the cost of external capital is convex with respect to the amount of external capital, e.
Convex costs of external capital can be interpreted in the following way: In additional to paying interests rates on debt or dividends on
equity, there are additional dead-weight costs, C(e), associated with external capital. These can be costs associated with increased
financial distress, increased probability of bankruptcy, asymmetrical information between investors and the managers or the managers'
compensation for making their undiversified risks greater. The first derivative of the dead -weight costs with respect to the amount of
external capital, Ce , is thus positive. If this is the case, the second alternative is also undesired.

To show this Froot, Scharfstein and Stein used a net present value (NPV) model. The net present value of the investment is:


                                                             F(I) = f(I) - I, where

F(I)     = The net present value.
f(I)     = The present value.
I        = Investments, where

                                           I = w + e (= the sum of internal and external capital)

The profit function for the firm is thus

                                                 P = MAX [F(I) - C(e)] = MAX [f(I) - I - C(e)]

The first order condition for this problem is:


21
     About the Pecking order, see section 5.4.5.1, «The Pecking Order.», page 36.
                                                           FI = fI - 1- Ce = 0, where

FI        = The first derivative of the net present value with respect to investments.
fI        = The first derivative of the present value with respect to investments.
Ce        = The first derivative of the dead-weight costs with respect to external capital.

Rewritten, the FOC becomes:

                                                                 FI = fI - 1 = Ce

In a market without imperfections, the dead-weight cost of external capital would be zero. In that case
fI = 1 and the firm would invest optimally. With Ce > 0 the firm will underinvest, since fI > 1.

4.3              A Two-Period Model.
Froot, Scharfstein and Stein used a two-period model. At time zero the firms have to make its hedging decision, so that the amount of
internal capital at time one is optimal with respect to investment opportunities at time one. At time one the firms make their investment
                                                                           22
decisions, based on the capital available and the investment opportunities. At time two the firms repay their debt/equity holders.

 Time zero                                   Time one                                       Time two
 Make hedging decision.                      Observe investment opportunities .             Repay investors.
 Operate assets in place that will give      Receive cashflows from assets in place
 stochastic cashflows in period one.         and from the derivatives.
                                                                               23
                                             Maybe acquire external capital.
                                             Make investment decision.
                                             Make investment.
Table 2 Froot, Scharfstein and Stein's framework.
It may seem like a contradiction that the firm considers acquiring external capital at time one, since avoiding external capital is what this
framework is all about. This is however not necessarily the case. The goal is to maximise profits. This means that the costs of capital
should be as low as possible with respect to optimising the profits, which is not necessarily equal to not acquiring any external capital
at all. If the marginal income > the marginal cost, including the financing costs, the firms would be better off acquiring external capital.
The hedging strategy is aimed at avoiding unnecessary large acquisitions, not avoiding all acquisitions.

The cashflows from assets in place are w. The values of the risky assets that causes w to change are w. Any deviation from E(w)
causes the cashflows from assets in place, w, to change by an amount that is a function of  - E(). In the following, it is assumed that

                                                        Corr((  - E(  )), w) = 1 or -1,

even if this is not the case in the real world. Froot, Scharfstein and Stein used a great part of their paper to show what ha ppens if this
equation is not true. This is however of no interest in the outline of the following models. The equation above means that any change
in the risky asset is transformed into an equally large percentage change in the cashflows. The effect may however be differe nt for
different firms. An airliner and an oil producer will for instance have the opposite exposure to changes in the price of oil, which is why
it is assumed that the correlation coefficient is either –1 or 1.

The cashflows from assets in place correlates with the risky assets either because the unit revenues or the unit costs correlate with the
risky assets. The notation R is used when the risky asset changes the firms' unit revenue and thus the amount of liquid assets. The
change can be either positive, higher unit revenue, or negative, lower unit revenue. C will be the notation for a risky asset that changes
the firms' unit costs. If it is of secondary importance whether the changes in the amount of cash flows from assets in place comes from
changes in the revenue or from changes in the costs, w will be used to denote the risky asset. If one looks only at the extreme
situations four things can happen.




1
     24     R      Revenue             CORR (w, w)                = 1




22
     The core of Froot, Scharfstein and Stein's framework is to make these two correlate as much as possible.
23
     This should of cause be minimised, preferably eliminated.
24
     This table should be read the following way:
2          R      Revenue             CORR (w, w)                  = -1
3          C      Costs               CORR (w, w)                  = -1
4          C      Costs               CORR (w, w)                  = 1

Table 3 Sources of correlation between the risky asset and the cashflows.
The following illustrations show a multiperiod setting of Froot, Scharfstein and Stein's framework.


                                                                      and w
                                                                                    Investment opportunities
                                                                                    (Time)
Figure 2 Situation 1 and 4.

                                                                     

                                                                                    Investment opportunities
                                                                    w               (Time)
Figure 3 Situation 2 and 3.

4.4             Hedging in Froot, Scharfstein and Stein's Framework. 25
It is possible, by assumption, to trade derivatives written on the risky asset. The firm can thus redistribute the distribution of the
amount of liquid assets it has at time one in the two period model so that the amount of liquid assets correspond to the capital need,
given the fixed investment opportunities . The sum of the cashflows from assets in place and the money that the firms get from its
hedging activities is the internal capital redistributed with respect to the investment opportunities. This capital is denoted w R, since it is
               26
Redistributed. wR is thus the amount of internal capital the firm has at its disposal at time one for investments if it hedges and does
not acquire external capital. Without hedging, the firms will in some states need a lot of costly external capital. At time zero the firm
faces the following possible futures :

                   The risky asset changes in such a way that the cashflows increases.
                  The risky asset does not change, so cashflows does not change.
                   The risky asset changes in such a way that the cashflows decreases.
Figure 4 Stochastic distribution of the cashflow.
If the firm does not hedge and the cost of external capital is convex. Some states will be very expensive, either in payoff to the new
debt/equity holders, or as lost investment opportunities . Hedging will make all states more equal when it comes to external financing
and will thus reduce the cost of capital. In some cases hedging will give the firm less internal capital than if the firm had not hedged. In
those states, the firm would have been better off if it did not hedge. However, since the dead -weight costs are convex the bad states
are more bad then the good states are good and the informed investor should be willing to accept some losses in some states in order
                                              27
to avoid even greater losses in other states.

4.4.1           The Simple Model.
In this thesis this model of Froot, Scharfstein and Stein's framework is simply referred to as the «simple model». In this model the
investment opportunities are fixed. This model follows directly from the set-up of the framework. Since the cashflows from assets in
place are stochastic and the need for the money, the investment opportunities, is non -stochastic the firm will have too little money in
some states and more money than it actually needs in other states. Because of the dead -weight costs associated with external
financing the lack of match between the liquid assets and the investment opportunities means that profits are not maximised.

As opposed to the other models in Froot, Scharfstein and Stein's framework there are actually two different ways that the fir m can
hedge in the simple model.




Case 1: If a raise in the risky asset implies that the firm's revenue increases, this implies that the correlation between th e risky asset and
the amount of liquid assets that can be used for investments is 1. Correspondingly, if a raise in the value of the risky asset causes the
revenue to fall the correlation is –1.
25
   This section is really not an abstract, it only explains some features of Froot, Scharfstein and Stein's framework. However, only the
notation wR is added.
26
   This R has nothing to do with the revenue.
27
     In addition, this kind of hedging will make it difficult for the manager to overinvest in those periods where the cash flow f rom
    operations are greater than what is actually needed for optimal investments.
The simplest way for the firms to remove variations in the cashflows is to fix their costs and revenues. This can be done eit her by
                                                                   28                                                           29
going into long-term contracts with the suppliers and customers or by useing forward contracts with the business partners. This
way the prices of both the in- and outputs will be fixed no matter what happens to the prices of those commodities in the world market.
By the same way of reasoning the firm could also fix other costs as well, like using only fixed rate debt. Such a hedging strategy is
simple because the firms do not need to worry about what causes the variations in the cashflow.

                                                                  30
The other alternative is to take a position in the futures market or to use OTC forward contracts with the opposite exposure than what
the firms initially had. The cashflows from operations and the cashflows from the futures or forwards will cancel each other out. If one
uses a firm in the oil exploration industry as an example, a firm that is exposed to the price of oil but where the main asse t of the firm is
knowhow and technology, like the Norwegian PGS. Such a firm must continue to invest in R&D no indepen dently of the current price
of oil. The investment opportunities are therefor fixed. A fall in the price of oil will decrease the demand for the firms se rvices, due to
decreased investments by oil producers, and thus reduse the cashflows from assets in pla ce. If that firm holds oil futures contracts that
gives it the right and obligation to sell oil at the average oil price this will solve the problem for the firm. These contra cts can be sold or
settled financially. If E(oil price) is 18 USD and the oil price falls to 12 USD those contracts will be worth 6 USD. The low proceeds from
operations plus the cashflows from the derivatives will enable the firm to coninue its R&D at the same level as before the de mand for
the firm`s services dropped.

The following illustrations show a multiperiod setting of Froot, Scharfstein and Stein's framework.
                              (But the fluctuations do not matter)          wR through a long-term contract
                                                                             or an OTC forward contract

                                                                                     Investment opportunities

Figure 5 The Simple model using long-term contracts.

                                                                   and w           wR

                                                                                     Investment opportunities
                                                                       Futures contract
Figure 6 The Simple model using futures.


4.4.1.1          Adverse Protection in the Simple Model.
The Simple model is based on the fact that the firms face non-stochastic investment opportunities . Linear hedging, using futures ,
forwards or long-term contracts, will be sufficient if the investment opportunities are fixed forever with 100% certainty. This
assumption is of cause not very realistic. One must assume that there are boundaries to how much the value of the risky asset can
change before the investment opportunities can no longer be regarded as fixed. Froot, Scharfstein and Stein said that while l inear
hedging using futures and forwards «can add value, they generally will not maximise value if other, non-linear instruments, such as
options, are available.» Froot, Scharfstein and Stein said that the firms should protect themself against very large adverse changes in
the risky asset. To accomplish this the firms should add non -linear hedging instruments, such as options. In this thesis this type of
hedging is called «adverse protection». This is of cause only an extension of the simple model.


                                                        Option              wR through a long-term contract
                                                                             or an OTC forward contract

                                                                                Investment opportunities

Figure 7 Linear hedging with added adverse protection.


4.4.1.2          No Protection in the Simple Model, Out of Boundary Exposure.
31
  One can however just as well make the opposite assumption as in the «Adverse Protection in the Simple model»-section described
above. If the risky asset changes too much in the adverse direction the firms will probably consider to scale down the production. I n
this case the investments will be negative. Likewise, if the risky assets change a lot in the favourable direction the firms will want to
increase the output to take advantage of the new situation. This shows that linear hedging will only be beneficial within cer tain
boundaries. Above and below these boundaries it would be optimal to stay unhedged. To cancel out their linear hedging position the

28
     See section 9.3, «Long-term Contracts.», page 54.
29
     See section 9.4.4.1, «Forwards on Commodities.», page 55.
30
     Se section 9.2.1, «About Futures.», page 53.
31
  While the basic simple model and the adverse protection model were explained by Froot, Scharfstein and Stein, the following model
 is one of my extensions.
firms can use deep out of the money call and put options . These options are not intended to give added protection, as they did above,
but to cancle of the linear hedging so that the firms stay unhedged above and bleow the bou ndaries. It seems fair to assume that the
borders of the boundaries between where linear hedging is optimal and where staying exposed is optimal will be smooth. Becaus e of
this exotic options should be considered. Since these options are supposed to cancel out the linear hedging in the Simple model this
type of hedging strategy is called «Out of boundary exposure» in this thesis. Just like the adverse protection model mention above this
is also just an extension of the simple model.


                              Option                                   wR through a long-term contract
                                       Boundary                         and an option

                                                                            Investment opportunities

Figure 8 Out of boundary exposure.


4.4.2              Changing Investment Opportunities.
In this model the investment opportunities are no longer non-stochastic. It is assumed that there is a marketable risk, D, that correlates
with the variations in the demand. D is assumed to have a mean equal to one and standard deviation equal to . One can interpret this
as follows. The risky asset changes the demand for the goods t hat the firms produce. The changes in the demand may not necessarily
mean that the customers demand more or less of the goods, it may just as well be that the customers demand different properties of the
goods. In such a case the changes in the investments will not increase or decrease in the amount of output, only the quality of the
products will change. The important thing is that the changes in the demand means that it is optimal for the firms to change the
investments. Whether these investments results in increased output or increased product quality is of less importance.

This model of Froot, Scharfstein and Stein's framework comes in three different versions. In the first the firm should not he dge and in
the second the firm should actually speculate. In both of these models the risky asset that changes the amount of liquid assets,
through changes in the revenues or the costs, also changes the demand. It is the same asset that causes both changes. The cor relation
between the amount of the liquid assets in the firm and the investment opportunities is consequently equal to one. The third model is
similar to the ones above with one important difference. In this model there are two risky assets. One risky asset causes the amount of
liquid assets to fluctuate and the other causes the demand to fluctuate. The correlation between the two assets is < |1|.

4.4.2.1             The No-Hedging Model.
There are situations where, according to Froot, Scharfstein and Stein, the firms should remain exposed to the risky asset and not
hedge. This hedging strategy greatly contradicts other risk management theories. An oil producer has a different exposure to the price
of oil than the oil service firm mentioned above. An oil producer will face increasing/decreasing marginal profits of additional
                                                                                                                          32
investments at the same time as the price of oil and thus the liquid assets available for investments increases/decreases. This means
that the investment opportunities correlate with the amount of the cheapest form of capital the firm has, internal capital. Hedging will
therefor be counterproductive because it will lead to a situation where the firm has too much money when the investment opportunities
are low and too little money when the investment opportunities are good. This clearly shows the benefits of Froot, Scharfstei n and
Stein's hedging strategy compared to other hedging strategies where the objective is to stabilise the cashflow. Such a hedging
strategy will actually decrease the oil producer's future earnings because it will make investments more expensive if the firm has to
acquire expensive external capital.

One can look at the no-hedging model as a special case of the «Out of boundary exposure model». In the no-hedging model the range
where the firm's investments are fixed is set to zero.
                                                                   , w and investment opportunities




Figure 9 Correlated investment opportunities and cashflows.



a
     33, 34   D      Demand   CORR (D, Investment opportunities) = 1
b             D      Demand   CORR (D, Investment opportunities) = -1

32
   This was actually the case during the summer of 1998, when the historically low oil prices made all three Norwegian oil companies,
  Norsk Hydro, Saga and Statoil, reduce their activities on new oilfields in the North Sea.
33
   This table should be read the following way:
  Case a: If a raise in the risky asset implies that the demand for the firm's products increases, then the correlation between the risky
  asset and the investment opportunities is one.
34
   D is the value of the risky asset that changes the demand for the firm's goods.
Table 4 The risky asset and the investment opportunities .


4.4.2.2            Adverse Protection in the No-Hedging Model.
35
  While Froot, Scharfstein and Stein did not mention it, the adverse protection strategy makes perhaps even more sense here than in
the Simple model. The use of options will give protection against any large adverse changes in the risky asset. Without a safety net in
the form of options a large change in the risky asset may be devastating.
                                                                    , w and investment opportunities
                                                Option


Figure 10 The No-hedging model with added adverse protection.


4.4.2.3            The Speculation Model. simen
In yet other cases it may actually be beneficial to overexpose the firms to the risk that correlates with the amount of cashflows from
assets in place. This is the case in good states when an increase in the risky asset increases the investment opportunities to a level
higher then what the internal wealth can finance. The investment opportunities are very good and the cashflows form assets in place
are not enough to finance the investments. This will be the case if the changes in the demand, and thus the investment opport unities,
are more sensitive to changes in the risky asset than the cashflows are. A decrease in the interests rate will be positive for the amount
of liquid assets for a leveraged house constructor. The demand for new houses may however change by more than the chan ge in liquid
assets. In this model elasticity is a key word. This version of the changing investment opportunities model is called the «Speculation
model». Theoretically both linear and non-linear instruments can be used. Using linear instruments would however be extremely risky.
The firm can stay unhedged for values when the risky asset stays around its mean. An in the money call will be valuable when the
risky asset increases in value. This means that the firm gets money from assets in place at the same time as the option becomes
valuable. A safety net in the form of a deep out-of-the money option will reduce the risk in this model in the same way as for the other
models.

             w and investment opportunities
                                       Option
                                                                         


Figure 11 Highly sensitive demand and less sensitive cashflow. The Speculation model.


4.4.3           The Uncorrelated Model.
There may be situations where the amount of internal capital is uncorrelated with the investment opportunities and both fluctuate. This
will be the case when the cashflows from assets in place correlates with one risky asset and the investment opportunities correlate with
another risky asset and the two assets do not correlate.
In such a case the firm should hedge fully. The reasoning is the same as for the simple model. In some states hedging will give the firm
less internal capital than if it did not hedge. This means that the firm would have been better off if it did not hedge. However, in other
cases the firm would be better off with hedging. Since the dead-weight costs are convex the bad states are more bad then the good
states are good. Applying Froot, Scharfstein and Stein's framework will still be beneficial for the firm, even if hedging is
counterproductive in some states. In this thesis this version of the changing investment opportunities model is called the
«Uncorrelated model».

By combining tables 3 and 4 one gets the relationships between the risky asset, unit revenues, unit costs, the amount of liquid assets,
demand and investment opportunities showed below. Of special interest are of cause the risky asset, the investment opportunities and
the liquid assets. D is the risky asset that changes the demand and thus the investment opportunities. R and C changes the firms
revenues and costs, respectably. These variables changes the firms cashflows from assets in place.
 1a D   Demand   Inv. opp.  And Unit revenue                                            w0 
2a      D             Demand               Inv. opp.      And    Unit revenue             w0    
3a      D             Demand               Inv. opp.      And    Unit cost                w0    
4a      D             Demand               Inv. opp.      And    Unit cost                w0    
1b      D             Demand               Inv. opp.      And    Unit revenue             w0    
2b      D             Demand               Inv. opp.      And    Unit revenue             w0    
3b      D             Demand               Inv. opp.      And    Unit cost                w0    
4b      D             Demand               Inv. opp.      And    Unit cost                w0    


35
     This model is added by me and is not an abstract of Froot, Scharfstein and Stein’s paper.
Table 5 The investment opportunities and the w in the no-correlation model.
In the no-hedging and speculation models R = C = D. Only one risky asset moved both the cashflows from assets in place and the
investment opportunities .

Just as options can increase the usefulness of Froot, Scharfstein and Stein's other models they can also increase the usefulness of this
model. The firm will benefit from having extra protection when the investment op portunities are good and the cashflows from assets in
place are low. The linear hedging will help somewhat, but options will help the firm further in this highly undesired situation. The
options should yield a high payoff when this situation occurs. This means that the payoff must be function of both risky assets.

Even though this version of Froot, Scharfstein and Stein's framework was mentioned last, it is actually the most general. All the others
are special cases of this one.

4.4.4         The Reversed Simple Model.
36
   Viewing Froot, Scharfstein and Stein's framework as different cases of the uncorrelated model makes it evident that they missed one
possible state. What if w is almost or completely non-stochastic and D fluctuates randomly? In this case the amount of liquid assets is
fixed, but the investment opportunities varies. C = E(C) and R = E(R). From the same reasoning as in the simple model, one will see
that the optimal hedging strategy will be to change the distribution of the liquid assets so that the amount fits the investment
opportunities. The firm will use the derivatives to move away from its naturally complete hedged situation to a situation where the
                                                                                                                37
amount of internal capital correlates with the investment opportunities. This will be a fixed for floating swap. This of cause violates all
«common sense» if one is used to the ordinary way of thinking about hedging. The natural hedges that the non -stochastic amount of
liquid assets would be viewed as a perfect situation. However, according to Froot, Scharfstein and Stein's framework it will be a
situation where the profits are not maximised. The reason is of cause that in order to exploit the exceptionally good investment
opportunities that will occur from time to time the firm would have to acquire external capital. Alternatively, the firm would have to
forgo those investment opportunities. Both of these options will result in less than maximum profits. This model is called the
«Reversed simple model».
                            , investment opportunities and wR through speculation.

                                                                              w



Figure 12 The Reversed simple model.

4.5           The Hedge Ratio.
In the examples above the arguments have been limited to the extreme situations where the correlation coefficients have been 1, 0 or -1.
Froot, Scharfstein and Stein spent a great deal of space in their paper outlining the intermediate cases. The company should hedge
partly in the intermediate cases, using a hedge ratio of less than one. Period one capital is given by:

                                                        wR = w0 [h + (1 - h) X], where

h      = The hedging ratio.
wR     = The redistributed cashflows from assets in place with respect to X. wR is the sum of w and the cashflows from the derivatives.
w0     = The expected cashflows from assets in place, w 0 = E(w).
X     = The value of the risky asset. The risky asset is either D, C or R, which is the risky asset that either changes the demand (the
          investment opportunities), the costs or the revenue. The last two can be the same asset in which the risky asset will change
          the amount of internal capital, w, which is thus denoted w.

Assume first that X = w. A hedging ratio of one gives a complete elimination of all risks due to changes in unit revenue and unit costs.
This is the hedging policy used in other hedging theories as well as in Froot, Scharfstein and Stein's simple model and the uncorrelated
model. This will be optimal when there is no correlation between the cashflows from assets in place and the investment opportunities.
A hedging ratio of zero leaves the amount of internal capital completely exposed to the changes in the risky variable, w, like in the no-
hedging model. In the speculation model h < 0.
If X = D the equation fits the reversed simple model. In this reversed simple model the investment opportunities are stochastic and the
cashflow is fixed. In this situation a hedge ratio of one still gives the firm a complete elimination of all variations in the risky asset. This
is however accomplished without the use of derivatives to redistribute the cashflow since it was assumed that the cashflows f rom
assets in place were non-stochastic to start with. Therefore, the firm arrives at a fully hedged situation without hedging. This is
however not optimal, since the investment opportunities are stochastic. To understand this seemingly puzzling situation one must
realise that a hedge ratio of one means that the «speculation ratio» is zero. In order to make the amount of internal capital fit the
investment opportunities the firm actually has to speculate. In order to accomplish this the speculation ratio must be set to one and the


36
     This model is added by me.
37
     About swaps, see section 9.8, «Swaps.», page 61.
hedging ratio and thus h, is therefor equal to zero. In the reversed simple model wR = w0 (1 - 0)D gives the optimal distribution of
internal capital with respect to the investment opportunities.

In the changing investment opportunities model the NPV model is rewritten:

                                                          F(I) = f(I) - I, where
                                                         = (  - E( )) + 1, where

     = The correlation between the investment opportunities and the risk to be hedged.
E( ) = The expected value of the risky asset.

The present value term, f(I), is a function of both the correlation between the proceeds form assets in place and the investment
opportunities and how much the risky asset deviates from its average. Inserting
 = ( - E()) + 1 into f(I) gives.

                         f(I) = Present value * corr(inv. opportunity risk)*(abnormal return) + present value

In the case mentioned above with plant enlargement the correlation is close to one t he investment opportunities are good when there
are a lot of internal cash. If the risk, , is far above its average, E(),  will be large and the firm should invest much.

4.6           Changing Financing Opportunities.38
So far, the cost of external capital has been a function of the amount of the capital. The more money the firm acquires the higher the
interest payments of the debt, or the dividends to equity holders will be. Both will be a linear function of the amount of capital
acquired. In addition, there are dead-weight costs, which are a concave function of the capital acquired.

There may be situations where the cost of external capital is also a function of the risk that correlates with the cashflow from assets in
place and that may or may not correlate with the investment opportunities . If this is the case, the cost of external capital will be even
higher in the state where the firm has little internal capital. The extra cost of external capital means that the firm should use even more
internal capital than otherwise. However, the firm does not have the internal capital required. The firm should therefor hedge even
more. The firm should also hedge even if the investment opportunities are fixed.
The firm should use non-linear instruments, i.e. options , caps or floors, for this hedging.

4.7           Multinationals.
For multinationals Froot, Scharfstein and Stein defined the risk to be the exchange rate . Movements in the exchange rate will change
the investment opportunities . The profit function is:

                                            P(w) = MAX fH(IH) + fA(IA) - IH -  IA - C(e),where

fH(IH)   = The production function at Home, the present value of investment IH.
IH       = Investments Home.
fA(IA)   = The production function Abroad, the present value of investment IA.
IA       = Investments Abroad.
        = (  - E( )) + 1.
        = (  - E( )) + 1.
        = Parameter between zero and one of the sensitivity between foreign revenues and the risk, the exchange rate.
        = Parameter between zero and one of the sensitivity between foreign investment costs and the risk, the exchange rate .

The hedging ratio is how much money is held in each currency. h = 0 means that period zero wealth is held entirely in foreign currency.
Foreign investment costs are thus protected against any change in the risky asset, the exchange rate . If h = 1, all currency is held in
domestic currency.
If  = , the optimal hedging ratio is such that investments both home and abroad are independent of the exchange rate . The shadow
value of capital should be equal in both countries. Linear hedging, forwards or futures, are sufficient in this case.
If  = 0 and  = 1 only the costs of foreign investments are exposed to changes in the exchange rate. Investments become more
expensive, but the revenues from the foreign investments remain unchanged. The firm should hold more domestic currency than
foreign and invest more at home. In this and the next case, options are required for optimal hedging.
If  = 1 and  = 0, only the revenues form foreign investments are sensitive to exchange rate changes. Since the costs of foreign
investments are not sensitive to the exchange rate, there is no reason for holding foreign currencies. The firm should howeve r hold
some foreign currency in the case that the domestic currency depreciates. This will make returns of foreign investments high.

4.8           Further Extensions.
In an intertemporal setting, the use of futures can lead to large margin payments if the value of the risky asset changes a lot. This is a
«vast» of internal capital and an argument against the use of futures.


38
     In order to narrow the scope of this thesis only the simple model and the changing investment opportunities models have been
    evaluated, not the changing financing model. Because of this, not much space will be spent on this model.
Forward contracts do not have this problem, but there is however a larger credit risk associated with the use of forward contracts.

4.9            Implications of Lognormal Distribution of the Risky Asset.
39
   In this section one of Froot, Scharfstein and Stein’s implicit assumptions are questioned: The distribution of the risky asse ts. Froot,
Scharfstein and Stein argued that linear hedging might have too drastic effects if the risky asset moves a lot. In order to optimise the
benefits of the hedging non-linear instruments should be used in the adverse protection models.
It can however be shown that non-linear instruments can be used as a substitute for linear hedging, not merrily in addition to futures,
forwards and long-term contracts. In the following it is assumed that a raise in the risky asset is associated with favourable conditions
for the firm, increasing oil prices for an oil producer for example, and that a fall in the risky asset is associated with un favourable
             40
conditions. If a raise in the value of the risky asset is as likely as fall and the cons equences of a fall is more bad than a raise is good,
because of the dead-weight costs of external capital, this is a reason for hedging. This was Froot, Scharfstein and Stein's argument.
With the use of linear hedging, all variations in the cashflow from assets in place will be removed. This way the dead -weight costs are
removed and the firm's value is maximised said Froot, Scharfstein and Stein.
This is however true only if a raise in the value of the risky asset is both as likely and as large as a decrease. The short-run return of
the risky assets will be normally distributed, true enough, but the long-term value of the risky asset will be lognormal. Since Froot,
Scharfstein and Stein mentioned exchange rates and commodity prices as risky assets in their examples, one must assume that their
framework is developed for long-term hedging strategies. This is true although their framework is only a two -period model. This
assumption is fortified by the fact that the changes in the risky assets are to result in corporate investments. Because of thes e things, it
                                                                                                                                 41
seems fair to assume that it is more correct to assume that the distribution of the risky asset will be lognormal than normal.
The result is that while a long-term raise in the value of the risky asset is as likely as a long-term drop, a raise in the risky asset will be
larger measured in absolute terms. This means that a linear hedging strategy carries with it major po tential «losses» on the upside.
Losses in the sense that profits are forgone.
While linear hedging would be the better choice if the long -term distribution of the risky assets were normal, using options just to cut
                                                                                 42
away the downside will however be better when the distribution is lognormal.

The firm is faced with two choices.
 The first option is to invest in options . The options should cut of the downside of the distribution of the risky asset. The firm
   should have a long put on the risky asset.
 The second choice is not to hedge at all. The firm should use the extra revenues from the good states to finance the extra de ad-
   weight costs associated with raising external capital. This is possible because the revenues in the good states will out -weight the
   losses and, at least parts of, the additional dead-weight costs in the bad stated. This requires that one looks at this in a multiperiod
   perspective, where the firm has the ability to set aside money during the good states. Froot, Scharfstein and Stein came up with this
   rather strange conclusion too, that the firm should manage the risk by not hedging. Their assumptions where however different , as
   were the conditions for not managing the risk.

