A210 Corporate Finance

Document Sample
A210 Corporate Finance Powered By Docstoc
					AC210: Corporate Finance

       Norvald Instefjord
 Introduction to Corporate Finance
• What is finance?
• What is the distinction between financial
  and real assets?
• What is corporate finance?
• What is the role of financial assets in
  corporate finance?
              Week 1
  Financial Markets and Financial
• How do firms finance their investments?
  – Earnings (free cash flow, internal capital)
  – Equity capital (external – public or private)
  – Debt capital (external)
• Public and private capital
• Trading of public capital
  – New issues
  – Secondary trading
              Equity Issues
• First time a firm seeks public equity is
  called an initial public offering (IPO)
  – Primary issue: new equity is issued
  – Secondary issue: existing private equity is
    sold to outside investors (most privatisations
    take this form)
  – Legal and underwriting services provided by
    investment banks
              Debt Issues
• Bank loans – not publicly traded
• Corporate Bonds – traded actively in the
  secondary market

• Debt capital and equity capital account for
  most of the firm’s financial capital
           Definition of Debt
• Fixed claim
  – Specifies what needs to be repaid to the
    investor and when
  – Default risk – risk that the repayment plan is
    not fulfilled
  – Conversion options – covenants that allow
    debt to be reclassified as equity
          Definition of Equity
• Residual claim
  – Does not specify a repayment plan
  – Repayment is defined as the residual:
    whatever is not claimed by other claim
    holders should go to the equity holders
  – Voting rights: Equity holders normally have a
    right to vote on important corporate decisions
    • Mergers, takeovers
    • Large investments
    • Board representation
    Trends in Corporate Finance
• Globalisation
• Deregulation
• Financial innovation
• Technological advances in the financial
• Securitization
   What you should take home
• You should be able to
  – Understand the distinction between a fixed claim and
    a residual claim
  – List the main attributes of a debt claim
  – List the main attributes of an equity claim
  – Describe the ways in which firms raise funds for new
  – Describe the difference between private and public
  – Describe the difference between bank loans and
    corporate bonds
• Grinblatt/Titman: Financial Markets and
  Corporate Strategy
  – Ch 1: overview of the process of raising
    capital for investment
  – Ch 2: overview of the process of raising debt
  – Ch 3: overview of the process of raising equity
1. Why do firms use underwriters when
   they issue new equity?
2. In what ways do you think it matters that
   debt holders have a fixed claim when
   equity holders have not?
3. In what ways do you think it matters that
   equity holders have voting rights when
   debt holders have not?
              Review problems
1.   Invest 95 and sell for 102 – what is the return?
2.   Invest 95 and sell for 102. Each transaction is charged
     a 1% trading commission – what is the return?
3.   Invest 95 and sell for 102. You receive additional
     interest payments/dividends of 2 during the holding
     period. What is the return?
4.   Invest 95 and sell for 110 three years later – what is
     the annual return on your investment?
5.   Invest 95 now and another 98 next year. In the
     following year you sell your investment for a total of
     202. What is the annual return on your investment?
             Week 2:
 Valuing Financial Assets: Portfolio
• Tool box
  – Expected portfolio return
  – Portfolio variance
  – Covariance between the return on two assets
• Optimal investment
  – “Fair” price of an asset means that the value equals
    the purchasing price
  – Even if prices are “fair” there are still ways of
    investing your money that is better than others
• Risk Aversion
  – Investors demand compensation for including risk in
    their portfolio
             Portfolio weights
• A portfolio of financial assets can be represented
  in a number of ways
  – The number of shares held in the various stocks (e.g.
    1000 shares in BT, 250 shares in Marks&Spencer
  – The dollar-value held in the various stocks (e.g.
    £2,500 in Lloyds Bank, £10,000 in Jarvis etc.)
  – As portfolio weights: the dollar-weight of the various
    stocks (e.g. if total portfolio is £100,000, then the
    portfolio weight of Lloyds is 0.025 and the portfolio
    weight of Jarvis is 0.1 etc.)
 From portfolio weights to portfolio
   expected return and variance
• To determine the expected return and
  variance of a portfolio we need to know
  – The portfolio weights
  – The expected return on the individual assets
  – The variance of the return on the individual
  – The covariance between the return on any
    pair of assets
       Expectation, Variance and
• Expected return (“average” return) is a location
• Variance of return is a spread measure
• Covariance is a measure of how the return of
  two assets are “related” (they can move in the
  same or opposite directions, or they can be
• If the returns move in the same directions,
  covariance is positive, if the returns move in the
  opposite directions, covariance is negative, and
  if uncorrelated, covariance is zero
     The input data for a portfolio
             of N assets
• N expected returns
• N variances
• N(N-1)/2 covariances