                                                                                                                                      43
There are two problems associated with relying on cash from good states. The first is the agency problem of overinvestments. If the
manager maximises his own utility he may be inclined to use the extra cash from the good states to make investments above the level
that maximises the shareprices. This is not possible if the firm uses linear hedging, as in Froot, Scharfstein and Stein's fr amework.
Second, the money has to be put somewhere while the firm waits less favourable times. Putting the revenues aside in the money-,
                                                                                                                                  44
stock-, or bondmarket may be something that the shareholders will dislike. It can actually turn the firm into a take -over target. This
means that of the two choices the first is preferred.

If one reverses one of the assumptions and views a raise in the risky asset as a negative thing, like an increase in the oil-prices for an
airline company, things change dramatically. A raise in the oil-price is still as likely as a fall, the raise in the oil-price will however be
larger over time in absolute terms than a fall. The change in the risky asset will be large and with it comes a dramatic decr ease in the
cashflow from assets in place due to the high fuel-costs. The need for external capital will therefor be very large. Linear hedging will rob
the firm for some gains in the good states, but the firm will avoid large capital acquisitions and the huge dead -weight costs associated
with it. Linear hedging will indeed be the better choice in this situation.

This means that while linear hedging, perhaps combined with out of the money options , will be the optimal hedging strategy for some
firms in some of Froot, Scharfstein and Stein's models, options alone will be optimal for firms with the different exposure and perhaps
longer time horizons.




39
     This model is developed by the author of this thesis.
40
     As opposed to an airline company, where the raise in the oil price will be a disadvantage for the firm.
41
     This is the same argument that Black and Scholes used in their optio n pricing model.
42
     Caps and floors would probably be even better and even swaps can be used in this framework.
43
     See section 5.4.1, «The Manager Acts in Self-Interest.», page 31.
44
     See section 7.1.2, «Does the Firm Become a Take-over Target?», page 45.
                                                             PART II.
It is clear from the previous section that Froot, Scharfstein and Stein's framework is based on several assumptions. The following three
sections deal with the more fundamental assumptions; the costs of external capital, the costs of using Froot, Scharfstein and Stein's
framework and whether the framework will be accepted by stakeholders.
The subsequent two sections cover the more operational assumptions and deals with what actually makes a firm invest and different
derivatives/hedging strategies.


 5.            ARE THE COSTS OF EXTERNAL CAPITAL NECESSARILY
                                 HIGH?
Froot, Scharfstein and Stein's paper took the dead-weight costs associated with external financing as a given. They came up with three
categories of possible explanations for why external capital is expensive. These explanations did however only cover a couple of lines
in their paper and was by no means an attempt to give a full coverage of the topic. Froot, Scharfstein and Stein's explanations were:
 Bankruptcy costs and costs of financial distress . These can be both direct costs, like legal fees and indirect costs, like the cost of
    underinvestment in a firm in financial distress.
 Costs associated with asymmetric information between the manager and the investors.
 Costs associated with private benefits that the manager has at the expense of the shareholders. This is an example of an agency
    cost.

In order to understand the following arguments, an understanding of the simpler Modigliani and Miller (M&M) (1958 and 1963) models
           45
is needed. If the world was as simple as described by M&M there would not be any need for Froot, Scharfstein and Stein's
framework. The dead-weight costs associated with external financing would be zero. This result is due to the lack of market
imperfections in M&M's models. Different tax regimes will not change the conclusions and taxes are there for of little interest here. To
understand the cost of external capital, one needs to relax M&M's assumptions.

This section starts with a brief résumé of resent capital structure research, then the q uestion why one need outside financing at all is
answered. The section then goes on to explain what happens to the cost of capital when one relaxes M&M's assumptions, first the
assumption of no bankruptcy risk then capital structure problems when there are agency problems and asymmetric information .
There are however things that the firm can do to minimise the effects of these imperfections. In addition to describing the market
imperfections, this section also shows how the firm can reduce the effects of these market imperfections and thereby the dead weight
costs external financing. The theories are supplemented with empirical findings. If the dead-weight costs of external financing are
reduced the firm approaches a situation where it can rely on external capital and not on Froot, Scharfstein and Stein's framework, just as
in M&M's models.

5.1           Research on Capital Structure.
Financial theory used to be based on statistics, a trend that investment theory benefited largely from. The last two decades has
however seen a shift toward explaining corporate and investor policies from an agency problem/asymmetric information point of view.
The following sections are to a very large degree based on this framework. Recently however, there has been some work in financial
theory where the starting point has been industrial organisation and strategy. These theories will not be discussed here for several
reasons. First, Froot, Scharfstein and Stein based their framework on the agency problem/asymmetric informatio n literature. Second,
several of the industrial organisation papers take the firm's financing opportunities into account. Since this thesis does no t cover that
part of Froot, Scharfstein and Stein's framework, literature dealing with capital structure in a n industrial organisation framework has
been omitted.

In their survey of literature on capital structure Harris and Raviv (1991) found that although the financial literature has recently seen a
sizeable amount of theories, there is still much empirical work left to be done until we can evaluate the theories and start to develop a
good understanding of the capital structure choices that firms make. The empirical work that has been done does however large ly
support the theories. The fact that many theories lack empirical backing must be kept in mind in the following sections.

Myers (1984) said that while there may well be an optimal leverage ratio, firms usually do not issue equity and debt simultan eously.
This means that the firm will not usually be at its absolute optimal leverage, but it will be near it on average. Myers called this the static
trade-off hypothesis. The optimal capital structure is determined by the trade -off between the benefits and costs of debt. Brealey and
Myers (1991) said that firms within the same industry usually have about the same leverage. Myers (1984) did however say that there is
a wide range of leverage ratios. He explains this by the costs of adjusting the capital structure. Since an adjustment is costly, firms will
have different leverages.

Levy and Sarnat (1990) said that the optimal capital structure is not constant. If the interests rate raises the firm should increase the
proportion of equity since the debt becomes more expensive. Volatile interests rates will make it difficult for the firm to keep the capital
structure optimal because of the costs associated with acquiring capital.



45
     For an abstract of M&M see section 12.2.1, «Modigliani and Miller.», page 71.
The ratio of equity should also change with the age of the firm. Ward (1994) said that firms usually have to be financed with debt in the
beginning, because shareholders only have a claim on the residual. It is a large change that the residual is slim. The firm's business risk
tends to decrease as the firm matures. If the firm continues to be financed primarily by debt, both the financial and the business risk is
low. This low-risk firm will be undervalued and it will be a take-over target. The take-overs will decrease the leverage of the firm and
earn a lot.

5.2           Why Use Outside Financing at All?
                                                                                       46
The reason why we see outside financing, despite the costs that the owner-manager must carry, is according to Jensen and Meckling
(1976) the manager's desire to diversify. By selling a part of the future cashflow to outside investors, the owner-manager can use the
proceeds to buy other investment assets and thereby reduce the diversifiable risks. Another reason is that the extra money en ables the
owner-manager to take on projects that he would otherwise not be able to start. The sum of the lost future cashflow from the stock sale
and the costs from the agency problems are less then the gains from the new projects that the owner-manager can take on after the
capital acquisition.

5.3           Debt Financing.
The two main differences between debt and equity are the fact that equity holders are the owners of the firm and thereby exer cise
control and that they have a residual claim on the financial returns. Only th e latter is of interest here. There are three main reasons why
a firm should use debt in its capital structure.
 According to M&M's Proposition II a firm should use 100% debt in its capital structure if the debt payments are ta x deductible.
     Fama and Miller (1972) said that the firm value increases by the value of the tax subsidy.
 Myers (1984) said that because of the collateral debt financing would be cheaper.
 If there are agency problems and/or asymmetric information between the manager, the board and the owners debt financing can
     solve several of these problems, thus making outside equity cheaper.
The first point is discussed here, while the other two are discussed in the section about equity financing .

Myers (1977) said however that there are many reasons why we do not observe firms financed entirely by debt. For one thing St iglitz
(1972) and Miller (1977) found that by including personal taxes in M&M 's model these taxes might reduce or eliminate the effect of the
tax advantage. Other explanations are, according to Myers:
 Market imperfections.
 Managers may act in self-interest and protect their jobs and wealth.
 Financial decisions can be used as signals.
 The costs of bankruptcy and financial distress.
 And finally Myers' own argument that debtholders fear that the manager will act in the shareholders interest. This fear increases
    with leverage and would eventually make debt extremely expensive as the leverage approaches 100%.

These things will make debt more expensive than it would in the perfect world of M&M.

5.3.1         The Use of Debt.
Gowami, Noe and Rebello (1995) found that the reliance on debt has doubled over the last half-century and that long-term debt / long-
term capital has tripled. Smith (1986) said that of public securities underwritten between 1980 and 1984 63% were straight debt, 24%
were common stocks, 6% were convertibles, 5% were preferred stocks and the remaining 2% were convertible preferred stocks.

5.3.2         Why Risky Debt Matters.
The reason why risky debt matters is that it affects the weighted average cost of capital (WACC), which determines the value of the
firm. With risky debt the WACC becomes.

                                     WACC = (1-t) * k b(B/(B+S)) * B/(B+S) + k s * B/(B+S), where

WACC = The weighted average cost of capital.
T       = Taxes.
kb      = The cost of debt (bonds).
ks      = The cost of equity (stocks).
B/(B+S)          = The leverage.

As one can see, the cost of debt, kb, is a function of the leverage.

                                                             k b = k b(B/(B+S))

                                             47
This would not be the case M&M's world.           According to M&M only the cost of equity is a function of the leverage, because of the
increased risk.



46
     The term «owner-manager» is used by Jensen and Meckling (1976) in their paper and is therefor used here too.
47
     For an outline of the WACC in the simpler M&M world see section 12.2.1.2, «Corporate Taxes.», page 71.
                   Cost of capital
                                                Cost of equity

                                                WACC
                                                 Cost of risky debt
                                                        B/S (leverage)
                                      Optimal capital structure
Figure 13 Optimal capital structure with risky debt.
It is clear that bankruptcy costs and the cost of financial distress causes the cost of debt to increase. If a bankruptcy is costly the cost
of debt will increase by more than the cost of equity will decrease, causing the sum of the cost of the external capital to increase. This
would support the assumptions that Froot, Scharfstein and Stein made. This must of cause influence the value of the firm too.

                                                    VL I = VU +  BL - PV(P), where
     I
VL       = The value of the levered when it is insured against bankruptcy.
VU       = The value of the unlevered firm.
BL       = The value of the debt (bonds).
        = the aggregate effect of the tax, depending on whether it is a world without taxes, with only                  corporate taxes,
or with taxes on all levels.
PV(P) = The present value of the direct bankruptcy costs.

          The value of the firm              Levered firm
                                             without bankr.
                                             costs
               Unlevered firm                                              Levered firm with
                                                                           bankruptcy costs

                                                                           B/S (leverage)
Figure 14 The impact of bankruptcy costs on the firm value.


5.3.3        The Costs of Bankruptcy and Financial Distress.

5.3.3.1        Bankruptcy.
M&M argued that if debt is advantageous the firm should exploit this by employing 100% debt. By relaxing M&M 's assumption of no
bankruptcy risk, this conclusion changes. The shareholders hold a residual claim on the value of the firm. This makes their claim to the
firm a real option. By using option-theory one can show that this means that the claim that the shareholders hold on the firm is equal to
holding an asset and a put option on the same asset. If the value of the firm is less then the fixed debt payments to the debtholders, the
                                                    48
shareholders will exercise their option and default, and the debtholders will take over the firm. Brealey and Myers (1991) said that this
is not costly in it selves since the shareholders only had a residual claim on the firm. When this claim is worthless they simply give
away any further claims on the firm. Coast's theorem (1960) implies that if the debtholders could costlessly take over the firm, a
bankruptcy would not be a negative thing in itself, only a sad consequence of poor management.
Bernardo and Talley (1996) said that there are however several reasons why this is not costless in reality.
 If there are many debtholders it is costly to co-operate.
 The investment decision may be done before the bargaining and may be irreversible.
 Tendering of debt and debt restrictions happen simultaneous.

Levy and Sarnat (1990) found that between 30 - 70% of the value of the assets are used during a liquidation. This shows that there
must be a direct cost involved in a bankruptcy. This means that even if the shareholders hand over the firm to the debt holders exactly
when the value of their shares are zero, the value of the firm will still not equals the value of the debt. This is because the debtholders
would have to take from the value of the firm to finance the bankruptcy.
Weiss and Wruck (1998) studied the bankruptcy in Eastern Airlines. They found that out of the firm's value of 4 billion d ollars at the
time when the bankruptcy process started about half was lost during the process. The main reason for this was that the judge decided
to let the operations continue even though it turned out that the firm did not have the necessary economic st rength.
Because of these things the value of the debt after a bankruptcy will be uncertain. Copeland and Weston (1992) said that risky debt
should be priced according to the CAPM. The higher the systematic risk the larger should the risk premium for the debt be.

The interest payments are fixed costs. These payments increase with leverage, which implies that the probability of bankruptc y
increases with leverage. To compensate for the possible cash outflows associated with a bankruptcy the debtholders will charge a
premium in addition to the interests rate for the debt. This increases the cost of debt, making internal capital relatively cheaper


48
  Technical insolvency means that the firm is unable to meet its current obligations. Bankruptcy means that the firms liabilities exceeds
 a fair valuation of the firms total assets. This means that a liquidation of the firms assets will not be enough to pay the liabilities.
compared to debt. Since this premium will make the interest payments larger, the probability of a bankruptcy will increase. Because of
this the debtholders will have to increase the premium even more. This is a vicious circle.

Since a bankruptcy is just a change in ownership, the firm will often continue its operations. Brealey and Myers (1991) said that the
indirect bankruptcy costs are the difficulty of running a company when it is g oing through a bankruptcy. This includes the cost of
underinvestments. Since the firm has little money to invest, because of the debt payments, it may have to reject profitable projects just
because there are no money to finance the projects. The creditors may not allow the firm to sell off inventory to reinvest because they
do not think that the new investments will do better than the old ones. The only result will be reduced collateral, so the de btholders
reject the sale.

Rajan and Zingales (1995) found that across the G7 countries leverage increase with firm size. This is in despite the fact that tax rates
and bankruptcy codes differ in those countries. Rajan and Zingales said that one possible reason for this is that larger firms are better
diversified and thus have less bankruptcy risk. This means that large firms should be able to obtain lower interests rates fo r their debt.
Large firms would therefor be less likely to use Froot, Scharfstein and Stein's framework if one only looks at the bankruptcy costs.

5.3.3.2          Financial Distress.
A bankruptcy is of cause an extreme situation. One will however arrive at much of the same conclusions if the firm only runs into
                                                                                49
financial difficulties. Brealey and Myers (1991) used Myers' option approach when they said that when the firm is in financial distress
the shareholders are in an everything-to-win-nothing-to-lose situation. It is in their interest to take on projects with higher risks. This is
because the owners have only a residual claim on the firm. If the project goes well the shareholders will take the entire gain. However, if
the project fails the shareholders will be exactly where they were without the risky project. They are holding virtually worthless stocks.
The risky project may however reduce the firm's resources, making the firm and thus the debtholder worse off. To finance furt her
operations shareholders may for example cut expenses on R&D or maintenance. This will reduce the value of the collateral of the debt,
which hurts the debtholders. Because of this risk, the debtholders will demand a premium for holding the firm's debt. Harris and Raviv
(1991) called this the asset substitution effect.
Mella-Barral and Perraudin (1997) said that in many cases, the debtholders are persuaded by the equity holders to accept concession
prior to a formal bankruptcy. This take-it-or-leave-it offer will usually give the debtholders less then they really should have had. This
may however eliminate direct bankruptcy costs and the risk of underinvestments . The result is that the shareholders gain from such an
offer.

Opler and Titman (1994) said that it is difficult to quantify the net cost of a financial distress . It is however clear that financially
distressed firms loose market shares. If this is driven by costumes or competitors it is costly an d it will reduce the firm's value. However
if poor management causes the financial distress, the manager will eventually be replaced and the firm will be reorganised. O pler and
Titman said that such a reorganisation will result in a better firm and the firm value will increase. According to Opler and Titman capital
structure matters. Highly leveraged firms lose market share and experience lower operating profits. The relation tends to be stronger for
firms with significant R&D expenditures and for those in more concentrated industries. This is contrary to Ross' (1977) theory that only
high quality firms dare to have a high debt ratio.

5.3.4         Reducing the Costs of Bankruptcy and Financial Distress.
In order to avoid asset stripping, reduction of the collateral during a bankruptcy process, the debt contracts should have a clause that
prohibits further operations during a bankruptcy process unless it is very clear that further operations will benefit the debtholders.

In order to prevent asset substitution, wealth transferred from the debtholders to the shareholders, the debt contract should also have
a clause that prevents cuts in expenses on R&D or maintenance d uring financial distress. This will prevent a reduction of the value of
the collateral.

Brealey and Myers (1991) said that one possible solution for a firm that is in financial distress is to issue new equity so that it would be
able to take on new projects that will help the firm out of the distressed situation. However, the reduced chance of a bankru ptcy
benefits the debtholders, making their debt more valu able. The shareholders loose what the debtholders gain so that they do not get
back what they invested. Because of this the shareholders will not be willing to buy the newly issued equity unless it is und erpriced.
This makes the new external capital more expensive.

Smith and Stulz (1985) claimed that risk management should be used to reduce the likelihood of a financial distress and the costs
associated with it. This will have several effects. The direct effect is that the interest payments will be reduced, thereby r educing the
costs of external financing. The reduction of the likelihood of financial distress also increases the firm's debt capac ity. More debt
payments will reduce the over-investment problem which also reduces the cost of debt. There are also tax benefits associates with
increased debt. If one look at this argument twice one will see that by implementing the «old fashion» type of hedging, where all
variations in the cashflow is removed, one can make the debt less expensive. This will reduce the need for Froot, Scharfstein and
Stein's framework.




49
     See section 5.3.6, «Agency Costs of Debt.», page 27.
5.3.5          Other Disadvantages of Debt.
Lang, Ofek and Stulz (1996) found another reason why firms should avoid issuing debt and rather rely on cashflow from assets in
place. They found that a dollar less of operation cashflow reduces investments by less that a dollar less of cashflow brought about by
an increase in debt services. So, leverage matters more than operating cashflow when it comes to investments. The negative
                                                     50
relationship holds only for firms with low Tobin's q. This supports Froot, Scharfstein and Stein's framework.

5.3.6          Agency Costs of Debt.
The agency problem increases the cost of debt in several ways. Most of the theories concerning this problem deal with the fact that
the shareholders have an incentive to cheat the debtholders. This is the wealth expropriation problem. Jensen and Meckling (1976)
gave an explanation for why 100% debt financing is not possible, contrary to what M&M said. Jensen and Meckling said that because
of the agency problem it would be impossible to use far more debt than equity. If a project were to be financed by far more d ebt that
equity the debtholders would take a huge loss if the project fails, while they will only ge t a modest interest payment if the project is
successful. The shareholders on the other hand would lose a small amount of money in the case of a failure, while they would take
most of the gain if the project was a success. Debtholders will not accept this risk and it will be impossible to finance a project this way.
Jensen and Meckling said that the firm's leverage should be a trade-off between the agency cost of debt and the tax benefits. There are
several things that will determine this trade-off.
 Debt contracts should include features that prevent asset substitution
 Industries where asset substitution is less likely will have more debt.
 Firms with slow or negative growth will have large cash inflows and should therefor have more debt.

Another theory concerning the agency problems of debt is Myers' (1977) theory that also described why firms do not use 100% d ebt.
This theory is based on much of the same line of reasoning as the one that Jensen and Meckling used. Myers said that the
shareholders' claim is a real option. In the case where the debt matures after the option expires the debt will be worth MIN [Firm value,
Debt payments] and the equity will be worth MAX [0, Firm value - Debt payments]. The shareholders will invest only if the firm value
exceeds the sum of the investment outlay and the debt payments. Myers said that the creditors will get nothing if the option is not
exercised and if it is exercised the firm's value will exceed the debt payments, so MIN [Firm value, Debt payments] = Debt payments.
Viewing the capital structure this way makes it clear that the capital structure influences the firm's growth options . Because of the debt
payments some projects that would otherwise be NPV > 0 projects will not be undertaken because the net present value of the
cashflows - investments - debt payments turn the projects into NPV < 0 projects. Going along with such projects will only benefit the
debtholders, so the shareholders have no incentive to take on those projects. In Myers' framework the most important question in
determining the optimal capital structure is the ratio between the part of the firm's value that can be viewed as a c all option and the part
that is not dependent of the managers future decisions, such as assets in place. Potential debtholders will foresee the risk that the
manager will not undertake all NPV > 0 projects and require a higher interests rate. Firms with growth options should be financed with
equity, while firms where most of the value lies in assets with good collateral should use debt. If Myers' approach is put in to Froot,
Scharfstein and Stein's framework one can see that debt financing not only makes the project less profitable, but the NPV may end up
being negative and unless the project is financed with internal capital it may be rejected.

Martin (1996) used a similar way of reasoning. He said that firms with risky debt outstanding might fail to exercise a real o ption, an
investment opportunity, because doing so may transfer wealth from the shareh olders to the debtholders. The firm has a certain value
which is the sum of the value of the debt and the value of the equity. Exercising the option will bring money into the firm, which the
firm can use to pay its debtors. This will increase the value of the debt and the entire proceed from the option will benefit the
debtholders.

Brealey and Myers (1991) came up with another agency problem. They said that the firm could first issue a little debt. The risk for this
debt is small and so will the risk premium charged by the debtholders be. Later the firm can issue more debt, raising the risk of the firm,
including the risk of the existing debt. The average price of the debt will be too small compared to the average risk. The sh areholders
would thus have financed their firm with cheap, risky debt.

Given the fact that there are bankruptcy costs, financial distress costs and agency costs associated with debt financing one should
assume that debt is not very popular among decision -makers. After all Weinstein (1978) found that debt, just as equity, is underpriced
when it is issued. New debt yields abnormal returns during the first month. This support Froot, Scharfstein and Stein's framework. If
the things mentioned above really do make debt financing more costly than internal capital the firm should rely on Froot, Scharfstein
and Stein's framework instead of issuing debt.

5.3.7          Reducing the Cost of Debt.
Given the extensive use of debt there must be ways to reduce the agency problems mentioned above. Financial theory has severa l
proposals to how a firm should avoid the problems of expensive debt, although several of them carry wit h them new agency problems.
This is of cause important here, since Froot, Scharfstein and Stein's assumption that external capital is more expensive than internal
capital is questioned here. If this is only true in some cases, or if firms can easily take actions that reduces the cost of debt , this will
reduce the usefulness and importance of Froot, Scharfstein and Stein's framework.



50
     Tobin's q is the ratio of market value / replacement value.
5.3.7.1          Low Risk Projects.
As mentioned above Jensen and Meckling (1976) explained why one could not finance a project with much more debt than equity.
They did however come up with a possible solution to the problem. The manager can tell the potential debtholders that the p roject is a
low-risk project. A low-risk project implies that the interests rate that the firm has to pay for the loan should be low.
There is however an obvious problem with this approach. If the debtholders accept these terms the manager still has an incentive to
use the «cheap» money to finance a high-risk project. Such a project should have had higher interest payments than the interests rate
that the firm got from the debtholders. The shareholders can actually benefit by increasing the risk so much that the total value of the
firm decreases. Debtholders will foresee this threat and always require a high interests rate. Because of the lack of credibility this
approach will most likely not work.

Another agency problem may actually help solve the above problem, according to Hirshleifer and Thakor (1989). They said that the
manager prefers low risk projects and will not get involved in asset substitution even if it benefits the shareholders. This is because the
manager has undiversified risk. The manager's utility maximisation will help the debtholders, not, as one might assume the
                                                                                                                51
shareholders. This is a very good example of the point that was made in the very beginning of this thesis. It is not credible that the
manager will maximise the shareholders wealth, although many financial theories take that view as a starting -point. Rather, all actors
will maximise their own utility. The problem with this approach is that the debtholders can not verify the managers risk aversion. They
have no guarantee that the manager will not take on high -risk projects.

5.3.7.2          Contracts.
One way to reduce the cost of debt is through the use of contracts. Jensen and Meckling (1976) said that the firm should use contracts
to avoid the asset substitution and that industries where asset substitution is not a problem firms should use more debt.
Myers (1977) said that this is not as easy as it might sound. It would be optimal for the debtholders that all NPV > 0 projects were
accepted. Such a contract would require that all the shareholders would be personally responsible for the firm's investments. This
would raise a ferried problem among the shareholders.
Myers said that one possible solution is to re-negotiate the debt contract whenever the NPV is positive but less than the payment
promised to the creditors. This would leave both investors and creditors better off than if the project was not undertaken.
Renegotiating the debt contract will however be costly. These costs are both the direct costs of the re -negotiations and the costs for
the debtholders to verify the fact that the cashflow from the proposed project is in fact too low to handle the debt payments .
Shareholders have an incentive to lie and gain from the fact that they do not have to pay all the debt payment s. The creditors can not
trust the manager, who may act in the interest of the shareholders and who «always» will claim this. Re -negotiations will be easier if the
debt is short term but Myers said that rolling over short-term debt would not solve the problem, just potentially lower the costs.

Bernardo and Talley (1996) claimed that the debtholders can be offered to exchange their old debt for new debt with lower fac e value
and higher priority, equity or cash if the firm is in financial distress . There are however agency conflicts attached to this approach too.
The shareholders may select inefficient projects in order to enhance their bargaining power.

In order to avoid asset stripping, reduction of the collateral during a bankruptcy process, the debt contracts should have a clause that
prohibits further operations during a bankruptcy process unless it is very clear that further operations will benefit the deb tholders.

5.3.7.3          Dividend Restrictions.
One form of contracts is of special interest in this context; dividend restrictions . If the manager practices wealth expropriation the
wealth must be transferred to the shareholders in one way or another. Myers (1977) said that restrictions on dividends could protect
against sub-optimal investments. The problem is that there are so many ways that wealth can be transferred to the shareholders that
restricting this is difficult.
If the firm is in financial distress Brealey and Myers (1991) said, the shareholders want to increase the dividends. The increased
dividends will reduce the value of the firm, but this reduction is shared with the debtholders, while the benefit of increased dividends
go directly to the shareholders. Dividend restrictions will prevent this.

Long, Malitz and Sefcik (1994) studied wealth expropriation and said that if the manager does employ a wealth expropriation strategy,
debt issuance would be followed by dividend increases. They found that firms issuing straight debt are more likely to increas e
dividends and less likely to decrease dividends than firms issuing convertible debt. They also increase dividends slightly more often
than the market on average. This shows that debtholders are right in charging a high price for the debt. Despite these findin gs, Long,
Malitz and Sefcik said that they found no evidence of wealth expropriation. They come up with two different explanations for why firms
do not want to engage in asset expropriation. First, the covenants may do exactly what they are supposed to do, that is restr ict such
practice. Second, the reputation may be more important than a one-shoot gain.
Gowami, Noe and Rebello (1995) found that 55% of all firms have dividend restrictions .

Although dividend restrictions are a way to reduce the likelihood of asset expropriation and thereby make the firm less likely to use
Froot, Scharfstein and Stein's framework, dividend restrictions are important in Froot, Scharfstein and Stein's framework t oo. Myers and
Majluf (1984) said that in order to save as much as possible of the cheapest form of capital, internal capital, the firm should restrict the
dividend payments.


51
     See section 2.1, «The Goal of the Firm.», page 9.
5.3.7.4           The Manager.
Yet another form of contracts can be used to reduce the probability of wealth expropriation and thus reduce the firm's interest
payments. John and John (1993) said that the compensation plan should be tied to stability and the size of the cashflow. This will
ensure the debtholders that the manager will not accept very risky projects and this again will reduce the cost of debt . Since the
shareholders hold the res idual claim this will benefit them too. John and John said that the sensitivity of the manager's compensation
to the shareholders wealth is of importance. If the agency costs of debt are high the [managers compensation / shareholders w ealth]
ratio should be low. Levered firms should have a lower sensitivity than all-equity firms should. Jensen and Murphy (1990) found that
the [managers compensation / shareholders wealth] ratio is in fact low. It is however larger for smaller firms than for large firms.

The manager can be involved in other ways than just through the compensation plan . Leland and Pyle (1977) argued that the cost of
debt is related to the managers portion of the shares and that the debt will be cheaper the larger the portion. However, Mikkelson,
                                                                                                                               52
Partch and Shah (1997) found that the manager's ownership stake decreases from, on average, 67% immediately prior to an IPO to
44% immediately after. Five years after an IPO it is down to 23% and after 10 years it has fallen to 18%. Viewed in isolation , this
argument means that older firms should be more inclined to use Froot, Scharfstein and Stein's framework.

5.3.7.5           Monitoring.
The creditors can use a mediator to supervise that the firm does not follow an asset substitution strategy. Myers (1977) said that the
problem is to know when the mediator should be called in.

5.3.7.6           Reputation.
John and Nachman (1985) made a model where the firm issued debt more than once. In such a setting reputation plays an important
role. Since outsiders know less than insiders do and monitoring is costly, the cost of debt will be higher than in a full-information
Pareto optimal situation. Because of this the firm will invest less. Firms that know that they will need to acquire additiona l debt in the
future should not underinvest. If they do the price of the debt will increase. The interests rate will be a function of the firm's repayment
record.
Diamond (1989) argued that the firm should build a reputation for not expropriating debt. The best reputation for not choosing too
risky projects is that the firm has not gone bankrupt. Old firms signal that they do not start risky projects just by being o ld. If this
theory holds, old firms should have a lower cost of debt and thus have less need for Froot, Scharfstein and Stein's framework.
Myers (1977) said that although shareholders gain on wealth expropriation in the short term, it is a costly business in the long-term and
the shareholders are the ones who must pay the cost at the end of the day. The best way to avoid these long run costs is to avoid
cheating the debtholders in the short run.

5.3.7.7           Growth Options VS Collateral.
Jensen and Meckling (1976) said that wealth expropriation should not be a problem for firms in slow growing industries and that such
firms should use debt. The reason is that there are few growth options and less uncertainty about the manager's actions. This means
that a great many firms have access to fairly cheap debt and do not have to even consider Froot, Scharfstein and Stein's framework.

Titman and Wessels (1988) found that firms with unique or specialised products have low debt ratios. The reason for this is that the
second hand market for those assets will be very illiquid and the value of the collateral will thus be low. The firm will not be able to
issue secured debt and may just as well issue equity as unsecured debt. Titman and Wessels did however not find any evidence that
debt ratios are related to expected growth.
Martin (1996) also said that despite theories that said that firms with growth options should finance themselves with equity, there is
little empirical evidence to support these theories.
Barclay and Smith (1995) did however find evidence of this. They found that firms with growth opportunities have less long-term debt
and that large firms and regulated firms have more long-term debt.

5.3.7.8           The Debt Structure.
Barclay and Smith (1995) said that the underinvestment problem is more serious the more growth opportunities the firm has. They came
up with four ways to reduce the agency problem.
 Less debt.
 Shorter debt.
 Restrictive covenants.
 Shortening the maturity of debt.
The first is the most important. Barclay and Smith also said that if the debt matures before the real option is exercised the problem is
eliminated. This is the same result as Myers (1977) got.
Gowami, Noe and Rebello (1995) found that firms solve the asymmetric information problem differently depending on whether the
asymmetric information is near or far away and that Barclay and Smith 's conclusion of shorter debt only holds in some states:
 If there is asymmetric information about long-term cashflow the firm will use coupon-bearing long-term debt with dividend
    covenants.
 If on the other hand the asymmetrical information is about near-term cashflow the firm will prefer coupon-bearing long-term debt
    without dividend covenants.


52
     IPO, initial public offering. The firms goes public to raise equity for the first time.
 Uniformly distributed asymmetrical information result in short-term debt financing.
It seems rather obvious that the structure of the debt itself is important in determining the cost of the debt.

5.3.7.9          Bank Debt.
Johnson (1997) found that firms do not choose the debt type randomly. He found that firms use more public debt if borrowers'
incentive to take actions harmful to lenders can be decreased. Bankloans are concentrated around smaller firms, due to the monitoring
and information asymmetry costs. In general, bankdebt is negatively related to the market -to-book ratio and positively related to
leverage. Private non-bank debt has the opposite relations. Both are negatively related to age. Three quarters of the firms in the sample
borrow simultaneously from both sources.
Carey, Post and Shape (1998) found that banks and finance companies are equally likely to finance information -problematic firms.
Finance companies tend to serve observable riskier borrowers, particularly more leveraged borrowers. They also found, contrary to
what others have found, that banks are probably not better suited for monitoring borrowers.

5.3.7.10          Convertible Debt.
Kang and Lee (1996) found that issuing convertible debt gives excess returns of 1.11%. This shows that it is underpriced. The
underpricing is between that of IPOs and seasoned issuings, but above the underpricing of straight debt. The underpricing decreas es
with firm age and increases with equity . Convertible debt is used to reduce the asymmetric information problems. Of great importance
in evaluating Froot, Scharfstein and Stein's framework is Kang and Lees comments about underpricing of other securities. Without
mentioning references they said that several studies report that there is no, or only slightly underpricing of seasoned equit y and
unseasoned debt.

5.3.7.11          Repurchasing Stocks.
Smith (1986) surveyed empirical work on stock price reactions after firms announce that they will issue debt. There is an abnormal
negative return for both straight and convertible debt. Smith said that different empirical studies came up with different estimated for
underpricing of debt. The underpricing ranged from 0.05% to 1.6%. The results were however very different if the debt offering was
combined with other financial transactions. The abnormal return after the announcement of debt issuing was as follows.
 21.9% when it is combined with common stock repurchase.
 14.0% when it is combined with common stock exchange offer.
 2.2% when it is combined with preferred stock exchange offer.
Smiths (1986) findings confirmed that a common stock repurchase is a signal of increased future cashflows. The two-day abnormal
return after a firm announces that it will repurchase common stock is 1.6% for targeted small holdings, 3.6% for open market
repurchases and 16.2% for intra-firm tender offers. All these are higher than the 0.9% abnormal return associated with dividend
increases, which is usually seen as a way managers signal good future cashflows. This is especially important in Froot, Scharfstein and
Stein's framework because paying dividends will actually reduce the amount of the form of capital that Froot, Scharfstein and Stein said
that the firm should use for investments. The fact that there are ways to signal the future cashflows to the market is important and the
fact that there are alternatives to increased dividends is even better.
Ikenberry, Lakonishok and Vermaelen (1994) documented that firms that repurchase stocks over-perform.
Nelson (1994) showed that since 1926, NYSE firms that reduce the outstanding number of shares outperform those who increase t he
number of shares. Issuing firms have typically had a good period and exploit this. This was Myers' argument too.
The findings mentioned above may however not give the full picture. Nohel and Tarhan (1998) found that the improved operating
performance following a stock repurchase is limited to low growth firms. The improved performances are generated by a more efficient
utilisation of assets and assets sales, not improved growth opportunities . The goal for these firms is not to change the capital
                                      53
structure, but to shrink the assets.

5.3.8          Summary.
It has been shown that although there are agency problems concerning debt financing, there are also ways to deal with that problem.
This can be summarised as follows.




Facts about debt financing.
Leverage increases with firm size.
A bankruptcy is costly for the debtholders both directly and indirectly.
Financial distress is worse for firms with: high leverage, much R&D or firms in concentrated industries.
Firms with low Tobin's q use more internal capital and less debt.
However, firms in slow-growing industries should use more debt.
Firms with low market-to-book value or firms who are small or young use bank debt instead of private debt.
A financial distress may put the equity holders in an everything-to-win-nothing-to-loose situation.
Firms with unique or specialised products use less debt.
Agency problems prevent M&M's conclusion of 100% debt financing.
Financially distressed firms loose market shares.

53
     This shows the limitations of cross sectional studies.
The managers portion of the shares decrease after an IPO.
Firms with growth opportunities use less long-term debt. The result is however not clear.
In order to reduce the cost of debt the firm should emphasise the following things:
The firm should have debt contracts that regulate later issuance of debt.
It should be cheap to renegotiate the debt contracts .
It should be easy to resign the manager in the case of a financial distress .
The manager should not be diversified, he should have most of his wealth invested in the company.
The manager should own a large portion of the shares.
The firm should use a mediator to monitor the manager.
The manager's compensation-plan should be tied to cashflow stability, not to shareholders' wealth.
The firm should avoid any asset substitution and build a reputation for not using such a strategy. The most important element in the
reputation building strategy is never to increase dividends after a debt issuance.
The firm should match the debt term and divided covenants to the time of the asymmetric informat ion.
The firm should issue convertible debt.
The firm should repurchase equity when it issues debt.
Table 6 Avoiding the agency problems associated with debt financing.

5.4         Equity Financing.
Contrary to M&M's argument it was shown above that debt financing may be costly due to the direct costs and the indirect costs such
as bankruptcy costs, costs of financial distress and agency problems. With this in mind, equity financing seems like a good alternative.
Noe and Rebello (1996) even said that if there was a combination of managerial opp ortunism and adverse selection problems, equity
issuance would be optimal.
There are however problems involved with equity financing too. There are two types of agency problems that make equity expensive.
 The manager may act in self-interest and does not maximise the shareholder's wealth.
 The manager may act in the interest of the existing shareholders and therefor cheat new shareholders .
It will however be shown that there are ways to solve these problems too, just like the problems associated with debt financing .

5.4.1       The Manager Acts in Self-Interest.
Two papers form the basis of this theory; Jensen and Meckling (1976) and Stulz (1990).
Jensen and Meckling (1976) said that the manager is compensated in two different ways by the firm. One is in the form of money
(salary, bonuses, stock options , etc), while the other form of compensation is non-pecuniary. This can be such things as having a
larger then necessary office, overuse of the company's jet, having a larger staff than necessary, etc. If the manager owns 100% of the
firm, he carries 100% of the costs of such overuse of non -pecuniary. The agency problem occurs when the manager owns less than
100% of the firm. If the market expects the overuse it will discount the value of the firm when it prices the shares and the owner-
manager carries the entire cost when the firm issues equity for the first time. However, in all later equity acquisitions the old
shareholder, both the owner-manager and outside shareholders, will carry the cost. The manager's insensit ive to overuse non-
pecuniary increases as his share of the firm's stocks falls.

Another form of non-pecuniary is the size of the firm. Jensen and Meckling argued that the manager wants the firm to be as large as
possible because the larger the firm, the higher status of being the manager of that firm. The manager has an insensitive to increase the
production to a level above the optimal level. In microeconomic terms this means that the margin al cost of the last good produced
exceeds the marginal revenue.

Stulz (1990) said that the moral hazard problem between managers and shareholders is that the manager is willing to accept NPV < 0
projects. The manager will do this even if paying the money out as dividends would be better for the shareholders. If the amo unt of
perquisites is an increasing function of investments, the manager will invest as much as possible. The cost of this is obvious, there is
however yet another agency cost involved. It is the fact that with the overinvestments there may not be enough money left to fund
                      54
really good projects.

Harris and Raviv (1990a) described yet another agency problem between the manager and the other stakeholders. Harris and Raviv said
that managers want to continue operations even if shareholders would be better off with a liquidation.

                                                                                           55
5.4.2       Reducing the Agency Costs of Managers Who Act in Self-Interest.
It was shown above that there are several ways to reduce the problems that increase the costs of debt. There are also several ways to
reduce the agency problems of equity.


54
    There are however overinvestment problems with Froot, Scharfstein and Stein’s framework too, see section 7.2, «The
  Shareholders.», page 46.
55
   This section covers both theories and empirical evidence concerning the agency problem of managers who act in their own inter est.
  Empirical evidence concerning the asymmetric information between new and existing shareholders are covered in section 5.4.4 «The
  Manager Acts in the Interest of the Existing Shareholders.», page 34.
5.4.2.1        Growth Options.
Jensen (1986) said that the abnormal stock return, i.e. the stock underpricing, associated with a stock issue will depend on the firm. The
reaction will be positive for high growth firms and negative for other firms. The reason is that for the high growth firms it is obvious
that the money will be spent wisely on good projects, while the money may be used for overinvestments in the other firms.
Martin (1996) used the same line of argument as Jensen (1986) did when he said that debt financing maximises the value in firms with
poor investment opportunities . This means that overinvestments are not a problem for firms with good investment opportunities. Large
investments are only a problem when there is a risk of overinvestments. Firms with good growth opportunities should invest a lot. This
means that as long as the manager invests in relevant projects, large investments are a positive, not a negative thing. If Ma rtin is right,
firms with good investment opportunities can finance their growth with outside equity without having to pay more than the fai r price
for the money.
Denis (1994) tested Jensen's hypothesis that the stock price reaction will be positive for high growth firms and negative for other firms
when they issue equity and found evidence that support Jensen. There is a positive relation between several ex ante measures of
                      56
growth opportunities and announcement period abnormal stock returns. The result was however not monotonic, because the results
seemed to be driven by a small subset of high growth firms. For other firms there is no relation between investment opportunities and
               57
stock returns. Those smaller firms were on average much smaller, they had been incorporated for fewer years and were less inclined
to issue debt. If one controlled for this, the investment opportunity explanation was insignificant. If Jensen's theory and D enis'
findings are correct, very small high-growth firms need not worry about expensive external capital and it may not be necessary for them
to implement Froot, Scharfstein and Stein's hedging strategy.

5.4.2.2        Debt.
When Myers (1977) explained the optimal debt level he said that the firm's growth options are important. Here one can see that growth
options are important for the equity too. If most of the firm's market value is in its growth options, Jensen and Martin showed that the
firm should be financed with equity. Myers said that if most of the firm's market value is in the assets, these assets should be used as
collateral and the firm should be financed with debt.

Jung, Kim and Stulz (1996) said that debt financing will force the firm to forgo some NPV > 0 projects and because of this, firms with
good investment opportunities will prefer equity financing. This is of cause Myers' (1977) argument. Jung, Kim and Stulz found,
contrary to the previous two studies, that firms that issue equity experience considerable asset growth from the year prior t o the issue
to one year after the issue. Firms with the best investment opportunities did not experience adverse stock returns when they issued
                                                            58
equity. Some firms issued equity when the Pecking order suggested that they should have issued debt. These firms experienced
higher asset growth than similar firms financed by debt.

Jensen (1986) and Stulz (1990) argued that increased debt will reduce the amount of internal capital and that this will reduce the
manager's over-consumption. Stulz said that the gain is a reduction of the overinvestments. The cost is however that some NPV > 0
projects will not be carried out. The reason for this is that the debt payments will be so large that there will not be enough money left to
finance any good projects. This underinvestment problem is what Myers (1977) described, only he mentioned it as a reason why debt
is expensive. If not all NPV > 0 projects will be undertaken, this will hurt the debtholders and they will require an increased interests
rate. This means that increased debt to reduce the agency problem of equity will increase the agency problem of the very same deb t.
The optimal leverage will, according to Stulz, depend on the investment opportunities and the probability distribution of the cashflow.
If the manager acts in his own self interest he will not want debt financing , since it reduces the investment amount and thus the
perquisites. If the manager acts in the shareholders' interest he will pay out as dividend all excess cashflow above the amou nt used for
the optimal investment. So dividend payments will poss ibly hurt the debtholders, because it can be part of an asset substitution
strategy. It can however benefit both the shareholders and the debtholders because it reduces the manager overinvestments.
One example that shows that debt does discipline managers is Hanka's (1998) study of how leverage influences the terms of
employment. He found that firms with more debt used part-time and seasonal employees, paid lower wages and funded pension plans
less generously. This way the firms s hifted the risk onto the employees.

Noe and Rebello said that in the absence of asymmetric information the shareholders' optimal policy is to rely entirely on debt. This will
reduce the manager's benefits. If the adverse selection costs are relatively small, the firm will be financed with debt and will pay some
dividends if it has favourable information. If the adverse selection costs are moderate, the optimal policy if the firm has f avourable
information is a mixture of correctly priced equity and debt. Last, if the adverse selection costs are severe the firm will under-price its
claims if it has favourable information. Debt financing is always optimal in the presence of pure managerial opportunism prob lems.

5.4.2.3        Convertible Bonds.
Straight debt carries with it the underinvestment problem. If the shareholders try to solve the overinvestment problem of selfish
managers by using debt, the firm will not be able to take on all NPV>0 projects. Mayers (1998) said that convertible bonds would solve

56
   These are: the book-to-market ratio and Tobins`q, which is the most commonly used. In addition are dividend yields, R&D-to-sales
  and return on equity. Ex post measures, such as subsequent growth and actual capital expenditures may not measure investment
  opportunities.
57
   This clearly illustrates the limitations of cross -sectional empirical work. One will only find what one is looking for, so the result
  may not be a complete picture of what is actually going on.
58
   See section 36, «The Pecking Order.», page 36.
this problem. He developed a two period model where the firm invests in the first period and may invest more in the second pe riod if
the first investment is a success. With straight debt it would be cheapest to issue debt only once, at time zero. However, there would
be an overinvestment problem. The convertible debt should initially be out of the money, but it will be in th e money if the NPV of the
first period's investments turned out to be profitable. If the firm's investment option is valuable the debt is converted int o equity and
the firm's leverage decreases. The reduced debt payments help the firm in the second period. When there is uncertainty about the
maturity of the investment option a call provision is needed.

Even though Mayers' suggestion to use convertible bonds may sound like a good idea, investors may not agree. Smith (1986)
surveyed empirical work on stock price reactions after firms announce that they will issue debt and he found that there was a n
abnormal negative return for both straight and convertible bonds.
Lee and Loughran (1998) arrived at the same conclusion. They documented poor stock and operating performance in the years
following the offering. This, they said, can not be explains by an IPO, a SEO, or the level of proceeds. The profit margins and the return
on assets are approximately halved during the four years following the offering. Lee and Loughran assumed that the stock reac tion
could be explained by the fact that investors believe that the convertible bonds are issued when the firm is overvalued. This is of cause
                                                                                          59
Myers' (1984) argument for why a stock issuance is also followed by poor stock returns.

5.4.2.4          Monitoring.
To solve the agency problem and thus sell s tocks at a higher price, the manager should impose monitoring on his own actions.
But just as it may be costly to use debt to reduce the agency problem of equity, monitoring will be costly too. Jensen and Meckling
(1976) assumed that the effect of monitoring with respect to the monitoring costs is concave. The owner -manager carries the
monitoring costs. Monitoring the manager increases the value of the firm because it decreases the likelihood of overuse of non -
pecuniary, but it decreases the value of the firm by the monitoringcosts.

5.4.2.5          The Board.
As mentioned earlier, Rao (1994) said that it is the board's job to supervise the manager on behalf of the shareholders. Many boards fail
however to do so. This means that it is up to the manager, who wants to overuse non -pecuniary, to impose monitoring on himself. The
manager who acts in self-interest may not be to eager to implement this. This means that monitoring carries with it both costs and
incentive problems, just as straight debt did.

Brickley, Coles and Jarrell (1997) said that there have been many complaints about the fact that the CEO often also is the chairman of
the board. This is not only a problem in the US. They said that even though little empirical work has been done on the subject, they
found that it would be beneficial for the firm to separate the two.

5.4.2.6          The Manager's Ownership.
Jensen and Meckling (1976) said that the manager is more likely to overuse the firm's resources if his ownership in the firm is low. This
way his costs of the overuse are also low. Morck, Shleifer and Vishny (1988) confirmed this when they found that low managerial
ownership results in low firm value measured by Tobin's q value.
Rosenstein and Wyatt (1997) found that there is a negative stockprice reaction when the manager owns less than 5% of the shares.
There is, on the other hand, a significantly positive reaction when the ownership is between 5 and 25%, while there is no rea ction when
the ownership is above 25%. Denis and Denis (1994) may have the answer to that seemingly puzzling finding. They said that while
there are many mechanisms that can reduce the managerial efficiency, these mechanisms might be less efficient if the manager holds a
controlling block of stocks. This means that Jensen and Meckling's (1976) argument mentioned above may not hold. The truth may
rather be something like this. When the owner-manager owns 100% of the shares he himself carries all the costs of the overuse. If the
manager sells a minority part of the firm, he is still in control and can not be threatened, the costs of an overuse of the f irms resources
are however partly carried by the investors. If the owner-manager sells a controlling block other shareholders can impose restrictions
on him. Even though the owner-manager carries even less of the costs of the overuse, the other shareholders can control him. This
may explain why firm value is not a linear function of the manager's ownership. Denis and Denis found that firms where the owner-
manager holds more than 50% of the shares do not rely on other organisational constraints as a substitute for the control. Th is may
however not matter, since they also found that firms where the owner-manager controls a majority block do not perform poorly.
Denis, Denis and Sarin (1997) found that ownership structure has an important influence on internal monitoring efforts.

This should be viewed together with an argument mentioned earlier. Leland and Pyle (1977) argued that the cost of debt is related to
the managers portion of the shares and that the debt will be cheaper the larger the portion. W hen the manager owns a large portion of
the firm, both debt and equity will be cheaper. This is of cause not surprising, since agency problems account for a large pa rt of the
costs of external capital.

Despite the theoretical reas ons for high manager ownership and the empirical support for these theories Mikkelson, Partch and Shah
                                   60
(1997) found, as mentioned earlier, that the manager's ownership decrease after firms go public.




59
     See section 5.4.4.2, «Timing.», page 35.
60
     See section 5.3.7.4, «The Manager.», page 29.
5.4.2.7        The Manager's Duties.
Jensen and Meckling suggested that the manager should impose restrictions on his own decisions. Such limitations are meant to force
the manager not to take actions that hurt the shareholders. Examples of restrictions a re things like dividend policies, future debt, or
maintenance of working capital. These restrictions do however carry with them costs, which Jensen and Meckling called bonding
costs. These are the costs that occur when the manager can not make decisions th at would be optimal for the firm because the he has
limited decision power. Bonding costs also include such things as having a public account audit the financial accounts to pro ve that
the owner-manager does not overuse resources or over-expand the firm.

Stein (1997) said that because of the over-investment problem, managers should be given credit constraints when dealing with external
capital. In his opinion, management is not suited for acquiring new funds at all. The managers' d uties should be limited to the use of
funds, not to raising them. The manager should allocate the funds to the projects with the highest returns, provided they are positive.
The job of acquire external capital and to expand the size of the firm should, according to Stein, be given to someone who will not
benefit from it.

John and John (1993) said that a compensation plan that minimises agency cost between the manager and the shareholders may
increase the agency costs of debt. This means that incentives like these can not be used. Although the compensation plan can be used
to solve the agency problem of debt, it can not be used to solve the agency problem of selfish managers.

5.4.2.8        Take-over Threat.
Stulz (1990) said that the take-over threat would motivate the manager to do the right things. If the manager makes less than optimal
decisions, this will reduce the shareprices. It will be cheap to acquire the firm and then replace the manager with one who is willing to
                                                                                                                               61
make decisions that are better for the shareholders. This means that the firm should not have to many anti-takeover measures.

5.4.3       Summary.
The manager has an incentive to act in his own interest. This increases the cost of equity. Below is a summary of the facts concerning
this and a list of things that the firm can do to prevent these actions and thus reduce the cost of new equity.
 Facts about the agency problems between the manager and the shareholders.
 Small growthfirms do need not worry about overinvestments.
 In order to reduce the cost of equity the firm should do the following things:
 Have at least some debt.
 The debt should preferably be convertible.
 Not have to many anti-takeover measures.
 Monitor the manager.
 The manager should not have any of his «friends» in the board .
 The manager should own a relatively high number of the shares.
 The manager should be given credit constraints.
Table 7 Avoid the agency problems of overinvestments.


5.4.4       The Manager Acts in the Interest of the Existing Shareholders.
The previous sections explained the overinvestment problem of equity. This section covers the problem that arises when the manager
goes to the capital market to acquire more equity to finance a project. The new shareholders do not trust the manager when he
describes the good investment opportunities. The reason why the manager needs more money may just as well be that the firm is in
financial difficulties. There is asymmetric information between the manager, who knows the state of the firm and potential new
shareholders. Because of this the new shareholders requires a premium that makes the new equity more expensive. This results in an
underinvestment problem.

5.4.4.1        Firm Quality.
Myers and Majluf (1984) said that if the manager acts in the interest of existing shareholders , he may not accept all NPV > 0 projects.
The reason will be that by selling shares to raise money for the new pro ject the firm gives away future cashflow both from the new
project and from ongoing operations. The cashflow that the existing shareholders must give up in order to get funds for the n ew
project may exceed the cashflow they will get from the new project. The result is a net loss for the existing shareholders. So if the firm,
despite of this, still issues new equity, this is a signal to the market that the ongoing operations are not very successful. The existing
shareholders are after all willing to give up a part of the future cashflow from those assets in place to finance new projects. The capital
market will therefor take a stock issue as a signal of bad news about the profitability on the firm. In Myers and Majluf's model there are
only two types of firms. Firms with current assets worth a lot and firms where the current assets are worth less. Investors can not
observe what type a firm is. If a firm with low-value current assets issue equity to finance a project, the new investors will want a large
share of future earnings form the firm. The low-value firm can accept this and issues equity. However, if a high value firm want to issue
equity the investors will anticipate that it is a low value firm and require a large part of the future earnings from this firm too. The

61
  Anti-takeover measures are actions taken by the firm before it becomes a take-over target to prevent anyone to even try to acquire
 the firm. Examples of these actions are. supermajority voting provisions, staggered terms of directors, golden parachutes and poison
 pills that give present shareholders the right to buy at a substantial discount.
existing shareholders can not accept this and the firm will not issue equity. This means that only low-value firms will issue equity.
Since valuable assets is what makes secure debt secure, one would expect that firms with valuable assets would use this framework
and issue more debt than firms with less collateral. Growthfirms and knowledge firms will, according to this theory, have more equity.

Noe (1988) showed that by modifying Myers and Majluf's model it can be shown that some firms will prefer equity to debt. He
increased the number of categories of firms from two to three. This is because issuing debt would make investors believe that the firm
has lower current assets than it really has. Debt would thus be underpriced.

5.4.4.2          Timing.
Myers' (1984) argument for why equity is expensive is that new shareholders believe that the manager will only issue new equity when
                            62
the shares are overvalued. He claimed that the manager is able to time the issue in such a way that it benefits the existing
shareholders. New shareholders will therefor require to get a large share of the firm in compensation for the new equity. There are
strong empirical evidence that support this theory.

                                                                                                               63
Smith (1986) surveyed research on abnormal returns after the market gain knowledge about security offerings. The result was that in
no cases was the abnormal return significantly positive. An announcement had either no or a negative impact on the abnormal return.
For common stocks and convertible preferred stocks the reaction was negative by 3-4%, while for preferred stocks the result was not
significantly different from zero. He believed that the main reason for this is that managers are more inclined to issue equity when the
firms are overvalued. This is the same argument as Myers (1984) used. Since the manager has superior information about the true value
of the firm, an equity issue announcement is believed by the market to signal that the firm is overvalued and the market adjusts the
firm's value by lowering the shareprices. The stockprice change will be larger the more unpredictable the announcement is. If , however,
a common stock offering is combined with retirement of convertible debt , the markets reaction is not significantly different from zero.

Ritter (1991) found long-term under-performance of common stock subsequent to IPOs. He said that the manager takes advantages of
opportunities and that this supports Myers' theory. The shares were actually overvalued, thus the under performance.

Jain and Kini (1995) investigated the change in operating performance by firms as they made the transition from private to public
                                    64
ownership through public offerings . They found that IPO firms initially had high market-to-book and P/E ratios, but these declined
after the IPO along with the EPS. They found a decline in post -issue operating performance, measured as operating return on assets
and operating cashflow deflated by assets relative to the pre IPO levels. The firms experienced high growth in sales and capital
expenditures. Jain and Kini said that the poor results could not come from a lack of opportunities. They listed three possible
explanations why firms' performance fall after an IPO.
 Increased agency costs after the IPO (Jensen and Meckling).
 Managers make the accounting figures look especially good prior to an IPO.
                                                                           65
 The issue is timed to periods of extraordinary good performance. (Myers)

Loughran and Ritter (1995) found similar results. They found that between 1970 and 1990 IPOs had an annual return of 5% and SEOs
                                                                                                                                   66
had an annual return of 7%. An investor would have to invest 44% more in order to have the same return as in a non -issue firm.
Loughran and Ritter found, just as Jain and Kini did, that firms issue after a period of growth. They said that while sales grow, the
cashflow did not.

Mikkelson, Partch and Shah (1997) found that while the firms that went public exceed other firms in income scaled by assets o r sales
prior to the IPO, their income fell below other public firms after the IPO.

Rao (1994) said that there are negative abnormal returns associated with security offerings. He gives the     following list of possible
explanations.
 The offering brings the firm away from its optimal capital structure.                                        Difficult to test.
 Cash inflows must be matched with investments. Empirical studies show that the market does not believe       Supported empirically.
 that the firm will spend all the new cash.
 Unanticipated announcements. The price change is only a function of how unanticipated th e offering          Partly supported.
 was.
 Asymmetric information. The manager offer securities when they are overpriced.                               Supported empirically.
Table 8 Negative returns associated with security offerings.
Both the second and the fourth suggested explanation support Myers and Majluf (1984) and Myers (1984).