• Plus N portfolio weights

• For FTSE100 there are therefore
  100+100+100(99)/2 = 5150 data points that
  need to be estimated even before working out
  the portfolio weights

E (rP )   wi E (ri )
          i 1
Var(rP )   j 1 wi w j Cov(ri , rj )

              i 1
Var(rP )   wi2Var(ri )  2 wi w j Cov(ri , rj )
              i 1             i j
   Covariance and Correlation
• Covariance is a measure of relatedness
  that depends on the unit of measurement,
  so if the return is measured as a percent
  (e.g. 10 percent) or as a desimal (e.g.
  0.10) the covariance will be different
• Correlation is a measure of relatedness
  that is normalized to be independent of the
  unit of measurement
Covariance and Correlation

 Cov(ri , rj )  ij Var(ri ) Var(rj )  ij i j
                       Cov(ri , rj )
 Correlation  ij 
                           i j
    The Mean-Standard Deviation
      Approach to Investment
• Risk averse investors don’t like risk
• Variance averse investors don’t like risk that comes as
• This is not the same in general – variance aversion is a
  special case of risk aversion

• Portfolio theory takes the variance aversion approach –
  which in practice means that we assume investors wish
  to maximize their expected return given a certain
  variance, or minimize their variance given a certain
  expected return
 Mean-Standard Deviation
for Two-Asset Investments

E (rP )  wE (r1 )  (1  w) E (r2 )
 Var(rP )  w2Var(r1 )  (1  w) 2Var(r2 )  2w(1  w)Cov(r1 , r2 )
Portfolio Frontier
Mean-Std Dev for Portfolios of the
Risk Free Asset and a Risky Asset

 E (rP )  wE (r )  (1  w)rF  rF  w( E (r )  rF )
  Var(rP )  w Var(r )  Var(r ) w
 Covariance as Marginal Variance
• We can interpret the covariance between
  the return on a stock and the return on a
  portfolio as the stock’s marginal variance
• That is, if we increase the stock’s portfolio
  weight marginally, the portfolio variance
  will increase by approximately twice the
  stock’s covariance with the portfolio
           Algebraic “proof”

r  rP  mri  mrF  rP  m(ri  rF )
E (r )  E (rP )  m( E (ri )  rF )
Var(r )  Var(rP )  m Var(ri )  2mCov(rP , ri )

dVar(r )
          2mVar(ri )  2Cov(rP , ri )
dVar(r )
             2Cov(rP , ri )
  dm m 0
          What to take home
• Understanding of expected values, variances,
  and covariances
• Understanding of expected return and variance
  for a portfolio
• Understanding of risk aversion and variance
• Understanding of the portfolio frontier
• Appreciation of the linearity of expected return
  and standard deviation for portfolios consisting
  of the risk free asset and a risky portfolio
• Chapter 4 in Grinblatt/Titman
1. Variance: Prove that E(x-E(x))2=Ex2-
2. Covariance: Prove that E(x-E(x))(y-
3. Take a time series of returns 0.05, -0.03,
   0.10, 0.04, -0.10, 0.20. Estimate the
   expected return and the variance of
           Week 3:
From Mean-Variance to the CAPM
• Capital Market Line
  – Finding the market portfolio
• Two-fund Separation
  – Optimal diversification
  – Market vs idiosyncratic risk
• CAPM expected returns relationship
  – Expected return on assets depend on their
    covariance (i.e. their relatedness) with the market
  – Estimating beta risk
         Capital Market Line
• The line that goes through the risk free
  asset and the tangency portfolio
• Identification?
  – Maximization procedure
  – Simplifying “trick”, the excess return on any
    asset divided by its covariance with the
    tangency portfolio, is constant
  Maximization programme to find
     the Capital Market Line
• We can identify the frontier portfolios of
  risky assets
• Consider investments consisting of the risk
  free asset and a frontier portfolio – these
  are represented by straight lines
• For the frontier portfolio that is the
  tangency portfolio, the angle of the straight
  line is the steepest
Capital Market Line cont..