62
     According to Myers (1984) there is empirical evidence for this too.
63
     For the results when the offerings were IPOs, see section 5.5, «IPOs.», page 39.
64
     The period of the empirical test was from one year prior to the IPO to five years after.
65
     More on this theory follows below.
66
     The results were for five years after the issue.
Kooyul, Kim and Stulz (1996) found however no support for the timing theory. They did however study firms in general, not
specifically IPOs, so the result may not necessarily be valid in the context above.

Pagano, Panetta and Zingales (1998) did however find evidence for Myers timing theory. They found that both investments and
profitability decreased after the IPO. Prior to going public, those firms grew faster and were more profitable than other firms. Pagano,
Panetta and Zingales also found that IPO firms have a high market to book value, just as Jensen and Meckling argued.

In this context one particular capital market anomalous should be mentioned. Chan, Karceski and Lakonishok (1998) confirmed what
others have found several times before. There is a January effect. This means that if the firm wants to exploit the hot -market theory in
the short run then equity should be issued in January.

Lerner (1994) examined venture capitalists' ability to time IPOs to periods where equity values are high and use private information
when values are low. He found that venture capitalists go public when the firmvalue is at its absolute short-run peak. This too
supports Myers' theory.
The problem may however be more complex than this. Gompers (1996) found that the choice of venture capital firm matters when the
firm goes public. Young venture capital firms take companies public earlier than older firms in order to establish a reputation and to
successfully raise capital for new funds. The young companies have typically been in the board shorter, hold smaller equity stakes and
time the IPO to fit their next project.
This last point is a very important one because Bayless and Chaplinsky (1996) claim that the issuer would forgo a large amoun t of
capital if the firm issues in a cold market instead of in a hot market. Hot markets are markets where the volume of issued securities is
large. On average one would get 2.33% more capital from selling the same amount of securities in a hot, rather than in a cold market.
Loughran, Ritter and Rydquist (1994) documented that for the last 20-30 years 14 of 15 countries had a positive correlation between
annual volume of IPOs and the level of the stockmarket. This means that it may not be the manager who acts opportunistic, but the
venture capitalists. However, the effect is the same for potential new shareholders who consider buying the new shares.

5.4.5         Reducing Agency Costs Between Existing VS New Shareholders.

5.4.5.1         The Pecking Order.
It will be shown below that there are ways to reduce the agency problems between existing and new shareholders too. The Pecking
order theory comes however up with another and more drastic solution. Avoid equity financing all together for as long as possible.
This means that the firm should not eliminate the use of outside equity, but it should only be used as a last resort.
Myers and Majluf (1984) said that outside equity is so expensive that firms may choose not to go into new-equity financed projects at
all. In order to avoid this underinvestment problem. the firm should finance its new project with cash, secured debt, or by selling
securities. If this kind of financing is possible the firm can take on all NPV > 0 projects. Myers and Majluf said that the old shareholders
are better of ex ante and on average ex post, if the firm takes on all NPV > 0 projects. This shows that the asymmetric information
between the manager and the new shareholders have economic consequences for the shareholders. In order to avoid this, the firm
should issue securities that are as safe as possible. That way the new investors will not need to charge an extra risk premium and the
firm can take on more profitable projects. This is equivalent with minimising the dead -weight loss, C(e), in Froot, Scharfstein and Stein's
                 67
profit function.

                                                         P = MAX [f(I) - I - C(e)]

Myers and Majluf called the sum of cash, securities and borrowing opportunities slack. If the firm has slack it can avoid issuing stocks
altogether and it can especially avoid issuing stocks when the firm is undervalued. If projects are financed with slack potential
asymmetric information between the manager, existing shareholders and new shareholders will not affect the financing or investing
decisions.
Myers (1984) based his theory on empirical work that had been done at the time. He took Myers and Majluf's (1984) conclusion a bit
                                                                            68
further and said that a theory that is competing with the static trade-off hypothesis is the Pecking order theory. In short the Pecking
order theory suggests that:
                                                    69
 Firms prefer to use internal capital to external.
 The dividend policy reflects the investment opportunities .
 If the dividend payoffs and a reduced profitability mean that the firm does not have enough cash from operations for its
    investments the firm will first draw on its cash balance or marketable securities portfolio.

67
     See section 4.2, «The Framework.», page 14.
68
   The static trade-off hypothesis is the assumption that there exist an optimal level of leverage. This optimal level takes into account
  the bankruptcy costs, financial distress costs and agency costs of debt as well as the direct and indirect costs of equity. It means
  minimising the WACC. About WACC, see section 5.3.2, «Why Risky Debt Matters.», page 24.
69
   Myers (1984) say that this is not just a theory, but an empirical fact. A lot of work had been done in the field of finance t o explain
  firms’ capital structure choice. The focus of this work had been to expan d the model that Miller and Modigliani presented 20 years
  prior to Myers’ Pecking order, that is the trade-off between debt and equity financing. Little effort had been spent on understanding
  the third source of capital, the internal capital. In this sense, Myers’ Pecking order was a fresh breath into a somewhat stalled
  financial topic.
     If the cash and the securities do not raise the required amount of funds the firm has to use external capital. In such a case the firm
      will issue the most secure securities first. This means that the firm will first use secured debt, then hybrid debt and only if the firm
      can not get any more out of these sources will it issue new equity.

According to the Pecking order theory the firm does not aim at any particular leverage, instead it uses the cheapest capital available. If
the firm manages to always finance its projects with internal capital or debt and no equity it may end up with a capital structure similar
to the one in M&M's 1963 paper, with a lot of debt. However, the reason for that capital structure is not the tax benefit, but the fact that
the cost of debt is lower than the cost of equity. While the debt VS equity problem can be used to understand the agency costs and
thus the dead-weight cost of external financing, it is really the Pecking order that gives the full picture in Froot, Scharfstein and Stein
framework since it includes internal capital too.

5.4.5.2          The Pecking Order, Empirical Tests.
The Pecking order does however not do to terrible well when it is tested directly.
Jung, Kim and Stulz (1996) compared three different explanations for why some firms issue debt and others equity. The three models
                               70                      71                        72
were the Pecking order theory , the agency model and the timing model . Their findings supported the agency model. Firms with
valuable investment opportunities will prefer equity financing and the leverage will be less the better the investment opportunities are.
Equity financing has both direct and indirect consequences for the manager. The direct consequence is increased discretion, w hich he
values. The indirect consequences are that unused funds make the firm a take -over target. In addition, unused funds can tempt the
manager to overinvest. Because of this the monitoring costs will increase. Jung, Kim and Stulz found little support for the Pecking
order theory.
Helwege and Liang (1997) too found little support for the Pecking order. The probability of obtaining external funds was unrelated to
the shortcomings of internal sources of capital. Helwege and Liang modified Myers and Majluf's ranking and tested their model
empirically. They set up a theory where internal capital was the cheapest form of capital followed by bank loans. Bank loans are
preferred to bonds because of the monitoring the banks do. There is thus less asymmetric information associated with bank loans.
Risky debt offers little advantage over equity and equity should be preferred. Helwege and Liang's findings did however not
correspond with their theory. They found that some firms issued equity when they could have obtained bank loans. These findings
violate Myers' theory too. Helwege and Liang also found that private debt was the most common source of capital, although the
amount was decreasing during the sample period. The fastest growing and the most profitable firms used public debt or public equity.
Helwege and Liang's conclusion was that firms with greater cash surpluses tend to avoid raising private debt, not public debt or
equity. This, they said, supports the static trade-off theory in that there is a target capital structure.
Martin (1996) said that despite theories that claims that firms with growth options should finance themselves with equity, there is little
empirical evidence to support these theories, which includes Myers' theory.

Christie (1994) made an observation that supports the Pecking order in an indirect way. According to the Pecking order the firm should
use the safest sources of capital first. The safest of all forms of capital is the internal capital the firm gets from assets in place. This
means that paying out dividends is bad thing. Christie said that if it was true that firmvalue is a function of dividend changes, an
omission would have the largest effect. This, he found, was however not the case. He found that the reaction pattern is U-shaped. This
means that an omission gives signals that are not included in normal signalling theories or in agency cost theories. Christie believed
that the best explanation is that the stock market does not believe that the omission is permanent, while a dividend cut may on the
other hand be permanent.
Barclay and Smith (1995) also found evidence of the Pecking order. They found that firms with growth opportunities have less long-
term debt and that large firms and regulated firms have more long -term debt. This, they say, was consistent with Myers (1977).

5.4.5.3          Timing.
To solve the asymmetric information problem Myers and Majluf said that the firm should issue stocks in periods where there is no
information asymmetry and keep the money as slack. This is of cause not necessarily a good solution. For one thing, this will lead to an
overinvestment problem. The manager will have money «lying around» and he will be eager to spend it. He will want to spend the
money for two reasons. First it is in his interest to overuse corporate money, secondly, a firm that has unused capital can be a take
            73
over-target.
In addition, issuing stocks in periods where there is little asymmetric information and not spend the money will only be possible when
there is asymmetric information. No investors will be willing to buy stocks in a firm that does not have any good projects at the
moment, but hopes that good opportunities will come in the future. The only way the manager will ge t the investors to put their money
in the firm is by saying that the firm actually has good investment opportunities . But this is of cause only possible when there is
asymmetric information.

Bayless and Chaplinsky (1996) s aid that there is little evidence to support the theory that it is possible to time the issuance to periods
where the asymmetric information is low. As mentioned above, Jung, Kim and Stulz (1996) did not find evidence for timing either.


70
     Myers’ (1984) model is based on the asymmetric information between existing and new shareholders .
71
     The manager acts in self interest, Jensen and Meckling (1976) and Stulz (1990).
72
     Firms experience long-term periods with underperformance after issuing equity.
73
     See section 7.1.2, «Does the Firm Become a Take-over Target?», page 45.
5.4.5.4          Contracts.
Myers and Majluf (1984) said that one way to solve the problem of expensive equity is to make compensation plans and a corporate
culture that focuses on long-term value and not short-term effects of asymmetric information.

5.4.5.5          The Manager's Ownership.
Still another possible solution to the problem, according to Myers and Majluf (1984), is to guarantee that the manager shall always
have the same pro rata share of all new share issues. The problem is that the manager sooner or later will end up with a huge amount of
undiversified capital invested in the firm. Empirical findings suggest that this solution to the over-investment problem is not used.

5.4.5.6          Capital Structure.
Smith (1986) said that retiring debt is a signal of increased future cashflow. Brennan and Kraus (1987) said that it is possible to solve
Myers and Majluf's problem. If the firm issues enough equity to finance the project and to retire debt the firm will signal that it is a high
value firm.74

Constaintinides and Grundy (1989) used another technique to arrive at the same conclusion. They said that instead of retiring debt the
firm should use the new equity to finance the project and buy back old equity. It will be costly to understate the true value of the firm.
Ikenberry, Lakonishok and Vermaelen (1995) said that in 1990 the amount of capital repurchas ed was equal to almost half of the amount
paid out as dividends which shows that the amount was considerable. Even more interesting in this context is the fact that in the
period from 1985 to 1993 the amount of capital repurchased was nearly three times th e amount raised in IPOs? Ikenberry, Lakonishok
and Vermaelen came up with several explanations for the stock repurchasing: They assumed that the most plausible explanation was
signalling.
  Excess cash distribution.       Take-over defences.            Signalling.
  Capital structure adjustments.  Substitution for dividends. Wealth expropriation.
Table 9 Explanations for the stock repurchasing.


5.4.6          Summary.
Below is a summary of the facts about underinvestments due to distrust between existing and new shareholders and what the firm can
do to reduce this distrust.
 Facts about the consequences of underinvestments .
 Firms do prefer retained earnings.
 In order to reduce the cost of equity the firm should emphasise the following things:
 Retire old debt when it issues new equity.
 Buy back old equity when it issues new equity.
 Retire convertible debt.
 Use a reputable venture capitalist.
                                75
Issue equity in hot markets.
The compensation plan should focus on long-term value.
The manager should own a large portion of the firm's shares.
Table 10 Avoiding the agency problems of overinvestments.
In addition one can say that the firm should use Majluf and Myers' argument and use slack instead of issuing capital. This argument is
however not a very good one when one wants to compare Froot, Scharfstein and Stein's framework with the old fashion going -to-the-
capital-market strategy. Froot, Scharfstein and Stein's framework is after all an extension of Majluf and Myers theory and Myers
                      76
Pecking order theory.

Froot, Scharfstein and Stein's main argument was that the firm should co -ordinate its investment and financing opportunities. Their
hedging strategy may however be redundant since equity is at its cheapest in hot markets and when the capital market is hot t he
investment opportunities will probably be at its best too. One can also turn the argument around and say that the capital market is
willing to trade money for shares at a lower price when the investment opportunities are good. There must after all be a reas on why the
capital market is hot. As one can see, the capital market is already using Froot, Scharfstein and Stein's framework since it co -ordinates
the financing opportunities with the investment opportunities.

5.5            IPOs.
Because of the special circumstances surrounding an IPO, these transactions deserve special attention.

The initial public offering is special in two ways Smith (1986) said.
 The uncertainty about the market-clearing price is greater.

74
     This is the firm quality problem mentioned in section 5.4.4.1, «Firm Quality.», page 35.
75
     This will probably not reduce the asymmetric information, but it will limit the cost of capital.
76
     See section 4.1, «The Origin.», page 14.
 Examination of stock price reactions to initial announcements is impossible.
Smith surveyed different empirical work and found how much stocks are underpriced when they are issued. The results of the differen t
studies differ, as did the sample size and the time periods. The estimated underpricing ranged from 10 to 50%! This potential huge loss
for older shareholders is eliminate if Froot, Scharfstein and Stein's framework is used.

5.5.1          Firm Size and Age.
Mikkelson, Partch and Shah (1997) found that small and young IPOs underperform after going public while larger, older firms perform
similar to other publicly owned firms. This means that larger, older firms do not have to rely on Froot, Scharfstein and Stein's
           77
framework.

Pagano, Panetta and Zingales (1998) studied why Italian firms went public. They found t hat the likelihood of a firm going public
increases with firm size and the industry's market to book value. The first is consistent with the theory that there are econ omics of scale
in going public 78 and it supports Froot, Scharfstein and Stein theory becau se it is believed that there are more asymmetric information
in smaller firms 79, so equity would be more expensive for them. Pagano, Panetta and Zingales found that firms do not go public in order
to get money for investment but to rebalance their accounts. According to these findings investors are right when they demand a
premium for IPOs. The fact that firms with high marked-to-book ratios go public support the theory that growthfirms use equity. This
means that growthfirms do not need to implement Froot, Scharfstein and Stein's framework.

5.5.2          Why IPOs Underprice.
Noe and Rebello (1996) said that underpricing is the last resort, it will only be used when informational asymmetries are great.

Rao (1994) came up with a couple of explanation for why IPOs are underpriced.
 Pricing IPOs is difficult. The underwriters prefer to err with a price that is too low rather than one that is too high.
 The investment bank wants to be protected from legal liabilities and damage to reputation.
Rao also came up with a couple of explanations for underpricing that have not been supported empirically. These were:
 Risk-averse underwriters.
                                                                                                                  80
 The firms must pay to get access to the underwriters' knowledge about investors and the capital market.
                                                                                                                  81
     The investment banks must underprice the shares to ensure that the informed investors stay in the market.

By using Froot, Scharfstein and Stein's framework the firm will avoid all of these problems.
Booth and Chua (1996) said that one reason for underpricing that is often given is that it reviles price elasticity of the stock. Booth and
Chua found however that broad ownership and liquid secondary markets are the reason. Broad ownership will in itself also contribute
to liquid markets. Both of these do increase investorborne costs, but investors also benefit from liquid markets. Underpricin g is a
positive function of ownership dispersion.
Using internal capital and hedging makes the question of ownership an irrelevant one.

5.5.3          Who Carries the Costs?
Brennan and Franks (1997) said that non-managers carry 75% of the underpricing costs. It is however predominately non-managers
who benefit from the IPO too, because the manager looses the benefits that the control gave him. On average, a large majority of the
shares owned by pre IPO shareholders are sold at, or right after the offering. In addition to the obvious costs that underpricing brings
about it brings with it some benefits too. If the stocks are underpriced the likelihood of an over-subscription increases. This makes it
possible to reduce individual blocks, which increases the liquidity. This again will decrease the required rate of return and thus
increase the shareprice. However, many, smaller shareholders will reduce the monitoring , which benefits the managers.

5.5.4          Underwriters.
Dunbar (1995) found that underwriter compensation matters for the total costs of an IPO. Normally the underwriter receives a cash
settlement that is a percentage of the offering price. Dunbar found that the underpricing cost, underwriter compensation and expenses
are negatively related to the use of warrants. This contradicted previous findings.

Carter, Dark and Singh (1998) found that the choice of underwriter matters too. The underperformance of IPOs' stocks relative to the
market over a three-year holding period is less severe for IPOs handled by more pres tigious underwriters. These stocks are also less
short-run underpriced. The firm should therefor use a prestigious underwriter if it wants to use the capital marked instead of Froo t,
Scharfstein and Stein's framework.




77
    Section 5.5.5, «The IPO and SEOs.», page 40 covers underpricing for SEOs and find that there are non. This means that the older
  firms do not have to rely on Froot, Scharfstein and Stein’s framework after the IPO either.
78
   Se section 5.6, «Transaction Costs.», page 41.
79
     See Miams arguments in section 10.1, «Research on the Topic.», page 67.
80
     This is known as the Baron model.
81
     This is the Rock model.
Several studies have been made on the underwriters' actions after the stock issue. The conclusions may actually explain must of the
underpricing problem.
Ruud (1993) studied the price support phenomena and tried to come up with an explanation. He started by referring to several studies
that showed that IPOs are indeed underpriced. In his own study he looked at the distribution of the returns and found that th ere is a
peak at zero. This means that most IPOs are neither underpriced nor overpriced. The interesting thing is that the distribution is
skewed towards positive returns with almost no negative returns. He suggested that the reason is the stocks that start to tra de at
prices lower than the offerprice are supported. This may explain why one observes a significant positive return. If no stocks had been
supported the returns may have been normally distributed and no one would never find any evidence of general underpricing . Ruud's
findings are supported by later research.
Schultz and Zaman (1994) studied underwriters aftermarket support during the three first days after an IPO. The underwriters can buy
stocks in order to increase the demand and thereby hold the price higher than it would have been without the interventi ons. They
                                                                                     82
found evidence that the underwriters do in fact support the stockprices. For hot IPOs the underwriters do not intervene. If however
                                                                 83
the IPO is cold the underwriter trade much at the inside bid. Schultz and Zaman also found that the underwriters buy more of the
stocks that do not have a positive initial return.
Mikkelson, Partch and Shah (1997) found that underwriters attempt to stabilise the prices around the offer date. In the case of IPOs this
is accomplished by placing a limit order to purchase shares with the specialist on the exchange.

This means that it is reason to question the entire foundation that much of the capital structure literature is built on. One should in
particular be aware of the fact that Myers actually started out with what had been found empirically, that is that equity is underpriced.
He then made the Pecking order theory to try to explain what had been found empirically. If it turns out that it is the underwriters'
interventions makes equity appear expensive, Froot, Scharfstein and Stein's framework may not be much else than a nice calculus
exercise. This may also explain why researchers have had a hard time finding any good explanations for why IPOs are underpric ed.
They may not be.

5.5.5          The IPO and SEOs.
Smith (1986) said that in his survey he found that the underpricing of SEOs is almost zero. If this is true, the entire foundation of Froot,
Scharfstein and Stein's framework falls apart completely. All firms that have already gone public may not care about the framework at all
because they can just as well issue equity whenever it is needed.

Jegadeesh, Weinstein and Welch (1993) investigated the signalling hypothesis associated with IPO underpricing. They said that there
are two signalling theories. The first is based on the fact that the manager has superior information. The second theory expl icitly
considers the possibility of future equity issues in deciding on IPO pric es. High quality firms underprice the IPO to be able to raise
capital favourable in a SEO. In the long run the more expensive SEO will be more important than the underpriced IPO.
Jegadeesh, Weinstein and Welch confirmed the latter hypothesis, but it is weak. Underpriced IPOs are more likely to return to the SEO
market and with larger offerings. However the aftermarket return explains future SEO better than the IPO explains it.
Jain and Kini (1995) rejected the entire theory that firms underprice the IPO and make a seasoned offering later when the asymmetrical
information is lower.

There is actually a way that the firm can use this to reduce the cost of capital in a SEO. Chaney and Lewis (1998) found that how well a
stock performs after an IPO is positively related to how incomes are reported. It is possible for the firm to present earnings that are
smoother than what the cashflow would have suggested. By smoothin g the reported earnings the firm will perform well in the stock
market after the IPO. This will lead to cheaper capital in a SEO. Unlike several of the other suggestions to how the firm sho uld reduce
the cost of capital this method is not based on removing asymmetric information or agency problems. If fact this suggestion is actually
based on the fact that there are asymmetric information between the manager and the investors.

Welch (1996) said that there is another way the firm can influence the amount it can raise during a SEO. A high quality firm can wait
between the IPO and the first SEO. During this period the market will have time to find out that the firm is really a high quality firm. A
low quality firm will not dare to wait and thereby risk that the market understands that it is a low quality firm. Welch found empirical
evidence that supports his theory.

Mello and Parsons (1998) said that when the firms make an IPO, it should have the final ownership structure in mind. This must be
reflected in the way the IPO is marketed. In their model the firm should first address the IPO to small investors. The IPO sh ould not be
used to sell control. A controlling block should be sold later durin g a SEO. The monitoring that a controlling investor can perform will
benefit all investors. It is acceptable to sell the controlling block at a discount.

If SEOs are not underpriced and the firm is already publicly owned the firm does not need to implement Froot, Scharfstein and Stein's
framework. If the firm is a privately owned growth firm that wants to acquire equity 84 the firm should have the SEOs in mind.




82
     Hot IPOs are offerings where the stocks are traded above the offering price.
83
     The inside bid is the highest price any dealer will pay.
84
     It was shown earlier that growthfirms do not need to use Froot, Scharfstein and Stein’s framework, they can use the capital market.
5.5.6          Privately Held Firms.
Boye and Dahl (1996) said that according to a British study unlisted firms are traded at a discount of 20%. They assumed that the
figure for Norway is 30-40%. This means that privately held firms should indeed rely on Froot, Scharfstein and Stein's framework
instead of issuing more private equity.

5.6           Transaction Costs.
When a firm wants to decide whether to use Froot, Scharfstein and Stein's framework or external capital all costs must be cou nted for.
                                                                                                                         85
This includes a comparison between the transaction costs in the capital market and the costs of trading with derivatives.

Titman and Wessels (1988) studied determinants for the choice of capital structure. They found that short -term debt was negatively
related to firm size, reflecting the high costs that small firms face when they issue long-term debt. They said however that the
transaction costs are small compared to other leverage-related costs and that their impact may not be particularly significant.

Rao (1994) said that the underwriter's fee is the difference between the price the underwriters get for the shares and the price they pay
for them. The underwriters typically buy shares at 15% discount. There are economics of scale in the underwriter fee. Rao sho wed that
the fee and the administration cost differ for different securities and different sizes of the offerings.
 Type of offering          Value (in USD)            Fee (in % )               Adm. Costs (in % )
 Stock offerings           < 2 mill                  7-13                      4-9
                           > 2 mill                  <5                        <1
 Rights offerings          < 20                      Lower than for general offerings
                           > 20                      Same as for general offerings
 Bond offerings            <5                        Not much less than for equity
 Bond offerings            >5                        <2                        <1
Table 11 Direct Costs of Offerings.
Tufano (1996) supported Rao when he said that it is reas onable to believe that informational asymmetries and transaction costs are
larger for small firms.

80% of all firms that issue securities use an underwriter instead of rights offerings, according to Smith (1986). This is despite the fact
the using an underwriter is 3-30 times more expensive. The reason is that the investment bank not only provides a distribution channel
between the firm and the investors, it also performs a monitoring function. Because of this investors are willing to pay more for
securities that are sold through an underwriter. This means that the new investors carry the costs. They are also the ones th at benefit
from it. If this price is fair this is thus not a market imperfection and it should not influence the firm's decision to whether it should use
Froot, Scharfstein and Stein's framework or not.
Rao (1994) said that the investment bank also helps the firm in the choice of capital, timing, pricing and characteristics. A nother very
important reason for having an underwriter is the fact that the underwriter may agree to buy the remaining if the offer is not fully
subscribed. However, the underwriter may choose not to guarantee this. This is often the case for new, speculative projects, such as
IPOs. The firm may voluntarily choose to drop the guarantee too if it is confident that all the shares will be sold and it wa nts to keep
the expenses as low as possible. This is because the underwriter buys the shares from the firm at a discount.
If the firm wants to issue several times, it can use a master registration statement. This means that it registers the entire issue once and
it can then choose how much it wants to issue and when it wants to make the actual offerings.
Smith also said that competitive bid offerings on average have 1.2% lower total costs than if the firm used negotiated offers. Despit e
this, most firms choose negotiated offerings, probably because of the monitoring .

Kothare (1997) found that fewer than 10% of US firms issue seasoned equity directly to current shareholders through rights offerings
despite the fact that it is cheaper. The reason is that rights offerings impose a significant indirect cost on issuers by red ucing the
liquidity of the issuer's stock, contrary to a public offering, which increases the liquidity. With rights offerings the spread increases
which reduces the liquidity.

All this suggests that transaction costs are considerable for small firms .


5.7           Conclusion.
When one wants to draw conclusions from this chapter it is important to remember that there are still white spots on the fina ncial
theory map when it comes to fully understand the choice of capital structure. Myers suggested that the the ory of optimal capital
structure is obsolete. When Myers' own alternative, the Pecking order theory, has been tested the conclusions have however be en
ambiguous.

The costs of bankruptcy, financial distress and agency problems make debt expensive. There are however actions the firm can take in
order to reduce these costs. In addition, debt reduces the overinvestment problem of equity financing. In general large firms , firms in
slow growing industries and firms with tradable collateral uses debt. The fact th at interest payments are tax deductible is of cause a
great advantage.


85
     These costs are studied in section 6.1.1, «Transaction Costs.», page 42.
Two kinds of agency problems makes equity financing expensive, one results in overinvestments and the other in underinvestmen ts.
There are however actions that the firm can take in ord er to reduce the costs of equity financing too. It seems however that
growthfirms can acquire external equity at a fair price.

These findings suggest that not all firms need to consider Froot, Scharfstein and Stein's framework, they can use the capital market
instead. However, other studies have found that firms in general prefer to finance projects with retained earnings. This stro ngly
supports Froot, Scharfstein and Stein's paper.

A different area of research has however come up with findings that may make Froot, Scharfstein and Stein's framework redundant.
Resent research on IPOs and SEOs suggests that the earlier findings that indicated that particularly IPOs were underpriced ma y only
be a result of bad research design and statistics. If this is true the entire foundation that Froot, Scharfstein and Stein's framework were
built on disappears. This means that the main argument against external capital is the relatively high transaction costs tha t small firms
in particular face.



 6.            WHAT ARE THE COSTS OF FROOT, SCHARFSTEIN AND
                         STEIN'S FRAMEWORK?
Froot, Scharfstein and Stein said that external financing is expensive and that the firm should use internal capital to finan ce new
projects. This means implicitly that hedging with derivatives is cheaper than acquiring external capital. In the section above it was
shown that the old theories of costly external financing do not necessarily hold. The price of external capital may after all be fair. What
then about the costs of Froot, Scharfstein and Stein's framework? Is it true that financing investment opportunities with derivatives is
less costly than using external capital?
Again, there are both direct costs and indirect costs involved.

Derivatives in Froot, Scharfstein and Stein's framework are discussed in more detail later, 86 this chapter only deals with the costs so
that the costs of using Froot, Scharfstein and Stein's framework can be compared with the costs of using the capital market.

6.1           The Direct Costs of Hedging.
Marshall and Larson (1983) said that there is a great deal of literature on the cost of hedging and they conclude that hedgin g is cheap.
This is especially true for futures contracts. However, it is not costless.

6.1.1          Transaction Costs.
Errington (1993) said that in active cash markets the bid-ask spread is only 12.5 basis points. Decovny (1992) said that for illiquid
instruments the bid/ask spread would increase. One thing that points in the direction that hedging is not too expensive is that
Ackermann, McEnally and Ravenscraft (1997) found that hedge funds out perform mutual funds. They did this even when they were
risk adjusted.

Abken (1994) said that firms prefer over-the-counter instruments to dynamic hedging. OTC instruments are more convenient, they are
however more expensive.

6.1.2          Opportunity Cost.
There are other costs involved in holding derivatives in addition to the transaction costs. There are opportunity costs of ho lding
instrument that do not pay interest. Smith, Smithson and Wilford (1988) called this the cost of carry.

6.1.3          Liquidity.
One very important thing is the fact that the liquidity of the second -hand market is very low for many instruments. If a firm uses Froot,
Scharfstein and Stein's framework and the risky asset(s) moves so that the business environment for the firm is favourable an d the
derivatives become valuable the firm may have problems liquidating the derivatives and get the money it needs for the investments.
                                                                                                    87
This may not be a large problem because 98% of all futures contracts are closed before maturity. Forward contracts are also usually
                           88
settled before maturity.

6.1.4          Managing the Derivatives Portfolio.
There are also costs associated with managing the derivatives trading and keeping the right derivatives portfolio. Wisbey (19 92) said
that there are several options when a firm wants to establish a risk management system. They have different advantages and different
costs.
 Action                            Pro                                  Con
 Develop an in-house system.       One gets what one wants.             Costly.

86
     See section 9, «What Instruments Should Be Used?», page 52.
87
     See section 9.2.1, «About Futures.», page 53.
88
     See section 9.4.1, «About Forwards.», page 54.
                                 Ones owns priorities.                       Little value added for the costs.
Use an external system.          Known initial costs.                        Others priorities.
                                 Cheaper.                                    What system should one choose.
Use an external bespoken system. Professionally.                             Expensive.
Table 12 Pros and cons of different systems solutions.
It is not uncommon to invest less than optimally in such systems. Reynolds (1995) said that too little management resources are used
to monitor derivatives trading. Trading in derivatives can be done without approval, while operational decisions involving fa r less
money are monitored. While outside advise is used for other matters, derivatives trading is considered to be a «do it yourselves»
matter.
Another important aspect of the cost of hedging is the fact that not all hedges are efficient.
                 Costs

                                             Efficiency frontier

                                   
                                       
                                                               Basis risk


Figure 15 Efficient derivatives.


Hull and White (1993) said that sophisticated models are not needed in trading, only in research.

6.2           Conclusion.
Just as for the traditional, capital market financing of project Froot, Scharfstein and Stein's framework carries with it cos ts too. The
direct costs are probably far less than the direct costs of external capital since trading in derivatives is quite cheap, especially for the
liquid instruments. The opportunity costs can however be considerable and off-balancesheet instruments is preferred. Froot,
Scharfstein and Stein implicitly assumed that the costs of hedging were less than the costs of acquiring external capital. Th is is most
likely a fair assumption.

There is however an overinvestment problem but since this problem does not deal directly wit h the raising of funds it is discussed
      89
later.




                           7.          WILL STAKEHOLDERS APPROVE?
Agency problems are one of the major sources of the market imperfections that make external capital more expensive than internal
capital. These agency problems therefor increase the firms need for Froot, Scharfstein and Stein's framework. The framework does
however carry with it new agency problems, as the following sections show.

7.1           The Manager.
Even if Froot, Scharfstein and Stein's framework may increase the value of the shareholders' stocks, the manager may still choose not
to implement the framework. The manager will not willingly implement such a financial strategy if it decreases his utility. A s mentioned
before, Rao (1994) said that it is the board's job to supervise the manager on behalf of the shareholders, but many boards fail however
to do so.
There are two reasons why managers may not want to implement Froot, Scharfstein and Stein's framework.
First, if Froot, Scharfstein and Stein's framework increases the firms risk and the manager faces serious consequences if the firm gets
into financial difficulties the manager may not want to implement the hedging strate gy. This argument depends on the managers risk
aversion.
Second, the fact that the firm has to invest money in the derivatives market may turn it into a take -over target, which may also involve
unpleasant consequences for the manager.
In the first problem it is the fact that the firm may end up with too little money that is the problem while too much money is the problem
in the latter argument for not implementing Froot, Scharfstein and Stein's framework.




89
     See section 7.2, «The Shareholders.», page 46.
7.1.1        The Manager's Risk Aversion.
According to financial theory the firm can implement a risk management program for several reasons, however reducing unsystematic
risk is not one of those reasons. Investment theory suggests that investors should reduce such risks through diversification. Smith
and Stulz (1985) argued that managers are risk averse and that they may still implement a risk management program to protect
                                         90
themselves against undiversifyable risk.
In order to find out if managers will even consider Froot, Scharfstein and Stein's framework one must first find out if the manager has
any reason to be risk averse. This fear is the product of the consequences the managers face if the firm does in fact get into financial
difficulties and the likelihood that Froot, Scharfstein and Stein's framework will increase the probability of a bankruptcy or financial
distress.

The first question one must ask in order to find out if the manager will approve to Froot, Scharfstein and Stein's framework is: Does
increased corporate risk really hurt the manager? Gilson (1989) found that out of the managers employed two years prior to a financial
distress, only 34% remained in their job two years after the restructuring. Even worse, for the managers, is the fact that none of the
displaced mangers held a similar position in a listed firm three years after their departure.
Gilson and Vetsuypens (1993) made an empirical s tudy of the consequences of financial distress and bankruptcy for the managers of
                                        91
such firms. Contrary to earlier studies they found that managers incurred significant losses when their firms were in trouble. Almost
one-third of the managers in their survey lost their jobs, those who kept their jobs often took large cuts in salary and bonuses.
Houston and James (1993) found that less than a third of the CEOs in their sample of financially distressed banks retained th eir jobs.
Only one of 39 acquired a similar job in another exchange listed bank.
Cannella, Fraser and Lee (1995) studied banks in Texas and found that 2/3 of managers in non -failed banks continued to be employed
in the sample period. This means that the normal attrition rat e was 33%. However, only 22% of the managers from equity insolvent
banks regained jobs in the region. This clearly shows that the manager has reason to worry about financial difficulties.

The second question one must ask is about the probability of a bankruptcy or a financial distress. Froot, Scharfstein and Stein's
hedging theory may increase the probability of a bankruptcy or financial distress. This is because what separates Froot, Scha rfstein
and Stein's framework from other risk management theories is that they said that the cashflow should be optimised with respect to the
investment opportunities while other theories define risk management as a way to stabilise the cashflow. In the simple- and the no-
correlation models firms turn variable cashflows into fixed cashflows, just as in ordinary hedging. However, Froot, Scharfste in and
Stein said that if the cashflow correlates with investment opportunities , the firm should not hedge. They went as far as to said that the
firm may even benefit from speculating if the investment opportunities are extremely sensitive to the same risky asset as the one that
makes the cashflow fluctuate. Both these models and in particular the latter, increases the undiversified risk that the manager faces.
This is also true in the reversed simple model were the income from assets in place are fixed and the investment opportunitie s fluctuate.
The firm should redistribute the income so that is matches the fluctuating investment opportunities.
In addition to the increased risks associated with the redistribution of stable incomes are the imbedded risks associated wit h
derivatives, speculation can be very dangerous if something goes wrong. Unexpected market movements, model errors, less then
prefect hedges or hedging instruments etc can lead to serious problems for the firm.
This clearly shows why the use of options , as well as the extensions of Froot, Scharfstein and Stein's framework developed in this
thesis, are important as a safety net. These strategies will not only protect the firm from the costs of a potential financia l distress or a
bankruptcy, they also greatly reduce the managers risk. This will increase the likelihood that Froot, Scharfstein and Stein's framework
will be implemented.

These findings show that although investors may be able to diversify and therefor not care about undiversifyable risk, manage rs will
indeed care about such risks. From this perspective one can draw the conclusion that managers will be willing to implement th e simple-
and the uncorrelated model, but not necessarily the no-hedge, the speculation or the reversed simple strategies, unless they are certain
that the safety net provided by options really works. At first this may seem like an acceptable result for a theory as theoretical as Froot,
Scharfstein and Stein's. Two out of five versions of the framework will be implemented straight away. The problem is that there is no
difference between the simple- and the uncorrelated model and ordinary hedging where the aim is to stabilise and not to optimise, the
cashflow. The optimal cashflow will be a cashflow that is fixed. The manager may not want to implement the parts of Froot, Scharfstein
and Stein's framework that were revolutionary in a risk management perspective.

7.1.2        Does the Firm Become a Take-over Target?
The threat of too little money was the reason for concern in the problem above. Too much money may however also be a problem.
Jensen (1986) said that some firms become take-over targets because the manager inefficiently allocates free cashflow to unproductive
investments. Hendershott (1996) tested Jensen's (1986) hypothesis and he found that take -over targets do not have above average free
cashflow. However, the use of funds is important, since he found that target overinvestments play an important role in t he market for
corporate control.
Later papers take the same standpoint as Jensen. Jung, Kim and Stulz (1996) said that firms that have a lot of cash are take-over targets.
Giammarino, Heinkel and Hollifield (1997) agreed to this. If this is the case, Froot, Scharfstein and Stein's framework may turn the firm
into a take-over target. An example illustrates the problem. Suppose economic variables change in such a way that the firm's


90
    As mentioned before, Gèczy, Minton and Schrand (1997) found however no empirical evidence for this theory. However, while this
    may be true in general this section will discuss managerial risk aversion and Froot, Scharfstein and Stein’s framework explicitly.
91
   Jensen and Murphy (1990 a and b) among others.
derivatives become valuable. In this case the firm should use this money to invest because the movements in the economic variables
will also have increased the firm's competitiveness and therefor its need for investments. There are several possible reasons why the
firm may not want to invest right away. The manager may prefer to wait until the value of the derivatives increase further or he may
assume that the change in the economic variables are only a temporary phenomenon and that the business environment will not b e
effected in the medium or long-term. Thus the firm should not invest, but just keep its valuable derivatives portfolio. Either way, if
Jensen is right the firm is now a take-over target, even though it may very well be well managed. This is because the firm has valu able
assets that it does not use. This is a risk that managers may not want to take and they may reject implementing Froot, Scharf stein and
Stein's hedging strategy.

Jensen and Ruback (1983) surveyed empirical findings of stock returns during take-overs. For take-over targets there are large and
significant increases in the stock return after an announcement of a take -over. If stocks of unused capital motivated many of these
take-overs, it shows that investors really do not like unused money in the firms. This is bad news for the Froot, Scharfstein and Stein's
framework, because firms in their model will have to hold a certain amount of money in derivatives instead of in operations. Ward
(1994) confirmed Jensen's statement and said that during the 70's and early 80's, there were a build-up of large, low dividend
conglomerates with lots of cash invested in the moneymarket. In the late 80's, raiders targeted these firms.

Blanchard, Lopez-de-Silanes and Shleifer (1994) studied what happened to firms that got large cash windfalls, for instance from won
lawsuits etc. Here one can think of those cash windfalls as major changes in the value of the derivatives due to changes in t he value of
the risky assets. Those firms that chose not to spend the cash where acquired within a couple of years.

Bethel, Liebeskind and Opler (1998) found that raiders are not the only threat to the manager. Minority groups of investors a re actually
a larger threat. Of the firms on the Fortune 500 list in 1980 Bethel, Liebeskind and Opler found that 9% experienced leveraged buyouts,
8% hostile take-overs and as many as 33% experienced attempts by minority investors to influence corporate control. While the
activity of leverage buyouts and hostile take-overs decreased in the 90's, minority investors continued to try to influence corporate
policies. They also found that only firms that were performing poorly and firms that were diversified were attacked.
Mikkelson and Partch (1997) too found that the turnover is linked to p oorly performing firms.

Take-overs are serious for the manager. One thing that indicates that managers are afraid of take -overs is the fact that they require a
higher salary whenever there is a take-over threat. Agrawal and Knobler (1998) found that even though there is competition among the
managers, which drives the managers' compensation down, the take-over compensation is larger. The take-over compensation ranged
from USD 41.000 - 255.000. Martin and McConnell (1991) found that management turnover increases following take-overs. However,
the turnover typically occurred in firms that underperformed. Bethel, Liebeskind and Opler (1998) also found that take -overs are
followed by abnormally high rates of manager turnover. Mikkelson and Partch (1997) also found that the management turnover is more
excessive in more active take-over periods.

In addition, even if there is a take-over threat there are ways to cope with the problem. Giammarino, Heinkel and Hollifield (1997) said
that firms should use poison pills 92 or shareholder rights to protect them selves against general take-overs and green mail93 to protect
itself against raiders. This means that a general assembly that wants to implement Froot, Scharfstein and Stein's has means to protect
the manager against the take-over threat and motivate him to use the framework. Zwiebel (1994) went as far as to said that managers
prefer debt to decrease the take-over threat.

On the other hand, Giammarino, Heinkel and Hollifield (1997) said that the manager has an incentive to overinvest in anti-take-over
measures. This is because he has undiversified risk.
Comment and Schwert (1995) found that the level of take-overs fell dramatically at the end of the 1980's. The improved antitake-over
measures had increased the bargaining power of the managers and the shareholders received a premium during the take -overs.
However, the antitake-over measures did not prevent any take-overs, they only increased the bargaining power. So managers are in
fact faced with a risk.

The manager's ownership may however yet another time turn out to be an important variable. Denis and Denis (1994) found that no
firms in their sample where the manager owned more than 30% of the shares had ever been a victim of a hostile take-over.

7.1.3          Is Froot, Scharfstein and Stein's framework a Better Alternative Than to Go Public?
Before one can say that managers will not approve to Froot, Scharfstein and Stein's framework because it incre ases their
undiversifyable risk one must look at the alternative. If one assume that the firm is faced with investment opportunities , the manager
will only have two options , use Froot, Scharfstein and Stein's framework or go public. Doing nothing will surly not be a long -term
alternative. Pagano, Panetta and Zingales (1998) found that there was a significant increased turnover in control after Italian firms went
public. Faced with this fact the manager may consider implementing Froot, Scharfstein and Stein ideas and keep the firm privately held,
thus maximising his own utility.




92
     A poison pill provision gives the existing shareholders the right to buy shares in the new firm at a discount.
93
     The firm can buy back the shares of a substantial shareholder at a premium above the market price of those shares.
7.2         The Shareholders.
Agency problems are the main reason why external capital is more expensive than internal capital and therefor also the main reason for
using Froot, Scharfstein and Stein's framework. However, there are indeed agency problems associated with Froot, Scharfstein and
Stein's framework too. Many of the problems are actually the very same problems that cause outside capital to be expensive!
The most obvious agency problem is the overinvestment problem. As stated earlier, Stulz (1990) said that the moral hazard problem
between managers and shareholders is that the manager is willing to accept NPV < 0 projects because they will increase his perquisites.
In Froot, Scharfstein and Stein's simple model the investment opportunities are fixed and the income from assets in place fluctuate. The
firm can use two different hedging techniques. The firm can either use long -term contracts with its suppliers and customers that will act
as forward contracts. Agency problems can be eliminated by the way those contracts are written. Alternatively the firm can buy
forward, future or swap contracts with the opposite payoff of the operations. In good states the amount of money earne d, in excess of
what is needed for the period's fixed investments, should be used to buy derivatives. The manager will however have an incent ive to
spend the money on real investments, rather than on derivatives. He will overinvest if he feels that the size of the firm that he manages
is important for his social status. His eager to overinvest will be particularly strong if the compensationplan involves some kind of
convexity with respect to sales, income, market share etc. Timing will also be an issue and a source of an agency problem. Suppose that
the firm is an oil consumer, like an airline company. The amount of aeroplanes needed will not be correlated with the oil-prices, however
                                                  94
the profits will depend on the price of the fuel. Such a firm should hold derivatives that are valuable whenever the oil-price goes up.
But when should the manager sell the derivatives? A manager who wants to overinvest may be tempted to sell of the derivatives and
spend the money. If the manager sells them too early in the oil-price cycle the derivatives will not be as valuable as they will be later.
The firm may not even be in need of new planes yet.
In the uncorrelated model both the investment opportunities and the revenues from assets in place fluctuate. However, they fluctuate
independently. Just as in the simple model it is optimal for the shareholders if the firm's cashflow is transformed into a fixed
incomestream. This can be done the same way as in the simple model and the agency problems are consequently the same.
In Froot, Scharfstein and Stein's no-hedging model the cashflow from assets in place and the investment opportunities fluctuate and
they have a relatively high degree of correlation. It is in this model that Froot, Scharfstein and Stein probably make their most
controversial statement when they said that the firm should stay unhedged. In this case overinvestme nts are probably not a problem.
The reason is obviously that when the firm has money to invest it should invest and in periods where it is optimal to invest small
amounts the firm has little money to invest.
The speculation model carries with it another kind of agency problem. In this model the firm is supposed to use money on derivatives
in states where the cashflow from assets in place are already low. The derivatives will be sold during good times to finance
extraordinary large investments. The manager, as well as trade unions and suppliers, will probably not like such a hedging strategy,
because it increases their undiversifyable risk. So, even if the strategy will benefit the shareholders, the manager will, or can, not
implement it.

7.3         Will Other Stakeholders Approve?
Some of the following arguments have been mentioned before. They are however important in order to understand how the differe nt
stakeholders will react when confronted with a proposal to implement Froot, Scharfstein and Stein's framework.

Campbell and Kracaw (1993) said that all stakeholders benefit from risk management . According to investment theory investors should
diversify their investments to avoid unsystematic risk. Other stakeholders, like employees, managers and suppliers do not have that
opportunity and they will benefit from risk management. Smith and Stulz (1985) illustrated this when they said that debtholders favour
hedging since it reduces the probability of bankruptcy. In addition to the direct and indirect costs of bankruptcy, Shapiro and Titman
(1986) said that a bankruptcy means that the firm looses long-term relations with suppliers and customers. These relations have a value
and a loss of these relations is therefor a cost. This means that these stakeholders will dislike the implementation of the n o-hedging-
and the speculating models.

The problem with Froot, Scharfstein and Stein's framework is that in the no -hedging- speculating- and reversed simple models the firm
does not hedge. It manages its risk, yes, but it does not remove the risk. If all stakeholders agree to Froot, Scharfstein an d Stein's way
of thinking, that this sort of risk management will maximise the firm's sharevalue and if all stakeholders agree that sharevalue-
maximising is the best thing for everyone, then there is no problem. If on the other hand some stakeholders prefer the old -fashion risk
removing strategy, they will dislike an implementation of Froot, Scharfstein and Stein's framework.

7.4         Concision.
This section shows that the manager as well as other stakeholders will have no objections against the implementation of the s imple or
the uncorrelated models. In fact, lots of firms are already doing this, although maybe under a different name.
If the no-hedging, speculation or reversed simple models are to be implemented several things must first be done. An anti-take-over
strategy must be approved and implemented. Secondly the corporate culture must be so that it is accepted by both the managers and
trade unions etc that sharevalue maximisation benefits all. Only if the stakeholders believe that Froot, Scharfstein and Stein's framework
will lower their risk through increased financial strength will the stakeholders accept all the models in the framework. And only if the
shareholders think that they can control the managers' desire to overinvest will they want to implement Froot, Scharfstein an d Stein's
framework. A high debt ratio will normally discipline the manager from spending too much equity. A high debt ratio may however be
dangerous for a firm that wants to use any of Froot, Scharfstein and Stein's risk-increasing models. A combination of debt and
speculations will increase the risk of financial distress and bankruptcy. It is even doubtful that debtholders will accept this. This means

94
  One must here rule out extreme oil prices, like the ones sawn during the oil crises in the beginning of the 1970’s, which will influence
 the world economy and thereby also the demand for air travel.
that the existing shareholders are faced with the choice of the overinvestment problems associated with Froot, Scharfstein an d Stein's
framework and the costs of acquiring external capital. Since it has been shown that the latter costs may not be too large going to th e
capital market may be chosen over risking overinvestments.



                                                              PART III.
The preceding sections have evaluated the underlying as sumptions that Froot, Scharfstein and Stein made: external capital is
expensive, external capital is more expensive than Froot, Scharfstein and Stein's framework and decisionmakers will implement the
framework. The conclusions are that external capital is not expensive for all firms and there are ways to reduce the costs of external
capital, hedging is cheap and decisionmakers may reject the implementation of Froot, Scharfstein and Stein's framework.

The following sections take a closer look at the more operational aspects of Froot, Scharfstein and Stein's framework. Froot, Scharfstein
and Stein assumed that changes in some risky asset/assets causes the business environment to change in such a way that it is optimal
for the firm to change its investments. W hat then makes a firm invest? And is it possible to trade derivatives written on those
investment triggers? The following sections address these questions.


                           8.          WHAT MAKES THE FIRM INVEST ?
In their paper Froot, Scharfstein and Stein gave no list of possible risky assets that will suit their framework. They did however come
up with two specific risks. The price of the firm's output was used as the hedgable risk in a couple of examples and in their section
about multinationals the exchange rate was the risky asset. If their framework is as general as they claim in their introduction, more risks
should fit into the framework.
The risks have to be:
    Marketable, by assumption.
    The cause of the variations in cashflows from assets in place, or correlate almost perfectly with the variations in cashflows from
     assets in place.
 Changes in this/these risky asset/assets must be what cause the firm's investments to change.
8.1      Valuation of Investments.
8.1.1          The Discounted Cashflow Model and Why it May Not Be an Optimal Decision Rule.
An evaluation of the discounted cashflow method is important in this context. Froot, Scharfstein and Stein used the net prese nt value
model (NPV) in their framework. If this model is not an optimal evaluation method, Froot, Scharfstein and Stein's framework may lack
important parts in its basic set-up. The NPV model is a very simple model, perhaps too simple. Froot, Scharfstein and Stein argued that
changes in a risky asset, , will change the net present value of a project. The positive net present value of a proposed investment
project should, according Froot, Scharfstein and Stein , trigger firm investments. If investment decisions are a function of other criteria
than just a positive NPV, changes in the risky asset are not enough to determine if the firm should invest or not.
The problem here is that Froot, Scharfstein and Stein based their framework on an old model, the NPV model, and it is not at all given
that this model gives correct answers in an investment decision.

Modern finance started more or less when Modigliani and Miller published their paper «The cost of capital, corporate finance and the
                         95
theory of investment» in 1958. They claimed that, given their assumptions, it is not possible to create value through the way a project
is financed. Financing- and investment decisions are completely separable. Since then a lot of work has been done in the field of
finance to extend our knowledge about the relationship between financing and investments. However, almost all of the work has
focused on the financing part. According to investment theory, the investment decision for an investor is a tricky thing, involving
portfolio theory, diversification and riskaversion. Very little work has been done to understand the firm's investment process however.
Thousands of textbook pages are written on the subject, but they all tell the same story. The firm should discount the projec t's future
cashflows with the appropriate discount rate to get the projects present value. If the net present value of the project is positive, the
firm should accept it.
Kaplan and Ruback (1995) said that there is however little empirical work that supports this way of valuation. There are several
uncertainties involved in the DCF method.
 An obvious problem with the DCF method is to make good predictions about the future cashflows. This is however not the only
    problem.
 Pindock (1993) said that in addition to the uncertainty concerning the future cash in -flows is the fact that the investment costs are
    also uncertain. These investments are often irreversible.
 To further complicate the matter Pindock claimed that a firm may benefit from engaging in NPV<0 projects. The project may reveal
    information, for example about costs, which have a shadow value. This information will be private for the firm and it can be
    beneficial for the firm to have this information for use in future projects.
 Yet another problem with the NPV model is the difficulty of determining what discount rate to use. One of the hottest topics in
    finance lately has been the debate of whether the capital asset pricing model (CAPM) can be used. The most important question

95
     See section 12.2.1, «Modigliani and Miller.», page 71.
     has been whether or not  has any explanatory power when it comes to finding the correct discount factor. Fa ma and MacBeth
     (1973) did the first test of the CAPM. Their conclusion was that the model worked. However, when Fama and French (1992) later
     tested the model, they found that  did not have any explanatory power. Kothari, Shanken and Sloan (1995) tested Fama and
     French's controversial results and found that they were a result of the set -up of the model. Fama and French (1995 and 1996)
     rejected this and found in a new test that if one controls for firm size and the market -to-book value,  did not explain stock returns.
     Pettengill, Sundaram and Mathur (1995) claimed that Fama and French 's results were due to wrong assumptions in their model. By
     allowing realised market returns to fall below the riskfree interestrate, r f, an inverse relationship between returns and  is predicted
     and 's ability to predict returns is significant. Grundy and Malkiel (1996) found that  gave a good explanation for returns and
     claimed that Fama and French's results again were a result of statistics, rather than facts. This debate is far from over among
     academics within the field of finance. Fama and French (1997) said that the  may not explain expected stock return very well and
     Fama and French's three factor model, which uses E(RM) - RF, size and market-to-book, lacks theoretical founding. There are two
     problems with estimating the cost of capital. The risk loading vary over time and, because of large standard deviations, E(RM) - RF
     range from 0 to 10%. The standard error for industry cost of capital is 3% and it is even more for firms or projects. The latest
     contribution to this discussion is Loughran's (1998) thorough research into Fama an d French' claim that book-to-market and size
     matter. Loughran found that two features drove their results. A January seasonal in the book-to-market value and exceptionally low
     growth in small, young growth firms. In the largest quintile of all firms, which account for 73% of the total market value, book-to-
     market has no explanatory power on the cross section of realised returns during the 1963-95 period.

All this suggests that Froot, Scharfstein and Stein's framework may be based on a model that is less t han optimal, turning their own
framework into a less than optimal model.

8.1.2         Simple Methods.
Kaplan and Ruback (1995) found that using standard values for the industry was at least as good as using the DCF method or
comparison with companies involved in similar transactions.
Schwartz (1997) investigated commodity prices in natural resources investments. He said that it is common to assume that the price is
constant, or to use an industry prediction. The discount rate is usually set to 0,10-0,15.
Because of all the uncertainties associated with the DCF method, Fama and French also said that simpler methods, like the payback
method, might work just as well. This method was actually used durin g the Daimler - Chrysler merger.

8.1.3         Real Options.
Schwartz (1997) said that the market demands a premium that is higher than what it should have been according to the DCF mode l. The
DCF approach usually gives a result with to little sensitivity to commodity p rices. This is of cause of great importance when one wants
to evaluate Froot, Scharfstein and Stein's framework. Schwartz suggested that one should use the real option approach to price
projects. This makes sense since Myers (1977) said that the shareholders claim on the cashflow is a real option. In real option pricing
the cashflow is substituted with a certainty equivalent and discounted with the riskfree interests rate . The real option approach
requires a higher price to invest and thus reflects the real world better than the DCF method. This is important in Froot, Scharfstein and
Stein's framework because this shows that the investment opportunities are more sensitive to the risky asset then previously assumed.
Since the project's profitability is more sensitive to the risky asset than what one has been believed, the profitability of the projects will
be more volatile. For a given commodity price the firm should invest, only a small change in the commodity price will however turn it
                          96
into it a bad investment.
Froot, Scharfstein and Stein's framework is constructed in such a way that it protects the firm's profit maximisation against such
volatility. By using real option models the firm will be armed with a better tool for knowing when to use its derivatives. This means that
the framework should be redesigned in such a way that the firm's profit function is a real option.

8.2           About the Firm's Investments?
As mentioned earlier, financial theory suggests that the very simple role of positive NPV should be used as a determinant for corporate
investments. Froot, Scharfstein and Stein went a bit further and suggested that movemen ts in a risky asset are what actually determine
whether a project has a positive NPV or not. But is the investment decision as simple as Froot, Scharfstein and Stein suggest or is the
investment decision more complex? If there are several other determinants other than changes in the risky asset Froot, Scharfstein and
Stein's framework will not solve the firm's profit optimisation problem.

The following sections explain what kinds of investments the firm does, what the firm wants t o accomplish with its investments and
what triggers the investments.

8.2.1         Strategy.
Rao (1994) said that the firm makes two different types of investments; Strategic and tactical. Strategic investments change the very
characteristics of the firm. These investments include entries into new markets, M&A , joint venture projects, introductions of new
products. Tactical investments change the firm's cashflow, but not the characteristics. Tactical investments include introduction of
similar products, or entry into familiar markets, replacements of equipment, utilising economics of scale or scope. If the firm wants to

96
  The real option approach gives new research results. Kang (1995) examined financing and investment decisions of value maximising
 firms using option pricing on the corporate cash flow in order to find the effects of firm specific factors regarding financial leverage
 and corporate investments. He found that leverage will be lower and corporate investments higher if the firm faces higher tax-shield
 transfer costs, higher capital intensity and smaller insider ownership.
avoid the dead-weight costs of external capital Froot, Scharfstein and Stein's framework must work for both strategic and tactical
investments.

                                              97
Lazonick (1994) refereed to several theories that claimed that American managers had short-term horizons and that they were more
interested in cashflows in the near futures than in the far future. A long-term horizon involves having an innovative investment
strategy, which expands the firm's abilities. A short term horizon means adapting to changes in the environment. These sorts of
investments will reproduce the firms value-creating capabilities. This sort of investments may extract more value from the firm than it
creates, which obviously will not maximise shareholder wealth. Froot, Scharfstein and Stein's framework must work for both short- and
long-term strategies.