      E (rT )  rF
  w     Var(rT )
E (rT )  wE (rA )  (1  w) E (rB )
Var(rT )  w2Var(rA )  2 w(1  w)Cov(rA , rB )  (1  w) 2 Var(rB )

      w( E (rA )  E (rB ))  ( E (rB )  rF )
  w                  Var(rT )
    Capital Market Line cont..
• The maximization programme normally
  leads to a fairly complicated equation –
  with two risky assets we get a quadratic
  equation to solve
• In the class exercises you will be asked to
  have a go at such a problem
    Simplifying “trick”: finding the
        Capital Market Line
• We know the expected return on all risky
  assets and the risk free return
• The difference between the two is called
  the “excess return” for the asset
• The excess return, divided by its
  covariance with the tangency portfolio, is
  always constant
        Capital Market Line
Cov(ri , rT )  w1Cov(r1 , ri )  w2Cov(r2 , ri )  
 wiVar(ri )    wN Cov(rN , ri )

Cov(ri , rT )  E (ri )  rF

w1Var(r1 )    wN Cov(rN , r1 )  E (r1 )  rF
w1Cov(r1 , r2 )    wN Cov(rN , r2 )  E (r2 )  rF
w1Cov(r1 , rN )    wNVar(rN )  E (rN )  rF
           .002 .001         0
Var/Cov  .001 .002 .001
            0    .001 .002

                 .15 - .06
Excess Return  .17 - .06
                .17  .06
        Example cont..
  .002w1  .001w2  0 w3  .15  .06
.001w1  .002w2  .001w3  .17  .06
  0 w1  .001w2  .002w3  .17  .06

                      w1  40
                      w2  10
                      w3  50

                      w1  .4, w2  .1, w3  .5
      CAPM: Risk and Return
• Since the excess return divided by the
  covariance with the tangency portfolio is
  constant across assets, we can derive
  important relationships between risk and
• The covariance with the tangency portfolio
  is, if solved for the tangency portfolio itself,
  equal to the variance of the tangency
    Risk and Return
E (ri )  rF
               constant
Cov(ri , rT )
E (rT )  rF
              constant
 Var(rT )

E (ri )  rF    E (rT  rF )
Cov(ri , rT )    Var(rT )

      E (ri )  rF 
                     Cov(ri , rT )
                                   E (rT )  rF 
                      Var(rT )
      E (ri )  rF   i E (rT )  rF 
        Security Market Line
• The expected return of securities is linear
  in their beta-factors
• In the (beta,expected return) plane, the
  line crossing through (0,rF) and (1,E(rT)) is
  called the security market line
         Properties of betas
• Beta is linear: the beta of a portfolio of
  securities equals the portfolio-weighted
  average of the betas of the individual
• An implication is that the beta of the
  assets of the company equals the value-
  weighted beta of the liabilities of the
          Tracking portfolios
• A portfolio tracks another perfectly if the
  difference in the returns of the portfolios is
  a constant (possibly zero)
• Imperfect tracking: A portfolio consisting of
  a weight (1-b) in the risk free asset and a
  weight b in the tangency portfolio tracks a
  stock with beta β=b, because the two
  should have the same expected return
              Tracking Errors
• The two investments should have the same
  expected return, which implies that the tracking
  error has zero expectation and zero value
• Of course, investors do not like risk so they
  choose to hold the tracking portfolio instead of
  the stock
• Because such diversification is free of cost, the
  tracking error is also free of cost (i.e. it has zero
  Estimating the risk free return
• For risk free return use government bond
  or government bill data (long or short term
  instruments backed by the government)
• The return offered on such instruments is
  a good proxy for the actual risk free return
• Alternative, use the average return of a
  zero-beta risky stock, or the intercept with
  the y-axis if no zero-beta stock exists
  Estimating market risk premia
• Estimate the long-run average return on a broad
  stock market index and subtract the risk free rate
• Both the average stock market index return and
  the risk free return change over time
• The change in the difference is more volatile
  than the changes in the individual time series.
• Therefore, estimate the long-run average index
  return first. Do not estimate the difference
  between the market return and the risk free rate
                Beta estimation
• A raw beta estimate can be obtained from historical
  covariance and variance estimates (or by a regression)
• Average beta is one (this is the beta of the market index)
• If the raw estimate exceeds (is below) one, we know
  there is a possibility that the raw beta is an overestimate
• Raw beta estimates should be adjusted – i.e. they
  should be pulled down if they are above one or be
  bumped up if they are below one.
• There are ways of optimally adjust beta estimates
           Beta Adjustment
• Bloomberg adjustment
  – Adjusted beta = .66 times Unadjusted beta +
    .34 times One
• Rosenberg adjustment
  – Adjustment also incorporates fundamental
    variables (industry variables, company
    characteristics such as size, etc..)
• Also betas are adjusted sometimes to take
  into account infrequent trading problems
         What to take home
• Two-fund separation
• Capital Market Line vs Security Market
• Risk-Return relationships
• Tracking portfolio
• Parameter estimation: problems and
  current practice
• Grinblatt/Titman ch 5
• What is the tracking portfolio for a real
• How would you estimate the beta of the
  assets of a firm that has traded debt and
• How would you estimate the beta of a
  company that has never traded?
 Week 4: From CAPM to Arbitrage
         Pricing Theory
• Main purpose is to extend the valuation
  approach into more advanced and flexible
  valuation models
• CAPM can be thought of as a “one-factor” model
  (returns are determined by movements in the
  market portfolio only) but has important
  empirical problems (systematic deviations from
• APT extends to “multi-factor” pricing that can
  mitigate some of the CAPM’s empirical problems
          Risk Decomposition
• The Market Model
  – One-factor (the return on the market portfolio)
  – Related to the CAPM model
  – The regression estimates of the market model
    generates raw beta-estimates for the CAPM
• Risk Decomposition
  – Systematic (market) risk: asset risk that is explained
    by market movements
  – Unsystematic (diversifiable, idiosyncratic) risk: asset
    risk that cannot be explained by market movements
Market model regression