8.2.2          Strategic Behaviour.
Rao (1994) said that the firm must create some form of market imperfection in order to earn higher than required returns. The firm must
never reach a situation where it is in a complete-competition situation. The firm can obtain imperfect markets in a number of ways;
 Going into markets that other firms can not access.
 Utilising the fact that labour is usually not very mobile when it builds its plants and locate the plants to locations where the labour
   force has special skills.
 Product differentiation, brandname building.
 Economics of scale.
 Patents.
 Government regulations.
 Control of the distribution network.

The firm's investments must be directed toward creating a market imperfection. Rao further said that the optimal hedging strategy is not
only given from the variables mentioned above. One must also take game theoretical considerations. A firm will want to hedge more
when the competitor hedges less.

8.2.3          Replacement Investments.
Mauer and Ott (1995) said that there are two types of investments, expansions and replacements. The replacement investments are
very important when one wants to evaluate Froot, Scharfstein and Stein 's framework, since they exceed expansion investments by a
wide margin. Timing is an issue in replacement investments. The replacement investment should be viewed as a real option and if the
firm can time the replacement it should invest when the value of the investment exceeds the direct costs and the option value of
waiting. Mauer and Ott found that the optimal time between replacement investments is increasing in the volatility of costs, the
purchasing price of new assets and the corporate tax rate. The optimal time between replacement investments is decreasing in the
systematic risk of costs, the salvage value of the asset and the investment tax credit. The depreciation rate does not move the time
between investments in any particular direction. Uncertainty about the arrival of a technological innovation that would decre ase the
maintenance costs and operation costs results in a significant decrease in replacement investments.
When one think of corporate investments one will usually have expansions in mind, but since replacement investments count for larger
investments, they are of greater importance for Froot, Scharfstein and Stein's framework. If the firm considers using Froot, Scharfstein
and Stein's framework it must be possible to trade derivatives on the investment triggers mentioned above.

8.2.4          Technology.
One thing that may trigger an investment is the arrival of new technology . Grenadier and Weiss (1997) studied different investment
strategies for implementing new technology. There are several things to consider before one introduces new technology. A good
reason for being in the forefront in ones industry when it comes to implementing new technology is that there are several exa mples of
firms who implement one technology and later find it easier to adapt to the new next generation of tec hnology. There are however also
many examples of technological innovations that were never implemented because firms wanted to wait until the technology had
improved further.
Grenadier and Weiss listed five strategies.
 Compulsive: purchase every innovation.
 Leapfrog: Skip early innovations, but purchase the next improvement.
 Buy-and-hold: Buy the first innovation and not upgrade.
 Laggard: Buy the former when a new arrives.
 Bystander.

Two different firms will choose differently because of their different p revious decisions.
All this suggests that even if it is possible to trade derivatives on technological innovations it is not clear if or when th e firm wants to
invest. The firm must be able to hedge with respect to timing.

8.3           What Make Firms Invest?
Hull and White (1994) made a statement that may be devastating for Froot, Scharfstein and Stein's framework if it is true. They said t hat
hedging on the basis that there is only one source of uncertainty would be unrealistic. As a matter of fact, so would two or three risky
assets also be. This has serious implications as to which hedging instruments the firm can use. The firm must either use inst ruments

97
     Porter (1992), among others.
that are sensitive to more than one risky asset or construct a portfolio of instruments that can cope with the different risky assets in a
satisfactory manner.

Another problem with the fact that there may be more than one determinant for investments is that one risky asset may change in a
favourable way while another risky assets may not move, effectively making further investments useless. Business Asia (1995)
suggested that both the strong yen and excess capacity limits further Japanese corporate investments. If the firm use currenc y
derivatives and the yen weakens, the firm can still not invest due to the general over-capacity in Japan.

8.3.1          General Findings.
Pindock (1991) used option theory to describe corporate investments and he emphasises the value of delaying the investment and the
opportunity cost of investing. His list of investment-triggers included the following things.
 Managerial resources.
 Technological knowledge.
 Reputation.
 Market position and possibly
 Scale.
It must be possible to trade derivatives on all these factors if Froot, Scharfstein and Stein's framework is going to be used.

Dixit and Pindock (1994) said that the interests rate has, contrary to common believes, no explanatory power for corporate investments.
                                                                                                           98
This is easy to see if one apply the real option approach described above, Dixit and Pindock said. This model shows that the
following things determine the firm's investment decision.
 Product prices.
 Input costs.
 Exchange rates.
 Taxes and regulations.
If Froot, Scharfstein and Stein's framework is to be useful it should be possible to trade derivatives on these assets. This can of cause
be confirmed right away for the first three factors, without an in -depth analysis of available derivatives. These findings strongly
supports Froot, Scharfstein and Stein.

8.3.2          The Interests rate.
Dixit and Pindock's (1994) argument about the interests rate mentioned above deserves more attention. According to macroeconomic
       99
theory the level of aggregate real investments must equal the supply of new equity. There are both long -term equilibrium and short-
term equilibrium models where the investment demand is a function of the interests rate only.
Brown (1988) said that although the IS-LM model is based on a negative relationship between the interests rate and investments, other
things matter too, such as uncertainty, expectations and time. Brown said that economists have not been able to find a model to explain
or predict corporate investments. He said that although gross private domestic investment / GDP is 15%, it can fluctuate by 30% on a
year-to-year basis. This means that the investment decision must be determined by factors far more complicated than just the interest s
rate.
Dixit and Pindock's (1994) also said that it has also been found that interests rate cuts as a policy to stimulate corporate investments
have a limited effect.

8.3.3          The Exchange Rate.
Bartov and Bodnar (1994) said that it is a widely held view that changes in the exchange rate affect the value of the firm and thus also
its earnings and therefor its investment decisions. In their empirical work they found however no correlation between abnorma l returns
and changes in the value of the USD. They did however find a correlation between abnormal returns and the value of the USD lagged
             100
one quarter. The correlation is significant both economically and statistically. This indicates that the exchange rate is an imp ortant
variable in Froot, Scharfstein and Stein's framework.

He and Ng (1998) studied how the exchange rate exposure affected Japanese multinational corporations. They found that the exchange
rate movements has affected both the cashflow for the firms and the discount rate used to value those firms. The depreciation of the
Yen has had a positive effect on those firms. They found that the exchange rate exposure was positively related to firm size, which
should implicate that larger firms should be more inclined to use Froot, Scharfstein and Stein's framework. He and Ng also found that
firms with low short-term liquidity are less exposed to the exchange rate. This is contrary to what one would expect from Froot,
Scharfstein and Stein's set-up of their framework. A firm with little internal cash should be exposed to changes in the cashflow from
assets in place. This is because the total amount of money, the sum of liquid assets and proceeds from ongoing operations, will vary
relatively more, thereby forcing the firm to changes its investments or the amount of external capital raised. In Froot, Scharfstein and
Stein's model both these things are undesirable.



98
     The real option approach, see section 8.1, «Valuation of Investments.», page 47.
99
     See for example McCafferty (1990).
100
      They interpret this as an evidence that investors do not use all available information.
8.3.4       The Business Cycle.
Solow (1970) came up with a group of stylised facts about business cycles . When evaluating Froot, Scharfstein and Stein's framework,
two of Solow's facts are of special interest.
 Investments move abruptly procyclical and consumption moves smoothly procyclical.
 The fluctuations in GDP are persistent.
This means that investments correlate with the GDP.

Blanchard and Fisher (1989) presented several findings that described how US macroeconomic variables move together. They
described how quarterly leading and lagging variables move when there are changes in the GDP.
 Fixed investments and government spending move much in the quarter where there are GDP innovations and a little in the quarte r
    in advance.
 Inventory investments are a little negative in the following quarter.
 Real wages also move in the same quarter as GDP innovations.
 The nominal interests rate increases the two quarters prior to a GDP innovation and the quarter where the innovation take s place.
    Then it drops sharply the second and third quarter after the innovation.
 The real interests rate on the other hand decreases the two quarters prior to the innovation, is unchanged during the innovation
    and then increases. It finally falls in the second quarter after the innovation.
This last finding may explain why the studies mentioned above found little or no correlation between investments and changes in the
interests rate. The real interestrate may therefor after all explain investment levels.

Mork (1992) presented similar results as Blanchard and Fisher, again for the US economy.
 Investments decline sharply during recessions and increases equally sharp after recessions. Investments move much like
  consumption of durable goods, only more volatile.
 Residential investments are more sensitive to business cycles than non-residential investments. This means that changes in the
  GDP has different consequences for different firms.
 Changes in inventories are extremely volatile. The stock of inventory increases in the beginning of a recession. This can be due to
  either long-term contracts or that the management does not recognise the recession at once and continue to order inputs as before.
  Investments in inventory pick up very fast after a recession.

Hall and Taylor (1993) said that resent research in macroeconomics has shown, maybe not surprisingly, that firms change their prices in
response to changes in demand and supply. They also said that investments are caused by changes in demand, contrary to what
Keynes said. Hall and Taylor further said that business investments and the GDP are responding to the same underlying stimulu s. The
amount of investments is the joint outcome of three things;
 Investment demand by firms.
 Saving supply by customers.
 Investment supply by producers of investment goods.

Ariel and Pohlman (1994) said that business cycles are important for firm's investment decisions because they determine the credit
available. So, the amount of money available may be more important than the actual price of money, the interests rate . If firms use
Froot, Scharfstein and Stein's framework and rely on internal capital to finance projects then the credit available may not be an
important direct determinant for investments. However, customers and competitors may be exposed to interests rate risk or credit
available and thereby change the firm's competitive advantage and the need for investments. This means that a firm that applies Froot ,
Scharfstein and Stein's framework may still be exposed to interests rate risk or credit available to firms.

Yosha (1995) said that since small firms rely more on bank-debt they are more sensitive to monetary economics then larger firms are.
This should indicate that the business cycle is even more important for small firms that for large firms. This again should suggest that
small firms should be more eager to use Froot, Scharfstein and Stein's framework than large firms should.

Chan, Karceski and Lakonishok (1998) studied what forces determine variations in a stock return. This is important be cause ideally a
stock's return should be correlated with the firm's investment opportunities . They found however that macroeconomic variables
perform poorly when it comes to explaining returns. This means that variations in macroeconomic variables may not be good
candidates as investment triggers , or as assets that should be hedged.

8.3.5       Leading Indicators.
Froot, Scharfstein and Stein did not say anything about the length of the periods in their mode l. It does however seem fair to assume
that in some cases the investment decision should happen before the business environment actually changes, i.e. before the ri sky
asset changes. This will give the firm time to invest and to make other adjustments that take time. When the risky asset do change and
the business environment with it, the firm already has the ability to meet the changed demand.
This may not be an easy task for many of the risky assets that seem to determine the firm's investment decision. But from the findings
mentioned above it seems that the business cycle is an important variable in the firm's investment decision. The Index of Leading
Economic Indicators (Mork, 1992) is presented by the Bureau of Economic Analysis and is an average of 11 different indicators. The
indicators have previously signalled future trends in the economy. These indicators are;
 The length of the average workweek.
 Average weekly initial claims for state unemployment insurance.
   Manufacturer's new orders.
   Vendor performance.
   New plant and equipment orders.
   New building permits.
   Change in manufacturers' unfilled orders for durable goods.
   Changes in sensitive material prices.
   Stock prices.
   Money supply.
   Consumer expectation index.

If the firm can trade derivatives on thes e factors, instead of changes in the actual business cycle, it will be ready to produce goods as
soon as the demand develops.

8.4         Comments.
Now that it has been established what kinds of investments the firm makes (strategic/tactical, long/short term and
expansion/replacement), what the firm wants to accomplish (market imperfections ) and what triggers the actual investments one can
move on to study the derivatives. Only after the properties of available derivatives have been found s omething meaningful can be said
about whether there are marketable derivatives with the right properties to fit the firm's investment decisions.


               9.          WHAT INSTRUMENTS SHOULD BE USED?
9.1         How Derivatives Are Used.
Trade in financial derivatives started on May 16, 1972 in Chicago with the trade of exchange rate futures. The derivative market has
increased 20 fold from 1986 to 1994, when it was $20.000 billion. Derivatives separate risk and ownership. The owner of a sto ck forward
is exposed to changes in the stock value, but have no voting right. Derivatives make managers and investors desegregate risks , bear
those they can manage and transfer those they are unwilling to hold.
The use of derivatives can be classified into four categories:
 Derivatives can be substitutes for other investment assets, leaving risk and return unchanged.
 Hedging, reducing risk and return.
 Speculation, increasing risk and return.
 Financial engineering, redistribution of the cashflow.

These are some figures that describe the use of derivatives in the US in 1994.
 80% of all private firms considered derivatives important for risk management .
 83% of all financial institutions considered derivatives important for risk management.
 US commercial banks held $15 trillions worth in derivatives.
 The top 25 banks held 97% of the activity in derivatives trading.
Firms use a wide range of derivatives. The following is a list of the percentage of firms that used the different derivatives in the US in
1995.
          87% use interests rate swaps.                      64% use currency swap.
          78% use currency forwards.                         40% use interests rate options.
          31% use currency options .
... and for the UK.
          93% use interests rate derivatives.                84% use currency derivatives.
Despite this widely use of derivatives, Reynolds (1995) said that few firms believe that the derivative trading add value!

9.2         Futures.
9.2.1       About Futures.
Futures are only single period instruments. They give fixed protection, both upside and downside, and can be viewed as a seri es of
one-day forward contracts. Futures are the oldest of the derivatives. Modern futures date back to the 1860s. Futures have a time to
maturity of up to four years. When one goes into a futures contract the entire price for the futures contract is not paid for . The buyer
only pays a margin. This margin, the initial margin, will typically be only 1/20 of the value of the futures contract. Each party in the
agreement has an account that is settled daily. If the price of the futures contract has changed from the day before, one par ty pays a
variation margin to the other. Whenever the account dips below a certain level a maintenance margin has to be paid. All things being
equal, the price of futures contracts increase with the time to maturity. Below is a list of when the different futures contr acts were
introduced on modern exchanges.
 Foreign exchange rate futures in 1972.
 Interests rate futures in 1975.
 Commodity futures in 1978.

98% of all futures contracts are closed before maturity.

There are futures contracts written on the following assets.
 Agricultural commodities.
 Industrial commodities.
 Indexes, such as the S&P 500.
 Financial assets.
The direct costs for futures are the margins and the commissions.

9.2.2          Advantages of Futures.
One particular property is of especial importance for Froot, Scharfstein and Stein's framework. This is the fact that the initial margin is
only about 5% of the value of the contract. This is important in this context bec ause Froot, Scharfstein and Stein's framework implicitly
assumes that there are rationing on the internal capital. The less the firm has to pay up front for the contract, be better.
Futures contracts are traded at exchanges, which makes it easy to find a counterpart. The chance of finding a counterpart is further
increased by the fact that different exchanges specialise in different assets. The futures contracts are standardised, which makes
trading even easier. There also exist only a limited number of contracts, both when it comes to the number of underlying assets and t he
number of contracts on each underlying asset. This ensures that the liquidity of the contracts that are traded is good. This way it is
easy to get out of a deal by offsetting the first one. This is good for Froot, Scharfstein and Stein framework. All trading is between the
counterparts and the clearinghouse, not bilateral between the counterparts directly. This eliminates the defa ult risk.
Another feature about futures is the price limits. There is an upper limit to how much the price is allowed to move in one day. This way
                                                                       101
one reduces the risk of crashes that are not based on fundamentals. Everything that reduces the risk associated with the derivatives
is good for the robustness of Froot, Scharfstein and Stein's framework.
All this means that the main advantage of futures , with respect to the requirements that is put forward above, is that they are liquid and
inexpensive.

9.2.3          Disadvantages of Futures.
The limited number of available futures contracts is however a problem too. There may not be futures contracts on the firm's
investment triggers, such as inputs and finished goods. This problem arises if the risk that the firm needs to hedge is not traded. In that
                                           102
case the firm has to use cross hedging. Cross-hedging carries with it additional basis risk. This means that a cross hedge is a less
then perfect substitute for a single instrument.
Considering the length of business cycles and exchange rate cycles, futures contracts have maturity that do not span over a typical
cycle. The lifetime of futures contracts is too short for some investment triggers . This means that for some risks, the firm has to roll
                                                      103
over its hedging investments. As mentioned earlier this is undesired.
Another disadvantage it the fact that futures contracts have short times to maturity. The most liquid contracts run only for one year.
Froot, Scharfstein and Stein argued that one negative thing with futures is the potentially large maintenance margin payments that the
firm may face. This is however not as bad as Froot, Scharfstein and Stein make it sound. Consider for example a firm that fit s the simple
model. The hedge should be constructed in such a way that the derivatives give the firm a cash inflow when the income from assets in
place is low. That way the firm can continue to make its fixed investments without acquiring external capital. The flipside of the coin is
that such a contract requires that the firm pays maintenance margins if the risky asset moves in the opposite direction. This is however
a situation when the firm has much cash from assets in palace, perhaps more than is required for the nonstochastic investment . Volatile
risky assets can however turn this into a problem.
                                        104
As mentioned earlier, Hull and White        said that there would be several things that together causes the firm's revenue to vary. This
means that the firm has to use several futures contracts in order to hedge the risk. Each of these futures carry the risk of not being a
perfect hedge, with the implications this brings about.

The limited number of underlying assets and the short time to maturity is a problem with futures contracts.

Considering the pros and cons of futures , they seem to be a less than perfect instrument for Froot, Scharfstein and Stein's framework.

9.3           Long-term Contracts.
9.3.1          About Long-term Contracts.
Even though long-term contracts are not derivatives they fit very nicely into Froot, Scharfstein and Stein's framework and they are
therefor given some space here. A long-term contract covers the terms of a series of repeated transactions.
Using long-term contracts is very similar to going into joint ventures.

9.3.2          Advantages of Long-term Contracts.
The main advantage of these contracts in this context is that the firm can fix the cashflows without even having to think about the
sources of the fluctuations. This is especially important if there are several risky assets that determines the level of the cashflows. This


101
    One of the things that made the October 1987 crash in New York so bad was that the computers had been programmed to sell! sell!
    sell! if the prices moved below a certain level. DeGrauwe (1996) demonstrat ed how bubbles may move the market away from the
    fundamental values, even if all the actors are acting rational.
102
    Alternatively the firm can use forward contracts, which are OTC contracts.
103
      See section 3.5.1.1, «A Perfect Hedge.», page 13.
104
      See section 8, «What Makes the Firm Invest?», page 47.
will save time and money for the CFE's staff. The firm can fix the unit price of inputs, thereby fixing its costs a nd fix the unit price of
outputs in order to fix the revenues. The firm's profits will thus be a function of the firms operations and organisation. Th ese things are
however the firm's core competence and is thus a «good» risk.

9.3.3          Disadvantages of Long-term Contracts.
The most important disadvantage of long-term contracts is that it is difficult, if not impossible, to get out of the contracts without
violating laws or contract terms.

Long-term contracts fit Froot, Scharfstein and Stein's framework very nicely if the firm and its business partners do not consider the
last issue relevant. This is especially true if one can not trade derivatives on the risky assets.


9.4           Forwards.
9.4.1          About Forwards.
Forwards are agreements between two counterparts. All forward deals are over-the-counter, OTC, contracts. This means that they are
                          105
not traded on exchanges.
The trades used to be bilateral with the investment bank acting only as an intermediator. Forwards could be a deal between a supplier
and a customer. The two could agree on a fixed price that would be paid for deliveries in the future. These forward contracts are similar
to the long-term contracts mentioned above except that forward contracts cover only one delivery in the future.
Recently the banks have started to write the contracts with the firms or investors independently and only match groups of contracts
not the individual contracts. The use of these forward contracts will have much in common with the use of futures contracts. While
long-term contracts can only be used to fix the proceeds from the firm's assets, futures and forwards can be used to alter the entire
distribution of the cashflow.

Although forwards are supposed to be held until maturity they are often settled in cash prior to maturity.

The firm can use forward contracts to hedge or to speculate against assets that the firm does not have.
Only a few, large dealers can access the forward market directly, the rest have to use these large dealers as brokers. This will increase
transaction costs.

9.4.2          Advantages of Forwards.
Forwards are linear hedging instruments like futures . The difference is that forwards are OTC instruments. This is an important
advantage. Since a wide range of factors influence the investment decision there is a much g reater chance that the firm will be able to
hedge the risks it is facing if it uses OTC contracts than if it is limited to the contracts traded on future exchanges.
Since the contracts are OTC the maturity can be as long as necessary, at least in theory. Very long contracts are rare, swaps should be
considered instead if very long hedging is needed.
Forward contracts are off balance sheet instruments. This means that there are no payments until settlement. This is of cause of major
importance for Froot, Scharfstein and Stein's framework. The firm is low on internal capital by assumption. The fact that no payments
are required neither initially nor during the life of the contract means that the firm does not run a risk when it enters into the contract.

9.4.3          Disadvantages of Forwards.
Forward contracts are mostly for large actors in the market. This is obviously a negative thing in the framework of Froot, Scharfstein
and Stein since many firms will be excluded. Because of the credit risks, some institutions will require collateral for OTC c ontracts.
Some institutions will not even trade unless the participants are major entities. This is a serious problem for Froot, Scharfstein and
                                                                                                                      106
Stein's framework, since there are reasons to believe that there are economics of scale in acquiring external capital. This means that
small firms should be keener to use Froot, Scharfstein and Stein's framework than larger firms should. This is not a problem only for
forward contracts, but for all OTC contracts.

Forwards seems to be a better choice than futures because they are OTC, have longer time to maturity and are even cheaper than
futures up front. The only problem is the fact that small actors may not be able to use them.

9.4.4          Special Types of Forwards.

9.4.4.1          Forwards on Commodities.
 In Froot, Scharfstein and Stein's framework there is a marketable risky asset that makes the revenues from assets in place fluctuate.
Since forwards are agreements between two counterparts, the buyer and the seller, it is not necessary to limit the framework to
situations where the risk is a given risky asset, as Froot, Scharfstein and Stein did. The firm may not even care what causes the price of
its products to fluctuate. The contracts can be written on the firm's products directly. The forward contracts can be between the firm
and its customers, although a bank will stand as an intermediator. The firm can sell its products using one or more forward contracts

105
      There are however exchange traded forward contracts written on electricity.
106
      See section 5.6, «Transaction Costs.», page 41.
depending on the number of customers. Forward contracts on a specific risky asset are really not needed. T he firm can control its costs
the same way by going into forward contracts with its suppliers. It is of no interest what makes the spot prices of its produ cts
fluctuate. It may even very well be more than one risky asset. These kinds of forward contracts will be similar to FRA contracts, only
the rate decided on is the price of a commodity and the maturity is longer. This means that Froot, Scharfstein and Stein make s things
more complicated than they need be.
These contracts are similar to the long-term contracts mentioned above.

9.4.4.2          Forwards with Optionlike Characteristics.
Forward contracts give both upside and downside protection. While protection in one direction is good, protection in the opposite
direction is usually not desired. A break forward contract is a contract that is breakable if the risky asset reaches a certain value. Such a
forward has a payoff like an option. With a range forward the owner has the right to break both at the upside and at the downside. This
                                                                                    107
kind of instrument can be used in the No Protection version of the Simple model.

9.5            FRAs, IRGs and SAFEs.
Forward contracts on interests rates are called forward rate agreements (FRA). These are contracts on the interestrate itself, as
opposed to futures and ordinary forwards contracts which are written on securities that pay an interest, such as bonds. There are for
FRAs 3, 6, 9, 12 and 24 months and the market is liquid for deals up to one year. This is only a single period instrument. In a FRA, there
is no obligation to borrow at the specified rate. The parties only pay the difference between the forwa rd interests rate and spot rate on
the notional of the contract. Most FRA contracts are denoted in the LIBOR, since most of the trade occurs in London. However, the
deal can be in another currency than the British pound. With a FRA on e can buy a contract of getting for example a 6-month interests
                  108
rate in 3 months.
IRGs are interests rate guarantees. They are options on an interests rate. The only thing that separates them from a FRA is that they
only give protection in one direction.
Forward contracts on foreign exchange rates are called synthetic agreement for forward exchange. While FRAs are in absolute terms,
SAFEs are differentials, if the difference between two currencies is different from the contracted difference, there will be a payment one
or the other way.
The major advantage with these instruments is that there are no upfront payments. Their short times to maturity make them however
less suited for Froot, Scharfstein and Stein's framework than other instruments.

9.6            Options.
9.6.1          About Options.
Options give the holder the right, but not the obligation, to buy (put option) or sell (call option) an underlying ass et at a specific price
(the strike) at any time during the life of the option (American option) or at the option's maturity (European option) 109. The buyer pays a
premium to the issuer. Depending on whether one has sold or bought the option and whether it is a put or a call one can win or loose
both when the market value of the underlying asset increases or decreases.

Below are examples of underlying assets that options are written on:
        Common shares.
                         110                Bonds.                               Futures.
                                                                                              111

        Currencies.                       Bills.                               Stock indexes.
        Commodities.                      FRAs.                                Other options.
        Caps.                             Floors.                              Swaps
                                                                                            112

         Swap spreads.                          Yield curves.
Froot, Scharfstein and Stein's framework is based on state contingency where variations in the risky asset decide the firm's investment
decisions. Since it is highly unlikely that the firm will know in advance how the value of the risky assets will move, the firm has two
options when it comes to choosing what option contracts to use. If the secondary market for the options are liquid European options,
or Bermuda options, can be used. This way the firm can easily get out of the contracts. If the risky assets have mov ed in such a way
that the firm wants to invest the option contracts written on the risky assets will be valuable. The proceeds from the sold o ptions
should be used to finance the investments. If on the other hand the secondary marked is illiquid, American options must be used. Few
options are exercised, they are either out-of-the-money and thereby worthless or they are bought back. this will however happen close
to maturity. It is still important for the firm to be able to get out of the contract whenever it likes.

Options are traded both at exchanges and OTC; this goes for interests rate , currency, equity, index and commodity options . The
availability of exchange traded options is beneficial for Froot, Scharfstein and Stein's framework because it is easy and convenient for
the firms, as well as fairly cheap. The fact that options can be traded OTC is good because it enables firms to create option contract

107
      See section 4.4.1.2, «No Protection in the Simple Model, Out of Boundary Exposure.», page 17.
108
      This is called a 3 X 6 forward.
109
      Semi-American options, or Bermuda options, can be exercised on specific dates during the lifetime of the option.
110
    Most stock-options have a maturity of some months, which does not make them very suitable in Froot, Scharfstein and S tein’s
    framework. There are options with longer time to maturity written on shares in some large firms and for indexes.
111
    There are options on agricultural futures, industrial futures and financial futures.
112
      These are called swaptions or swoptions.
that fit their needs exactly. The drawback with OTC options is that they are more expensive and it may be difficult to sell them in a
secondary market.

9.6.2          Advantages of Options.
For exchange-traded options the exchange decides the underlying asset, the strike, American or European and t he time to maturity.
This is both good and bad for the firm in Froot, Scharfstein and Stein's framework. The good news is the liquidity and thus l owered
prices. This is good because the framework implicitly says that the firm has limited amounts of interna l capital, so cheaper options are
an advantage.
The benefits that options provide to Froot, Scharfstein and Stein's framework can be increased in two different ways. The firm can
combine different options into specialised pay-off profiles or the firm can use exotic options .

9.6.3          Disadvantages of Options.
If the exchange traded options do not fit the firms needs the firm has to use OTC options . This will be particular likely if the risky asset
moves in long cycles since exchange traded options have time to maturity of only three to 24 months. The main drawback of OTC
options is that they are more expensive than exchange traded options.
All things being equal options fall in value as time goes. This is not a good thing in Froot, Scharfstein and Stein's framework because
the firm may have to buy more options as time goes by to have the same protection.

9.6.4          Special types of options.
There are exchange-traded options on a lot of assets and OTC options can be written on any asset, so there is really no reason for
going into the different kinds of options, like forex options, stock options or options on bond futures. Some types of options differs
from the rest however and because of this they deserve special attention.

9.6.4.1          Macro Options.
                                                   113
A macro option is a means to hedge quantity.          Other instruments hedge the price of the assets but with macro options one can
hedge quantity. The underlying asset «price» that determine the value of the option is a macro variable. These can be single -period
                                                  114
contracts, or multiperiod caps, floors or collars.

9.6.4.2          Exotic Options.
There are a number of variations in the option contract terms. Playing around with these variations can further increase the benefits of
hedging.
Exotic options are options that deviate from standard call and put options. There has been a growing interest for exotic options lately.
Today they account for 5-10% of all derivatives traded and they are the fastest growing type of derivatives. There are several types of
exotic options.