     rit   i   i rMt   it

      i  (1   i )rF
          cov(rit , rMt )
     i 
           var(rMt )
     cov( it , rMt )  0
      Risk Decomposition
var(rit )  total risk

var( i rMt )   var(rMt )  market risk

var( it )  idiosyncratic risk

var(rit )   var(rMt )  var( it )
APT: The arbitrage principle
  behind factor models
~             ~
r i  ai  bi f

    ~             ~            ~                      ~
w ri  (1  w) rj  w(ai  bi f )  (1  w)(a j  b j f )
Set wbi  (1  w)b j  0
which has solution w 
                            b j  bi
    bj   ~          bj  ~
         ri  1          rj  risk free  rF
b j  bi       b b 
                  j    i 
APT: Factor pricing
    bj                   bj 
           ai  1             a j  rF
 b j  bi         b b 
                         j   i 

 b j ai  bi a j  b j rF  bi rF
 ai  rF a j  rF
                  constant  
    bi      bj

 E (ri )  ai  rF  bi 

 For CAPM,   E (rM )  rF
         Multi-factor models

K - factor model

E (ri )  rF  bi11  bi 2 2    biK K

~               ~         ~                ~      ~
rit  ai  bi1 f1t  bi 2 f 2t    biK f Kt   it
        We do not know what
          the factors are!
• Can be evaluated statistically – using a
  method called factor analysis
• The output generates portfolios associated
  with each factor

• Can use firm characteristics or
  macroeconomic variables as proxies for
  the factors
                Factor betas
• The betas determine the asset’s sensitivity to the
• A high loading on factor number 2 means that
  the asset is particularly sensitive to risks
  associated with factor 2

• Factor models extends into portfolio analysis
  since the factor betas of portfolio is just the
  value-weighted average factor beta for the
  individual assets in the portfolio
    Factor models: computing the
    variance-covariance structure
• Recall that computing the variance-covariance
  structure requires a large number of estimates
• For N assets, N variance estimates and N(N-1)/2
  covariance estimates
• N=100, 100 variance estimates and 100(99)/2 =
  4950 covariance estimates
• Using the market model, we can work out the
  covariance structure from the beta estimates, i.e.
  from the N beta estimates
Covariance structure estimation
                ~              ~        ~
                ri  ai  bi f   i

                                ~       ~   ~    ~
     cov(ri , r j )  cov(bi f   i , b j f   j )
                                    ~             ~    ~
                     bi b j var( f )  bi cov( f ,  j )
                                    ~   ~         ~    ~
                     b j cov( f ,  i )  cov( i ,  j )
                     bi b j var( f )
Variance estimation

      ~               ~           ~
 var(r i )  bi2 var( f )  var( i )
             Tracking Portfolio
• Objective: to design a portfolio that has certain
  factor betas (or factor loadings)
• Why? The use of tracking portfolios are many
   – Risk management: if the company is subject to risks
     beyond its control, e.g. currency risk, it may create a
     tracking portfolio that offsets the risk
   – Capital allocation: the company may wish to allocate
     capital to investments that yield a greater return than
     their tracking portfolio and to reduce its exposure to
     investments that yield a smaller return than their
     tracking portfolio
 Designing a Tracking Portfolio
• First, determine the number of relevant factors
  (guesswork, statistical analysis)
• Second, determine the factor betas of the
  investment you wish to track (statistical analysis,
  comparison with existing traded companies)
• Third, gather a collection of different assets with
  known factor loadings
• Forth, calibrate your portfolio such that the
  portfolio factor beta equals the target factor beta
  for each factor
              cking portfolio, target betas 1 and 2
Two- factor tra
for factor1 and 2 respectively