9.6.4.3          Digital Options
Digital options pay a fixed amount if the option is in the money and nothing otherwise.




Figure 16 Digital option.
Froot, Scharfstein and Stein assumed that investment amounts are continuos variables. In reality this will often not be the case as one
will not build 1.37 oilrig. If it is obvious that investments will be done incrementally, digital options can be used. The firm may not be
interested in small changes in the risky asset. These small changes may not be large enough to change the competitive advanta ge of
the firm, which means that the firm is not interested in exercising the options and start investing in new or larger plants, or to acquire
other firms. In addition, such minor changes in the risky asset may also be just a result of the risky asset's volatility . A digital option
with a high strike will be the answer to these problems. Such an option will not be in-the-money unless the risky asset changes
substantially and when it is in the money it will yield enough money for investments right away. In addition, these options w ill be
preferred to ordinary options because they are cheaper.

9.6.4.4          Barrier Options.
Barrier options are useful in Froot, Scharfstein and Stein's framework for the same reasons as digital options were; they are cheaper and
only valuable if the risky asset moves a lot. For barrier options the value of the risky asset has to be above or below a barrier before the


113
      There has only been one attempt to make macro futures the NYSE.
114
      About multiperiod options , see section 9.7, «Caps and Floors.», page 60.
option is activated. «In» options must exceed the barrier before they can be exercised, while «out» options can only be exerc ised if the
value of the underlying asset has not hit the barrier. An up-and-in barrier call option has a barrier that is in-the-money. If the value of
the risky asset at maturity is out-of-the-money, the barrier option is worthless, just like an ordinary call option. If the value of the ris ky
asset at maturity is between the strike and the barrier, it will still be worthless. The value of the risky asset has to be a bove the barrier
before the option can be exercised. If it is above the barrier, the payment is exactly equal to an ordinary opt ion.



                    Strike            Barrier




Figure 17 Up-and-in barrier call option.
In addition to up-and-in options, also known as knock-ins, there are up-and-out options. They can not be exercised if the risky ass et
exceeds the barrier. They are not of any interest here, neither are down -and-out barrier options. Down-and-in options have exactly the
same features as up-and-in options and will be used in Froot, Scharfstein and Stein's framework if the risky asset is such that the firm's
competitive advantage increases if the risky asset falls, like for an airline operator if the oil-price falls.

9.6.4.5          Non-linear Options.
As stated above, the firm may not want to invest unless the risky asset has move d quite a bit, however, once it chooses to invest, the
firm needs a lot of cash. One way to solve this problem is to use options where the payoff is a non-linear function of the value of the
risky asset. A deeply in-the-money power of two option will not be exercised unless the risky asset moves a lot and it will then pay off
a great deal. A deeply in-the-money option will in addition be quite cheap. non -linear options have however lost their popularity
       115
lately.

9.6.4.6          Pay-Later Options.
The whole idea behind Froot, Scharfstein and Stein's framework is the assumption that internal capital is cheaper than external capital.
This means that any cashflow out of the firm has a negative effect, including the cashflow to pay for the derivatives. A solution to this
problem is to use pay-later options, or contingent options. These are options where the buyer only pays a premium if the option is in -
the-money at maturity. These options have to be exercised if they are in-the-money and if the option is only just in-the-money the
option premium will exceed the value of the option. This is how the option issuer earns a profit from such deals. Because of this the
firm runs the risk of actually loosing money instead of getting money in a situation where it desperately needs money in order to exploit
the current market situation. On the other hand, it may be the only chance a firm has if it wants to use Froot, Scharfstein a nd Stein's
framework and it has no cash.
All options traded on the LIFFE (London International Financial Futures Exchange) are pay-later options.




Figure 18 Pay later option.


9.6.4.7          Chooser Options.
A chooser option gives the holder the right to choose at a predetermined date whether the option is to be a call or a put. The
                                                                                              116
predetermined date can be the maturity date. The option is like a spread optio n strategy. These options are used to hedge large
market moves when the trader anticipates large movements and is unsure of the direction of the market but will be better info rmed at a
later date. In a Black-Scholes set-up the fact that the option is a chooser will not add value. Later option pricing models have however
showed that the added degree of freedom has a value, naturally.

9.6.4.8          Asian Options.
Asian options are path-dependent options. That means that the payoff is dependent of the path of the price of the risky asset over a
period of time. These options are very popular, which is a good thing in the Froot, Scharfstein and Stein's framework because it will
make the availability good and the price low.
Two kinds of Asian options are popular.


115
      This is partly due to Bankerstrust’s disaster with a LIBOR squared swap.
116
      See Figure 20, «Bullspread. Out of boundary exposure in the simple model, and the reversed simple model.», page 59.
The average rate, or average price, option (ARO) pays out the difference between the average value of the risky asset and the strike,
given it is positive.
The average strike option (ASO) on the other hand pays out the difference between the value of the risky asset at maturity and the
average value of the risky asset during the life of the option, given it is positive. For these options it is the strike that is dependent of
the path of the value of the risky asset.

The advantage of Asian options in Froot, Scharfstein and Stein's framework is that the average value of the risky asset is probably a
far better investment determinant than the spot value. This means that the payoff from the options coincidence better with the
investment decision.

9.6.4.9          Options That Are Dependent of More Than One Asset.
Another form of path-dependent options are options that are a function of more than one asset. Hull and White (1994) said that it is
unrealistic to think that a firm can manage its risk by hedging only one risky asset. These options can be very useful for a firm that
considers using Froot, Scharfstein and Stein's framework. The payoff from these options can be a function of the value of the
underlying assets or of the spread between the values of the assets. This is an option strategy that obviously can be combine d with
any of the strategies mentioned above. One can for instance construct a barrier option wh ere the strike is dependent of the spread
between the LIBOR and the German repo rate in the power of 1,5.
A special kind of options on more that one asset is the quanto options. For these options the strike can be an index or the value of a n
asset, the payoff is however done in a different currency. A Nikkei options denoted in USD is an example of a quanto option.

There are also other types of exotic options , for example one-tough options, look-back options, delayed options and shouter options,
but these are not useful in Froot, Scharfstein and Stein's framework and are therefor not annotated here.

9.6.4.10          Compound Options.
Compound options are options on options. These are useful if a firm bids for a contract. If the firm gets the contract it may want to
protect it selves from movements in financial assets from the time it gets the contract until it receives the payments. The f irm can use
compound options for this kind of protection. If the firm does not get the contract it is not exposed to the risky asset and it will not
exercise the option to buy options. The risky asset in the first option may therefor be the bid for the contract. The underly ing options
will be written on currencies, commodity prices or whatever risks the firm faces if it gets the contract.

9.6.5          Different Types of Option Strategies.
Spreads are combinations of two or more options . The following things can vary.
 Long or short position.
 High or low strike.
 Long or short time to maturity.
 Calls or puts.
 The number of options of each type.
Froot, Scharfstein and Stein limited their adjustment of the hedge to the hedgeratio. However, different firms or different industries will
have different needs. It is doubtful that all firms can find one of Froot, Scharfstein and Stein's models that will fit in all possible states
of the economy or of the business environment.
With the unlimited number of pay-off profiles one can make with options , these should be used to tailor the protection to the firm's
needs. The following are examples of how the firm can use option strategies to increase the benefits of Froot, Scharfstein an d Stein's
framework. One should be aware of the fact that these strategies involves only regular options, there is no need for exotic options . This
                                                                                                                                            117
section covers single period options, however, the same strategies can be applied for the multiperiod options called caps and floors.


                                                                 Cashflow from assets in place.
                                  Bought put




Figure 19 Single option. Adverse protection in the simple model 118, the no-hedging model 119 and the uncorrelated model 120.




117
      See section 9.7, «Caps and Floors.», page 60.
118
      See section 4.4.1.1, «Adverse Protection in the Simple Model.», page 17.
119
      See section 4.4.2.2, «Adverse Protection in the No-Hedging Model.», page 19.
                                                                                 Bullspread

                                                                            Cashflow from assets in place.




Figure 20 Bullspread. Out of boundary exposure in the simple model 121, 122 and the reversed simple model 123.
In a bullspread the firm in long in one option (a call in the figure above) with a high strike and short in an opposite option (put) with a
low strike. The opposite spread is a bearspread. It is possible to get the same pay -off profil with caps and floors, the combination is
then called a corridor, and with swaps . Such swaps are called fixed floor-ceiling swaps.



                                  Cashflow from assets in place.
                                                                                 Bullspread




Figure 21 Speculation model 124, pure form.


                                           Cashflow from assets in place.
                                                                                 Straddle




Figure 22 Straddle. Speculation model with adverse protection.
In a straddle the firm is long in one option with a high strike (a call in the figure above) and also long in an opposite option (put) but
the latter has a lower strike.

                                    125
9.6.6          Synthetic futures.
A synthetic future is a combination of a call and a put. It is a spread with the same strike for both the call and the put. The following
relationship exists.

                                          Payoff on a call - payoff on a put = payoff on a long future.

This can be valuable in Froot, Scharfstein and Stein's framework for several reasons. First, it increases the number of possible assets
for linear hedging. One can create futures contracts on assets that no future exchanges offer contracts on. One can also create futures
that mature at different times than the exchange traded futures contracts. Many options have longer time to maturity that futures,
making them better suited for Froot, Scharfstein and Stein's hedging than futures contracts. Second, synthetic futures eliminate the
problem with futures that Froot, Scharfstein and Stein describe namely the potentially large margin payments. The payment up front is
however larger for a synthetic futures than for a normal futures. Third, since the firm can make a linear hedge out of heavily traded


120
      See section 4.4.3, «The Uncorrelated Model.», page 19.
121
      Cashflows from assets in place are redistributed through a long -term contract
122
      See section 4.4.1.2, «No Protection in the Simple Model, Out of Boundary Exposure.», page 17.
123
      See section 4.4.4, «The Reversed Simple Model.», page 20.
124
      See section 4.4.2.3, «The Speculation Model.», page 19.
125
      Some authors call this a synthetic forward or even a synthetic stock.
options, the default risk that Froot, Scharfstein and Stein mention as a problem for forwards is also reduced. The sum of the se pros and
cons make synthetic futures a good alternative for conventional futures/forwards.

9.7            Caps and Floors.
Options are single period instruments. There are however multiperiod equivalents. These are called caps , multiperiod calls, and floors,
multiperiod puts. Caps and floors can be combines in the same way as ordinary options can. This means that it is possible to create
any pay-off profile in a multiperiod setting.

9.7.1          Caps.
A cap is like a strip of call option contracts . The cap has several «maturity dates» at which the value of the risky asset is compared to
the value in the cap contract. If the value of the risky asset is above the strike value in the contract the issuer will pay the cap holder. A
notional is agreed upon in the contract. If the cap is written on the LIBOR with a notional of 10 million and a strike at 5% and the
present LIBOR is 5,4%, the issuer must pay the holder 0,4 * 10 million. If the LIBOR is less than 5% the cap will be out of t he money
and the firm will enjoy the low interests rates. There will be no transfer of money in that period.


High interests rate




Low interests rate


Figure 23 One-sided protection by a cap.
The protection periods can be 3, 6, 9, or 12 months. These instruments give one -sided protection for up to 10 years, just like the floors
below. This is of great importance for Froot, Scharfstein and Stein's framework, since one of the main arguments against othe r
instruments has been their short life. The fact that the protection periods can be as short as three months is a positive property of the
caps/floors because the firm gets cash immediately as soon as the risky asset's value changes and thereby also changes the busine ss
environment and the investment opportunities for the firm.

9.7.2          Floor.
A floor is the opposite of a cap; it protects against low interests rates. A floor is like a strip of interests rate put options. Floors are not
very actively traded.


9.8            Swaps.
9.8.1          About Swaps.
                                                                                    126                      127                              128
Since Froot, Scharfstein and Stein's framework is really floating -for-fixed          , fixed-for-floating         or floating-for-floating         hedges,
                                                                                          129
swaps are intuitively interesting instruments. Swaps are like a series of forward s.
Swaps do not necessarily need to be bilateral agreements, as they where in the early years of swap trading. Banks handle their swap
deals as a basket of agreements that together cancel out, even if all the deals have different times to maturity, diffe rent notional
principles etc. The banks will them selves take the counterpart position. Swaps are multiperiod instruments.
A plain vanilla swap has the following properties.
 A fixed cashflow is swapped with a floating one.
 The contracts lasts for 1, 2, 3, 4, 5, 7, or 10 years.
 The notional principle is constant.
The following is a list of different swaps and some of their properties.
 Money market swaps.              Up to three years.
 Term swaps.                      Three years or more.
 Spot-start swap.                 Starts two days after the agreement.
 Delayed-start swap.              Starts up to one year after the agreement.
 Forward swap.                    Starts more than one year after the agreement.
 Option on swap.                  Counterpart buys option on term of the swap.
 Swaption.                                                                    130
                                  Counterpart has option to change the swap.


126
      The simple and the uncorrelated models.
127
      The reversed simple model.
128
      The speculating model. This is actually floating-for-even-more-floating.
129
      Just like futures, only that the «futures forwards» lasts for only one day.
130
      The word «swaption» is more often used on options on swaps.
Buyout swap.                    Swap is closed and settled at current price.
Fixed-for-floating swap.
Basis swap.                     Both legs are floating.
Off-market swap.                The rate is not equal to the market rate.
Zero-coupon swap.               The fixed rate is zero.
Amortising swap.                The notional principal declines with time.
Uneven cashflows.               Notional principal takes on any pattern.
Participation swap.             The floating payments will not pay 1-for-1 according to the floating rate, but for example 1-for-0.5. The
                                floating payments are thus more or less sensitive to the floating rate.
Callable swap.                  Equals a fixed-for-floating swap and a put option on a swap. A way to get out of a swap.
Putable swap.                   Floating-for-fixed swap and a call on a swap.
Reversible swap.                The counterparts can change sides.
Extendible swap.                The option to extend the swap.
Cancellable swap.
Fixed equity swap.              Fixed rate payments are exchanged for dividends + appreciation/depreciation of an index.
Floating equity swap.           Floating rate payments are exchanged for dividends + appreciation/depreciation of an index.
Two-way equity swap.            One stock appreciation/depreciation is exchanged for another.
Yield curve swap.               Both are floating rates in the same currency, one in the short-term and one in the long.
Resetable swap.                 The coupon can be reset.
Step-up swap.                   The coupon is initial set below the market rate, but increases so that it ends up above the market rate.
Table 13 Different kinds of swaps.
This huge variety of possible terms makes it easier for the firm to find a contract that fits its needs. The market has grown to $1 trillion.
The creditrisk for swaps is larger than for futures contracts, but less than for forwards. In swap deals the principle is not at risk. The
only payments are the difference in the rates. This means that the only thing at risk is that the firm may lose the gain from the deal.
Although most swap deals are made through investment banks, more and more commercial banks trade in swaps. This increases the
availability and the increased competition may reduce the costs, both of which is good for the Froot, Scharfstein and Stein f ramework.
Swaps are more suited than forward contracts to hedge long-term foreign denoted debt. The reason is that the swaps can be
customised to reduce the basis risk.
The tendency is toward shorter maturity for swaps and that caps, floors and collars are taking over some of the market.
The direct costs of swaps are the broker's fee, since swaps are off balance sheet contracts .

9.8.2         Advantages of Swaps.
Such a broad range of possible contracts as the one listed above makes it easier for a firm to find a deal that fits its needs in Froot,
Scharfstein and Stein's framework.
There is no exchange of principals in swap trading. This makes them especially useful in this framework, since the firm does not have to
tie up unnecessary money in the hedging (like for futures margins).
Swaps are multiperiod.

9.8.3         Disadvantages of Swaps.
The size of the notional of a swap is anything above $10 mill. This may be bad news for Froot, Scharfstein and Stein's framework, since
small firms may not be able to go into such huge deals.

9.8.4         Different Types of Swaps.
Interests rate swaps are exchanges of cashflows in the same currency, but with different bases. The most common interests rate swaps
are fixed-for-floating, which is what a firm exposed to interests rate risk wants in Froot, Scharfstein and Stein framework. That interests
rate swaps are the most heavily traded swaps.
There are commodity swaps on precious metals, energy and physicals. However, since swaps are OTC instruments they can be used
by the firm almost like forward contracts. If the firm wants to get a fixed price for its products it can write forward contracts with its
            131
customers.       Another way to do this is to use a swap contract. The firm can simply swap a certain amount of goods with a certain
amount of cash. The advantage is that a swap can be customised in much greater detail.
Cross-currency swaps are exchanges of cashflows in different currencies, with the same or different bases..

9.8.4.1         Swaps with options like characteristics.
It is possible to construct swaps with the same pay-off profile as the option strategies mentioned earlier. Since swaps are multiperiod -
and OTC instruments they are a good alternative and it increases the CFOs' degrees of freedom when it comes to constructing a risk
management strategy.




131
      About forwards on commodities , see section 9.4.4.1, «Forwards on Commodities.», page 55.
9.8.4.2          Macro Swaps.
Just as there are options on macro variables there are swaps on macro variables. Other instruments hedge price, but these hedge
quantity. This is done by tying the floating leg to a macro variable. For interests rate swaps, the floating leg is tie to the interests rate
and for a cross-currency rate swap the floating leg is tied to a floating exchange rate. So the only way in which a macro swap differs
from the other types of swaps are in the choice of what should determine the value of the floating leg. Macro futures have not been a
success.

9.9           Conclusion.
9.9.1          Which Instruments Are Suited.
It seems that futures contracts are not very well suited for Froot, Scharfstein and Stein's framework. The main reasons are the time to
maturity, the margin payments and the fact that there is a limited amount of underlying assets.

For linear hedging the firm should prefer forward contracts. They do not have any of the main drawbacks that futures have. In addition
the firm can choose to use forwards to hedge a risky asset like the interests rate or the exchange rate just as Froot, Scharfstein and
Stein described, or the contracts can be used to fix the demand and the price of the in -/outputs directly.

Options can be used for several purposes. They can be used the way Froot , Scharfstein and Stein described, either in order to hedge
the firm's financing opportunities or to protect a firm from very large changes in the risky asset. If the firm believes that the distribution
of value of the risky asset will lognormal, options will be better than linear instruments. Options exchanges have a limited number of
underlying assets in their portfolio. The firm must be prepared to use OTC options .

In a multiperiod framework caps and floors will work better than single-period options. The long time to maturity for these non-linear
instruments makes them wellsuited for Froot, Scharfstein and Stein's framework.

In their paper Froot, Scharfstein and Stein discussed futures , forwards and options , while swaps were mentioned only briefly. It is
obvious that swap contracts are very useful in Froot, Scharfstein and Stein's framework. There are several reasons for this.
 Time to maturity.
 Swaps can have a wide range of assets as the underlying.
 Swaps can be written between two firms, so that demand and price can be fixed.
                                                                                    132
 There is a wide range of variables in the contract terms that can be customised. This makes it possible to tailor the deal so that it
    fits each individual firm.
 Since swaps are off balance sheet contracts they are inexpensive. They do not require any margin payments.

While ordinary call/put options, caps/floors and plain vanilla swaps are useful it is equally clear that exotic options and swaps will
further increase the usefulness of Froot, Scharfstein and Stein's framework. By using instruments with a pay-off profile like the digital
option, or any of the similar instruments, the firm can avoid the overinvestment problem. 133 Other instruments and combinations of
instruments can further benefit the individual needs of firms wanting to implement Froot, Scharfstein and Stein's framework.

9.9.2          Advanced Hedging Strategies.
It is clear that while Froot, Scharfstein and Stein limited their discussion to quite simple hedging strategies it is possible to tailor the
protection from derivatives in much greater detail by relying on more complex strategies. Especially spread and straddle type
instrument combinations are well suited for Froot, Scharfstein and Stein's framework.


9.10            Can the Investment Opportunities Be Hedged?
Now that the investment triggers and the properties of the derivatives have been identified it is time to find out if it is possible to trade
derivatives on the investment triggers. This was after all the main purpose of the last two sections.

One must differentiate the discussion into two different parts. One part where the firm writes long -term contracts, forward contacts or
swap contracts with its suppliers or customer and one part where the firm relies on exchange traded or OTC derivatives without
considering actions taken by suppliers or customers.
If the firm fits the no-hedging model there is of cause no problem.

9.10.1          Contracts Written Between Suppliers and Customers.
Firms that fit into the simple model or the uncorrelated model of Froot, Scharfstein and Stein's framework and have a limited number of
suppliers/customers with long-term relations can use this approach. The firm can write long-term-, forward-, or swap contracts with this
business associates that fix the costs and revenues that the firm faces. The firm does not have to worry about what drives the prices of
the inputs or outputs because the deals guaranties a fixed cash in - and out flow. This will solve the problems of identifying risky
assets, finding derivatives on those assets and maintaining a portfolio of the derivatives. The use of intermediators will eliminate the
risks too. Since swaps have the longest lives, they may be preferred.

132
      See Table 13 Different kinds of swaps.», page 61.
133
      See section 7.2, «The Shareholders.», page 46.
9.10.2          Marketable Securities.
If the firm has either many suppliers/customer, short term relations with those suppliers/customers or if the firm fits t he speculation- or
                                                                                                                                        134
reversed simple model of Froot, Scharfstein and Stein's framework the firm may have to rely on exchange or OTC traded derivatives.
For these firms it must be possible to trade derivatives that correlate wit h the investment triggers. This goes for tactical, opportunistic
investments as well as strategic investments into new markets, perhaps through acquisitions. It must be possible to hedge both
enlargement investments as well as replacement investments . Ideally the investments should have a long-term horizon that will enable
the firm to operate under monopolistic, or at lease not full competition, conditions.

There are liquid markets for derivatives written on a number of the investment triggers identified in the previous section. The most
liquid markets are of cause the exchange traded futures and options markets. These do however have one major disadvantage, they
have short lives. This can be solved by using forwards, OTC options , caps, floors and swaps.

9.10.3          The Risky Assets that Can Be Hedged.
The following is a list of the things that causes firms to invest. They are presented in the order that they were identified in the previous
section, not by importance or any such thing. Along with the assets are the derivatives one can.




 Assets.                             135                    Comments.
                         Derivatives.
 Managerial              No.                                Difficult to quantify.
 resources.
 Technological           1.Digital OTC options, including                                                          136
                                                            1.The strike should be an if-then function of a patent. Not
 knowledge.                caps and floors.                   accomplishable.
                         2.Stock options.                   2.Stock options on any firm that can come up with a technological
                                                              breakthrough, including itself.
 Market position.       OTC options, including caps         The strike should be a function of the market. Same problem as for
                        and floors.                         technological knowledge (1).
 Reputation.            No.
 Scale.                 No.
 Pindock (1991) found that the variables mentioned above are important determinant s for corporate investments. Firms that use these
 variables in their investment decisions can not use Froot, Scharfstein and Stein's framework.
 Exchange rates         All. Also SAFEs, caps and         Available for most currencies.
                        floors.
                137     These are both in- and output goods. Since the category is so broad it is split up below.
 Commodities.
 Oil.                    All.
 Metals.                 All.
 Agriculture.            All.                               Limited in many countries though.
 Other industrial in-    All.                               OTC must be used for several assets. They can be tailored for each firm
 /out puts.                                                 and for each commodity. They are however more costly than exchange
                                                            traded instruments.
 Finished goods.         OTC.                               There are only exchange-traded instruments on raw or semi raw materials.
          138            OTC.                               The firm can write contracts where the payoff is a function of the marginal
 Taxes.

134
      This is also the approach one would use in a changing financing opportunities framework.
135
      «All» means options, futures, forwards and swaps.
136
     This would be almost like a bet between the bank and the firm, where the firm has inside information about the likelihood of
    developing the new technology. Because of this the option would be extremely expensive.
137
    These are determinants for both expansion investments and replacement investments .
138
      These are determinants for both expansion investments and replacement investments .
                                                             tax rate or its tax payments.
 Regulations.             Digital OTC options, including                                                             139
                                                             The strike will be an if-then function of a regulation.
                          caps and floors.
 Dixit and Pindock (1994) found that the variables mentioned above are important determinants for corporate investments. Firms that
 use these variables in their investment decisions can use Froot, Scharfstein and Stein's framework.
                140       All. Also FRAs, IRGs, caps and Available for most currencies.
 Interests rate.
                          floors.
 Exchange rate.           All.                               Available for most currencies.
 Business cycle.          Macro options and swaps as                                                 141
                                                             In addition to the business cycle itself , macro derivatives can be written
                          well as any stock index            on any macroeconomic variable, including leading indicators.
                          derivatives.
 If the interests rate or the state of the economy, the business cycle, determines the firm's investment decision, there are a number of
 derivatives that can be used and the firm can use Froot, Scharfstein and Stein 's framework.
Table 14 Which derivatives fit the different investment triggers.


9.10.3.1          Investment Triggers and Derivatives.
It seems that Froot, Scharfstein and Stein's framework runs into trouble here. The reason is that only some of the investment triggers
can be hedged. While assets such as the exchange rate, the interests rate, the price of in- and out puts as well as macrovariables can be
hedged, the firm specific investment triggers that Pindock (1991) identified can not be hedged. This means that the firm can rely on
Froot, Scharfstein and Stein's framework for some assets, but not for others. In addition, many firms buy and sell commodities where
there are no derivatives. The entire human capital industry which include law firms, consultants, news agencies etc can not hedge their
in-/outputs. This is a result of Froot, Scharfstein and Stein 's assumption that the asset that changes the cashflow from operations and
that also changes that investment opportunities can be hedged. It turns out that this was an unrealistic assumption, because there are
no derivatives on the firm specific investment triggers or for the investment triggers of many firms.

                                                       142
9.10.3.2          Strategic VS Tactical Investments.
Strategic investments change the very characteristics of the firm, while tactical investments change the firm's cashflow. The derivatives
identified above seems to be suited for hedging tactical investments but not strategic investments. It seems that this shortcoming in
Froot, Scharfstein and Stein's framework will enable the firm to avoid using the capital market whenever it wants to adjust operations
but that the firm still has to acquire external capital when if it wants to make strategic investments.

Since the firms strategic behaviour comes from the strategic investments Froot, Scharfstein and Stein 's framework can not help the firm
creating competitive advantage.

                                                 143
9.10.3.3          Replacement Investments
Non off the things that determine the frequency between replacement investments are underlying assets for exchange traded
instruments. It may however be possible to write contracts where the price of the new assets or the second-hand value of the old
assets. This will however most likely not be the case for a lot of firms. Since replacement investments count for larger inve stments for
firms in general than expansion investments this is a serious problem for Froot, Scharfstein and Stein's framework.

9.11             Conclusion.
Several of the investment triggers identified in the previous section can not be hedged with exchange traded or OTC derivativ es. This
means that there may be situations where the investment opportunities change and the firm does not have enough internal capital to
finance the investments. The firms must either give up the investment opportunities or acquire external capital. Neither option will
maximise shareholder value. It is also doubtful whether strategic investments can be financed at all for any firms through the use of
derivatives.

Firms that fit the simple or uncorrelated models can use long -term contracts. This will enable them to keep the investments at a
constant rate. Expansions will however be difficult without acquiring external capital.

Firms that fit the no-hedging model can and should apply Froot, Scharfstein and Stein's framework. A critical success factor is however
whether it is possible to protect the firms from large adverse changes in the cashflow. If it is not possible to trade non -linear
instruments on the risky asset that changes the cashflow this model may be dangerous. This would however be the case with or


139
    As opposed to the problem with patents neither part will have an obvious informational advantage. See footnote 136 above.
    Because of this it will be easier to negotiate a derivatives deal with a regulation as the underlying asset.
140
    Even though several empirical tests reject the importance of the interests rate as a determinant for firms investments it is included
    both because macroeconomic tests have shown that it is an important determinant and because it has so good theoretical backing
141
    Measured as changes in the GDP.
142
      See section 8.2.1, «Strategy.», page 49.
143
      See section 8.2.3, «Replacement Investments.», page 49.
without Froot, Scharfstein and Stein's framework since the firm would have difficulties protecting itself by relying on other risk
management theories too. Long-term contracts may be a better choice than staying unhedged.

Using the speculation or reversed simple models may be difficult for many firms. Th e reason is that there are marketable securities on
only a limited number of the investment triggers that firms face. If other things determine the investments in the firms thes e models can
not be used.


                                                           PART IV.
The preceding the sections have evaluated the underlying assumptions behind Froot, Scharfstein and Stein's framework. The last part
covers empirical findings, the conclusion as well as the appendix.



       10.           DOES EMPIRICAL EVIDENCE SUPPORT FROOT,
                    SCHARFSTEIN AND STEIN'S FRAMEWORK?
10.1          Research on the Topic.
Froot, Scharfstein and Stein said that it is difficult to find empirical evidence for their framework. They were themselves n ot clear
whether their framework should be interpreted as an optimal way for firms to manage their risk in real life or whether this is only a
theoretical framework. In the latter case one should look at their model as an interesting startingpoint for further development in the
field of corporate finance, like the M&M models.
Even if there is little empirical evidence to support their work, this does not mean that the framework is not applicable as it is. There are
several reasons why this may be the case.
 Froot, Scharfstein and Stein said that most hedging is off-balance sheet operations, which means that they is not included in
    databases such as the COMPUSTAT. Even if firms do use Froot, Scharfstein and Stein's framework researchers will not be able to
    find evidence for such hedging. This has however changed since Froot, Scharfstein and Stein published their paper. It is now
    required in the US that off-balance sheet instruments are included in the footnotes in the financial statement.
 Smith, Smithson and Wilford (1990) said that many firms do not want to talk about their use of derivatives, supposedly because of
    the bad press that derivatives have got lately. This means that surveys may give a result in Froot, Scharfstein and Stein's d isfavour.
 It is also a possibility that financial theory has forestall corporate policy and that firms have not yet started to use Froot,
    Scharfstein and Stein's insight, although they would be better off if they did.

The following three examples were presented by Froot, Scharfstein and Stein in their paper.
 Nance, Smith and Smithson (1993) used survey data to compare characteristics about firms that use hedging and those that do not.
   While some of their findings support Froot, Scharfstein and Stein, others are less clear cut. One thing that they did find however
   was that firms with high R&D expenses tend to hedge more. This would mean that they are using the simple model. Their
   investment opportunities are fixed, so to ensure a certain level of continuity, they want to eliminate the impact of changing
   cashflows.
 Nance, Smith and Smithson also found weak evidence that firms that are highly leveraged hedge more. Froot, Scharfstein and Stein
   see this as support for their framework because the firms would have more difficulty getting external financing. The increased
   bankruptcy probability and the increased probability of financial distress due to the high fixed costs associated with debt will
   increase the dead-weight cost of external financing.
 Finally, Nance, Smith and Smithson found that firms with high dividends use derivatives more. In order to use this finding as a
   support for the Froot, Scharfstein and Stein framework one must look at the high dividends as high fixed costs and use the
   interpretation mentioned above.

Tufano (1996) investigated the use of derivatives in the gold mining industry in Northern America. More that 85% of the firms used
derivatives. The goldmining industry is an industry which should theoretically fit very well into Froot, Scharfstein and Stei n's
framework. The main risky asset is of cause gold and other risky assets that may influence the pro fitability of the production are
currencies. It is easy to trade derivatives in both assets. It also seems likely that investments are correlated with the price of gold. This
study should therefor give very useful information about the usefulness of Froot, Scharfstein and Stein's framework.
Tufano's study does not support the theory that firms hedge to avoid bankruptcy costs. Neither did he find evidence that firms hedge
to reduce the likelihood of financial distress either. This sort of hedging would enable the firm to increase its debt and thus reduce the
over-investment problem. The over-investment problem increases the cost of equity. One possible explanation for this is that the
asymmetric information may be low in the gold mining industry, making overinvestments a small problem. This would reduce the cost
of new equity and there would be little need for Froot, Scharfstein and Stein's framework.
Tufano's findings do however support Froot, Scharfstein and Stein's framework in that he found that firms with low cash balances
hedge more. If all firms prefer internal financing, one could imagine that those firms with the least internal capital would use hedging to
optimise the little internal capital that they do have.
Tufano did however not find any evidence that firms hedge to avoid costly external capital. This roles out the explanation for why firms
with low cash balances hedge more. This is clearly evidence against the Froot, Scharfstein and Stein's framework. Tufano did however
only look at one industry and he said that it is not possible to draw conclusions about other industries.
He did find a negative correlation between large on-going projects and risk management. This is consistent with Froot, Scharfstein and
Stein because those firms may not need to invest much and the need for capital for investments is ob viously low.
To sum up these findings one have to conclude that this study does not necessarily support Froot, Scharfstein and Stein's fra mework.
Since the firms did not hedge to avoid bankruptcy costs or to reduce the likelihood of financial distress they are probably not using
Froot, Scharfstein and Stein's changing financing opportunities approach. Since the firms did not hedge to avoid costly external capital
it seems that there is little support for the changing investment theory too. However, it may still be the case that many of the firms fit
into the uncorrelated- and the speculation models. The latter would explain the fact that as many as 85% of the firms use derivatives.
Tufano's paper did not look at different forms of derivatives, only the level of hedging. If the firms that did hedge used derivatives that
become valuable whenever the cashflow from the mines is high, Tufano's findings would certainly support Froot, Scharfstein and
Stein. If they did not use such hedging strategies, although it is such a simple industry and indeed one that would fit Froot , Scharfstein
and Stein's framework, this would be a serious setback for Froot, Scharfstein and Stein.

Gèczy, Minton and Schrand (1997) investigated why firms use currency derivatives and which firms used them the most. They found
several things that support Froot, Scharfstein and Stein.
 Firms with greater growth opportunities and tighter financial constraints are more likely to use currency derivatives. Gèczy, Minton
   and Schrand used the book-to-market value as a proxy for growth opportunities. They said that their findings suggested that firms
   use currency derivatives to protect them selves against variations in cashflow that would otherwise result in underinvestment s.
   Myers (1984) said that firms with intangible assets and growth opportunities tend to borrow less. These two findings fit together.
   High debt would make the firm's cashflow more volatile and thus interfere with the investments necessary to exploit the growt h
   opportunities. In order for the firm to protect it selves further from volatility it uses derivatives.
 Another finding that they did that supported Froot, Scharfstein and Stein was that firms with more internal capital hedged less.
 They also found that firms that are exposed to underinvestments hedge more. They found that R&D firms hedge significantly more
   than other firms. R&D firms fit into the simple Froot, Scharfstein and Stein model. The investment opportunities , the R&D, is
   nonstochastic, while the cashflow from other operations in the firm may fluctuate. This supports Lewent and Kearney's (1990)
   findings. They found that firms that are exposed to exchange rate changes and are engaged in R&D use foreign financing. The
   reason was that the R&D usually is centralised and that it therefor is exposed to exchange rate changes.
 Gèczy, Minton and Schrand also found that firms hedge, they do not speculate. This is contrary to what Froot, Scharfstein and
   Stein said that firms might do. Froot, Scharfstein and Stein argued that if the investment opportunities are very sensitive to the risk
   that makes the cashflow fluctuate, the firm can use derivatives that increase the exposure. The empirical evidence shows that this is
   probably not done in practice.

There are several things that determine whether a firm uses currency derivatives or not.
 The firms with the most exposure hedge the most. The exposures that triggers hedging the most were, according to Gèczy, Minto n
   and Schrand, foreign operations, foreign debt and competition from foreign firms in the domestic market. Changes in the exchange
   rate can make the foreign competitors change the price in the market and thus the demand.
 Firms with economics of scale advantages to hedging use currency derivatives.
 Typically, larger firms hedge more than small firms do. This is contrary to what Warner (1977) said. He argued that small firms
   should hedge more. The reason for this was that the largest part of bankruptcy costs are fixed. The cost of a bankruptcy for a large
   and a small firm is therefor similar in absolute terms. The bankruptcy costs is however a larger proportion of a small firms total
   assets. For this reason a bankruptcy is more expensive for a small firm and it should thus hedge more. O'Brien and Bhushan (1990)
   also argued that small firms should hedge more then large firms. Their argument was that there is more information asymmetry in
   smaller firms.
Foreign denoted debt is used as a substitute for derivatives.
Gèczy, Minton and Schrand found the following to be the primary sources for exposure:
 Foreign income.
 Foreign sales.
 Foreign debt.
 Foreign debt.
 Foreign tax.
 Foreign operations.

Miam (1996) used the fact that information about derivatives are no w included in the financial statement. Contrary to Gèczy, Minton
and Schrand's findings. Miam found no evidence that hedgers have a higher growth opportunities /assets ratio. Both studies studied
currency hedging. One should keep in mind however that several studies have actually found that firms that go public have lar ger
market to book values than those that do not go public.
Miam argued that there are several reasons why capital market imperfections should induce hedging. This would support Froot,
Scharfstein and Stein's framework.
 There is more asymmetric information in firms with high market-to-book values. These firms should hedge more because external
    capital is more expensive for those firms. Miam found support for this. However, interests rate hedging primarily drives this
    hedging. As mentioned earlier, it is not clear whether or not the interests rate has anything to do with the firms investment
    decisions.
 There is also more hedging in unregulated industries for the same reason as mentioned above.
 Since acquiring capital involves fixed costs smaller firms should hedge more. Miam fo und a positive relationship between firm size
    and hedging.
The Economist (July, 1998) explained how firms use a hedging technique quite similar to Froot, Scharfstein and Stein's no -hedge model.
They said that while firms used to ensure them selves against risks, they now try to offset the risk internally. Viewing all operations as
a portfolio of activities, where a negative change in one area may be offset by a positive change in another activity, does t his. The firm
                                                                                             144
does not have to insure either activity and saves a lot of money on reduced premium.             The firm does not have to turn into a
gonglomerat to do this.

10.2             Do Firms Rely on Internal Capital?
Brown (1988) said that in 1983, 76% of all corporate expansions used retained earnings. He also said that firms normally borrow with
longer maturity than the project so that the firm has money when the payment is due. Only in periods with high interests rate will firms
prefer shorter debt. Both these findings confirm the Pecking order.

Brealey and Myers (1996) presented the following list of the percentage capital sources for US firms from 1981 - 94.
                             Year '81 '82 '83 '84 '85 '86 '87 '88 '89 '90 '91 '92 '93 '94
 Internal capital                  63 78 66 66 71 62 67 64 70 77 90 78 84 72
 External financing                37 22 34 34 21 38 33 36 30 23 10 22 16 28
 Net stock issue                    -4    1    4 -15 -17 -16 -13 -20 -22 -12              4    5    4 -6
 Net increase in debt              29 16 18 39 34 43 29 35 34 21 -2                            7    7 17
 Increase in account payable       12     5 12       4 12 10 18 21 18 14                  8 10      5 16
Table 15 Internal vs. external capital, 1981-94.
As one can see, firms generally prefer internal capital by a large margin. On average, 72% of the capital used came from internal
sources. Also, debt was preferred to equity by an overwhelming margin. This supports Myers' Pecking order theory. It should however
be noted that this period was characterised by a sharp increase in leveraged buy-outs and that the following years were spent reducing
the debt.

Chowdhry and Coval (1998) said that of the 64 billion dollars spent in 1989 by foreign investors acquiring or establishing US
businesses, 44 billions, or about two thirds, were funded by the foreign parent.

10.3             Do Variations in the Cashflow Cause Investments to Fluctuate?
Lamont (1997) looked at the 1986 fall in the oil-prices. He investigated whether the reduced income caused firms to reduce investments
also in their non-oil activities. He found that investments in non-oil activities fell by 50%. Lamont said that it is not obvious that the
cause for the reduced investments are the reduced cashflows in it selves. One possibility is that the reduced investments wer e caused
by reduced credit ratings. The cost of capital increases due to reduction in the collateral. Either way, this supports Froot, Scharfstein
and Stein's framework. There is a link between the cashflows from assets in place and investments.
 If the reduced investments in non-oil activities were caused by a reduction in the amount of internal capital, then this is evidence of
    the simple model and of the changing investment opportunities models.
 If the reduced investments were caused by higher capital costs, it is evidence of the changing financing opportunities model
    (which is not discussed in this thesis).
This strongly supports Froot, Scharfstein and Stein's framework. If the firm wants to rely on internal capital to finance future
investments the cash should not necessarily be invested in things like the money market or long -term government bonds. The money
should probably, depending on the investment opportunities and the investment triggers , be invested in derivatives with a carefully
designed payoff.

10.4             Conclusion.
Even if empirical evidence do not support Froot, Scharfstein and Stein's framework this may be due to the fact that firms have not yet
started to use the framework.

The empirical findings mentioned above give strong support to Froot, Scharfstein and Stein's framework.
Firms with high fixed costs (R&D, leverage or dividends) hedge more, as do firms with low cash balances and tight financial
constraints. These firms may end up not having enough internal capital to utilise favourable investment opportunities. The sa me is true
for firms with growth opportunities and firms which are exposed to risky assets. Both kinds of firms hedge. Firms hedge to av oid
underinvestments.
Firms with enough internal capital hedge less, they can rely on the internal capital that they have. Firms with large ongoing projects do
not need to invest, these firms hedge less.

Financing is also important. There are economics of scale in hedging. This may explain why larger firms hedge more. Firms do not
hedge to reduce bankruptcy costs. These costs would make debt more expensive. However, the fact that firms in unregulated
industries hedge more may indicate that they rely on hedging instead of costly debt.

Firms use derivatives for hedging, not speculating. This means that firms do not use the speculatio n model. This is however the only
model that is disproven in these findings. There are however no evidence for the no -hedging model either but this has probably to do
with the research questions in the surveyed literature.


144
      An alternative is to use a captive, an internal insurance company.
Froot, Scharfstein and Stein's framework is also supported by the facts that most expansions are financed internally and that variations
in the cashflow changes investments.

All in all these empirical findings give strong support to Froot, Scharfstein and Stein's framework.


                                               11.            CONCLUSION
11.1           Conclusion.
Froot, Scharfstein and Stein argued that because external capital is expensive firms should use derivatives to redistribute t he proceeds
from assets in place so that the amount of internal capital would correlate with the investment and financing opportunities.

The main question of this thesis was whether firms should use hedging to avoid acquiring costly external capital when investment
opportunities change. The answer to this question is no for most firms.

It turns out that external capital is not expensive for all firms. More importantly is the fact that although the derivatives market is huge,
the amount of underlying assets is limited. If the firm can not trade derivatives on the risky assets that determines the inv estment
opportunities Froot, Scharfstein and Stein's framework can not be used. This is true for most firms.
Empirical work does however support many of Froot, Scharfstein and Stein's arguments. This thesis can not come up with any go od
explanation for this deviation. One poss ible explanation is of cause that financial theories and their tests have not been able to model
the real world. For this to be true almost all resent literature on capital structure, capital acquisitions and investment de cisions would be
wrong and Froot, Scharfstein and Stein's framework would be right. This does not seem very likely.

This thesis covers only the investment part of Froot, Scharfstein and Stein's framework, not the financing. This may reduce t he
robustness of the conclusion. However, since this thesis rejects the null hypothesis that firms should use Froot, Scharfstein and
Stein's framework this decreased robustness does not matter.

11.2           Are the Costs of External Capital Necessarily High?
The question asked under this heading was: «Are there situations where the firm might just as well issue new capital?»

Bankruptcy costs, financial distress and agency problems make debt expensive. However for large firms, firms in slow growing
industries and firms with tradable collateral debt is priced close to a fair price. Likewise growthfirms can acquire external equity at a fair
price. These findings suggest that not all firms need to consider Froot, Scharfstein and Stein's framework, they can use the capital
market instead. It has also been shown that IPOs may not be underpriced, contrary to what earlier research has found. It seems even
more likely that SEOs are priced at a fair price, especially for large, old firms. The transaction costs of acquiring externa l capital are
however high, and since they are almost fixed this hurts small firms the most.

The answer to the question raised above is therefor yes for large firms with collateral in slow growing industries and for growthfirms.

11.3           What Are the Costs of Froot, Scharfstein and Stein's Framework?
The question her was «…whether the cost of hedging is a problem for Froot, Scharfstein and Stein's framework.»

The direct costs are probably far less than the direct costs of external capital since trading in de rivatives is quite cheap, especially for
the liquid instruments. The opportunity costs can however be considerable and off-balancesheet instruments is preferred.

No, the costs of hedging are not a problem.

11.4           Will Stakeholders Approve?
In this section is was asked if the stakeholders will accept Froot, Scharfstein and Stein's framework. More specifically, is this «…a
Pareto optimal financial policy?»

Implementing Froot, Scharfstein and Stein's simple and uncorrelated models should n ot be a problem. Many forms already have risk
management systems similar to these models.
However, holding derivatives can turn the firm into a take-over target, something that will hurt the manager. Anti-take-over measures
are needed. These derivatives can also lead to an overinvestment problem. This must be controlled for if the shareholders are to accept
Froot, Scharfstein and Stein's framework. The shareholders will probably benefit from the no -hedging-, speculation- and reversed
simple models since it is showed by Froot, Scharfstein and Stein that it maximises profits. It is however doubtful whether the manager,
the employees, and debtholders will accept the implementation of these models.

No, it is not certain that stakeholders will accept Froot, Scharfstein and Stein's framework. The simple and uncorrelated models are
probably Pareto optimal policies, the others are not.


11.5           What Makes the Firm Invest?
«Is it really the case that changes in one or more risky assets are the mos t important determinant for changes in the firm's
investments?» This was the question in this section.
There are a great many things that determine the investments of a firm, some are firm-specific while others, like the exchange rate and
the business cycle, are general economic variables. The answer to the question raised above id yes, but there is a large «but», as the
following lines show.

11.6             What Instruments Should Be Used?
In this section the question was «Are there derivatives written on the risky assets identified in Problem IV?» is was also shown that
«A close substitute may however not necessarily be good enough.» and that «…business cycle recessions typically lasts for a h alf to
two years and booms can last for up to 10 years.», and that «…exchange rate cycles are often even longer than business cycles .»
Also, this section wanted to answer the question of whether «…other instruments and more complex hedging strategies will furt her
increase the usefulness of Froot, Scharfstein and Stein's framework.»

Several of the investment triggers identified in the previous section can not be hedged with exchange traded or OTC derivativ es. Of
special concern is the fact that the firm will not be able to hedge strategic or replacement investments.

The optimal instruments for hedging in Froot, Scharfstein and Stein's framework are caps, floors, swaps and long -term contracts.
Hedging strategies like spreads and straddles should be used.

The answer to the question in this section is for a lot of firms no.

11.7  Does Empirical Evidence Support Froot, Scharfstein and Stein's
Framework?
Theories aside, what goes on in the real world? «After all, it is what works that is interest ing, not what is supposed to work.»

The empirical findings gave support to the simple model, the no -correlation model and the reversed simple model. The no -hedging
model can not be evaluated with the available literature. Evidence against the speculation model was found.

11.8             Does the Framework Management Risk? 145
Froot, Scharfstein and Stein limited their risk management to only cover changes in investment and financing opportunities. T he firm is
however faced with several other risks as well. By applying Froot , Scharfstein and Stein's framework the firm remains unprotected
against these other risks.
Business risk is the risk of changes in the firms operations. These are good risks and what creates shareholder value. A firm that uses
the long-term contract version of the simple- or uncorrelated models must make sure that the contracts do not limit this risk.
As mentioned above, Froot, Scharfstein and Stein's framework does not reduce the firms strategic risk, nor does it reduce the political
risk.
Froot, Scharfstein and Stein's framework does however manage financial risk. The framework does manage credit risk in an indirect way
since the derivatives are traded at exchanges and through banks. The credit risk towards customers is however not managed. Liquidity
risks and market risks, both currency, interestrate and commodity, are managed, while operational risk and legal risk are not managed.
Since the firm may have to use less than perfect hedges it takes on basis risk as well as model risk and mathematical risk if it uses Froot,
Scharfstein and Stein's framework.

This shows that Froot, Scharfstein and Stein's framework is not a perfect risk management strategy. However, neither are other risk
management strategies. It seems fair to say that Froot, Scharfstein and Stein's framework does not cover the different risks any poorer
than other theories. So even though some risks are unmanag ed in Froot, Scharfstein and Stein's framework this is probably not a
strong argument against it.


                                                   12.             APPENDIX
12.1             Value-at-Risk.
The main thing that separates Froot, Scharfstein and Stein's paper from other theories on hedging is that the goal of Froot, Scharfstein
and Stein's hedging is to maximise firm value, while other theories stabilises the cashflows. As an example of the fact that theories
emphasise stability for its own sake is one of the hottest new developments in applied finance lately, the Value at risk (VaR) concept.
Jorion (1997) described VaR as a measure of the worst expected loss over a given time interval under normal market conditions at a
given confidence level. Historic data of a risky asset is used to make a probability distribution of the returns. Next, one makes a
                                   146
confidence level, for example 95%. The return that lies exactly on the 95% mark of the return distribution is of particular interest. On
19 out of 20 days (or weeks or months etc.) the return on the risky asset will lie within the confidence level, that is, at or above the
return on the 95% mark. The VaR is the investment amount times the return on the 95% mark. This number quantifies the risk th at the
firm is exposed to. This number will depend on several of factors, where the person who calculates the value at risk can choose most of
them.
 The investment amount.
 The risky asset.


145
      See section 2.2, «Risk.», page 9.
146
      Different actors use different confidence levels. Bankers Trust uses 99%, Chemical and Chase use 97,5%, Citibank uses 95,4% a nd
      Bank America and J. P. Morgan use 95%.
 The time horizon of the historic data, which determines what the return is on the 95% mark.
 The confidence level.
                                                                           147
 The length of the holding period, that is, daily, weekly, monthly, etc.
 The normal distribution.
Value at risk is an extension of duration that is used to calculate interests rate exposure. VaR can use other risks than the interests rate
and several risks can be added in the same number. This way VaR summarises the entire exposure for the firm, including interests rate
exposure, exchange rate exposure, commodity prices etc.
Froot, Scharfstein and Stein criticised other hedging theories because they have the extreme conclusion that the firm should hedge
completely, which implies that all variations are bad. Froot, Scharfstein and Stein have demonstrated that this is not the ca se. If VaR
should be optimised in Froot, Scharfstein and Stein 's framework it would not be value at risk, but rather real investments at risk . This
may explain why VaR is mostly used by financial institutions.
VaR is only applicable in two cases in the Froot, Scharfstein and Stein framework, the simple model and the no-correlation model. In
both cases the optimal hedging strategy is to hedge completely, hence, all variations in the cashflows are bad. Froot, Scharf stein and
Stein argued that even though there may not be a lot of empirical evidence for their theory yet, it is still an useful model. Despite of
this, the growth of VaR proves that applied finance has not accepted their view of «good» risks.

12.2             Capital Structure Models.
To understand why external capital is more expensive than internal capital one must first understand what determines the cost of the
external capital, debt and equity. This may be easier said than done according to Myers (1984), who said that we know little about
capital structure.
Most text on «modern finance», which is not very modern anymore, starts with Modigliani and Millers two highly influential pa pers
from 1958 and 1963.

                                            148
12.2.1           Modigliani and Miller.

12.2.1.1           No taxes.
                                                                            149 150
Modigliani and Miller (1958) argued that given a set of assumptions , , where the most important ones are that there are no taxes
and no bankruptcy costs, it is impossible to create wealth on the right-hand side of the balance sheet. Modigliani and Miller's
Proposition I said that the value of a firm must be equal whether the firm is financed by equity or by debt.

                                                                 VL = VU, where

VL         = The value of a levered firm and
VU         = The value of an unlevered firm.

Since investors can borrow themselves no one will be willing to pay a premium for a firm partly financed by debt. As the leverage of a
firm increases, so do the fixed costs. Modigliani and Miller argued that the higher leverage increases the rate of return. The increased
rate of return is however offset completely by higher risk.




147
      The rule of thumb is that this should be the time it will take to liquidate the entire investment.
148
    Few explicit references are given in this section, but this part is based on the following books in financial theory: Copelan d and
    Weston (1988), Levy and Sarnat (1990) and Brealey and Myers (1991).
149
    M&M’s assumptions were:
 Capital markets are frictionless.
 Individuals can borrow and lend at the risk-free rate.
 Firms issue only two types of claims. equity and risk-free debt.
 Managers always maximise shareholders wealth.
 Corporate insiders and outsiders have the same information.
 All cash flows are perpetuities.
 There are no bankruptcy costs.
 In their 1958 paper there was no taxes whatsoever.
 In their 1963 paper there was corporate taxes, but no other taxes.
 All firms are assumed to be in the same risk class.
 However, Stiglitz (1969) found that the only assumptions necessary for this conclusion is:
 Individuals and firms face the same interests rates.
 There is no bankruptcy risk.
 No transaction costs and
 Free flow of information.
150
    Modigliani and Millers theory was developed before the CAPM , so that they used risk classes to compare different firms. The
    variability in the cashflow within a riskclass is identical, only the size of the firms can differ.
12.2.1.2        Corporate Taxes.
M&M's 1963 paper included corporate taxes. Since interest payments on debt are tax deductible, the value of a leveraged firm will be
greater than the unlevered firm.

                                                         VL = VU + TCBL, where

TC      = The tax rate on corporate income and
BL      = The market value of the debt, the bonds.

This means that a corner solution is the optimal, as much debt as possible. This was M odigliani and Miller's Proposition II.
To understand the difference between different capital structures one needs a measure of the cost of the different types of c apital
employed in the firm. Copeland and Weston (1988) described this as the weighted average cost of capital, WACC.

                                           WACC = (1-TC) k b B/(B+S) + k s S/(B+S), where

kb      = The before-tax cost of risk free debt.
ks      = The cost of equity.
B       = The market value of the debt.
S       = The market value of equity.

Since WACC = WACC(TC) the deductible tax rates clearly influences the value of the firm.




                             %                                           ks

                                                                         WACC

                                                                         kb
                                                                          B/S (leverage)



Figure 24 The cost of capital in a world without taxes, M&M 1958.
The figure above shows the results of M&M's Proposition I, the financing is irrelevant for the firm's value, i.e., the WACC is constant.
Since the goal of the firm is to maximise the shareholders' wealth, the financing does not matter for the firm.

                             %                                           ks


                                                                        WACC
                                                                        kb
                                                                         B/S (leverage)



Figure 25 The cost of capital in a world with corporate taxes, M&M 1963.
This illustrates the corner solution from M&M's Proposition II. With corporate taxes the firm should have as much debt as possible,
since this will reduce the taxes. One must however remember that taxes do not increase the value of the firm, it is just that the
deductible interests rates reduce the amount paid in tax.
The results that M&M came up with are extreme, as are the assumptions. Several papers have criticised M&M 's conclusions.

Jensen and Meckling (1976) argued that debt was used even before interest payments were tax deductible and that M&M's concussion
therefor is not convincing. However, the fact that interest payments are tax deductible has a value. Myers (1984) argued that this
implies that the higher the tax rates the higher the value of the firm. Fama and Miller (1972) demonstrated that risky debt increases the
value of the firm by the value of the tax subsidy on the interest payments.

Fama and French (1998) did empirical research on the conclusion in M&M's Proposition II. Contrary to Proposition II they found that
there is a negative relationship between firm value and debt. They said that the reason must be that there is information about
profitability in the debt ratio.
The arguments against M&M's Proposition II mentioned above do however deal with agency problems. There are however situations
where the agency problems are absent or overlooked. When a parent firm finances a subsidiary in another country one arrive at a
situation with much in common with M&M's assumptions. Chowdhry and Coval (1998) showed that taxation decides whether debt or
equity is used. If the tax rate in the subsidiary's country is larger than the tax rate in the parent's country then debt is used, since the
interests rates are tax deductible. This shows that in its pure form, M&M is still valid.

12.2.1.3          Personal taxes.
With taxes on both personal income and capital gains in addition to corporate taxes, the relation between the value of a leve red and an
unlevered firm becomes.

                                              VL = VU + BL * [1 - [(1-TC)(1-Tg)/(1-Tp)]], where

Tg         = The tax rate on capital gains.
Tp         = The tax rate on personal income.

The optimal capital structure now depends on the three tax rates. The relationship is as follows.

                                            1 > [(1-TC)(1-Tg)/(1-Tp)] 100% debt is optimal.
                                           1 < [(1-TC)(1-Tg)/(1-Tp)]  100% equity is optimal.
                                   1 = [(1-TC)(1-Tg)/(1-Tp)]  100% capital structure does not matter.

                                                          151
Miller (1977) claimed that under certain assumptions,         the introduction of personal taxes would give the same result as M&M's
Proposition I, capital structure does not matter. The firm issues bonds with an interests rate equal to r, where

                                                            r = r0 / (1-Tp) where

r0         = The interests rate of the tax-free municipal bonds.

If the firm wants to issue more debt it must raise the interests rate of the bonds so that investors with a higher personal tax rate will
invest. Firms will continue to issue debt until

                                                         r0 / (1-Tp) = r0 / (1-Tg), so
                                                                   VL = V U




151
      Millers assumptions were.
     Perfect certainty, that is, no bankruptcy.
     The tax rate on equity income is zero.
     There are municipal bonds whose return is tax-free and certain.
     Firms and individuals can borrow at the same pre-tax interests rate.

				
DOCUMENT INFO
Shared By:
Categories:
Stats:
views:243
posted:6/26/2011
language:English
pages:72