Three assets with factor beta 1, 3, 1.5 for factor1
and - 4, 2, 0 for factor 2

Calibratio n :
x1  x2  x3  1
x1  3 x2  1.5 x3  1
 4 x1  2 x2  2

Output : x1  0.1, x2  0.3, x3  0.8
      Applying Pricing Theory
• Use pricing models to investment analysis
  (optimal investment strategies in financial
  markets – diversification)
• Use pricing models to calibrate
  investments (design of tracking portfolios)
• Use pricing models as a benchmark for
  real investment (comparing real
  investment returns to the return on
  tracking portfolios)
• Chapter 6 in Grinblatt/Titman
• There are three relevant factors driving
  asset returns
  – The factor structure of the debt of the
    company is (0.01, 0,0)
  – The factor structure of the equity of the
    company is (2,5,1)
  – The company consists of 1/3 debt and 2/3
• What is the factor structure of the
  company’s real assets (investments)?
Week 5: Investment Analysis – the
   case of Risk Free Projects
• Apply pricing technology to real
  investment analysis
• Net Present Value Rule
• Complications
  – Sunk cost
  – Opportunity cost
  – EVA and IRR
            Fisher Separation
• With different tastes, why should investors agree
  on investment policy?
  – Long-term vs short term
  – Risky vs Risk free
• Fisher separation
  – Agreement is optimal regardless of taste
  – Net present value rule: Invest in all projects that cost
    less than the value of the project’s tracking portfolio
  – NPV = PV(future investment) – Investment cost
• Cash flows of our investment
• Investment cost
• Discount rates (if risk free projects – use a
  risk free discount rate)
Present Value = sum of discounted
            cash flows
     Cash flows : C1 , C2 ,  , CT

     Tracking portfolios :
       C1              C 
             grows to  1 (1  r )  C1 in year 1
                       (1  r ) 
     (1  r)                    
       C2               C2 
              grows to             (1  r ) 2  C2 in year 2
                        (1  r ) 2 
     (1  r)2                      

     Present va  sum of the value of tracking portfolios
           C1        C2             C3
                           
         (1  r ) (1  r ) 2     (1  r )T
         Net Present Value

               C1      C2             CT
NPV   I 0                  
              1  r (1  r ) 2
                                   (1  r ) T
       NPV and Arbitrage
Adopt the project wh it has positive NPV
is equivalent to making money thro arbitrage

Cash flows : 20,40,10,30
               40     10       30
NPV  20              2
                                 3
                                     34.94
              1.05 1.05 1.05
Equivalent to undertaking the project, buying a
bond paying 40 in period 1, selling a bond paying
10 in period 2, buying a bond paying 30 in period
     Value Additivity of NPVs

Project A has NPVA
Project B has NPVB

Project A  B (i.e. the combined cash flows of the
two projects) has net present va :

    Mutually Exclusive Projects
• This is an “either-or” situation – you can invest in
  project A or you can invest in project B, but you
  cannot invest in both at the same time
• Both projects may have positive NPV so are
  worthwhile on their own
• “Either-or” situations often arise naturally. For
  instance, all timing decisions are mutually
  exclusive. You can invest now or you can invest
  in the future, but you cannot invest both now and
  in the future.
Which project to choose when they
     are mutually exclusive
• The choice criterion is to maximize the net
  present value of investment.
• Therefore, if you have two or more
  mutually exclusive projects to choose from
  you should choose the one with the most
  positive NPV.
          Capital Constraints
• There are situations in which you may
  have more projects with positive NPV
  available than you have funds for
  investment – i.e. you have a budget
• Then the choice criterion is to invest in the
  projects that offer the greatest profitability
      Profitability Index
Present Value cash flow PV
Investmentcost  I 0
Net present va NPV  PV  I 0

Profitability Index PI 
Project A : PVA  10m, I A  8m
Project B : PVB  100m, I B  90m
Project C : PVC  1000m, I C  950m

What is the optimal investment policy if the projects
are independent?
What is the optimal investment policy if the projects
are mutually exclusive?

Total investment budget B  100m.
What is the optimal investment policy subject
to staying within budget?
             Example cont.
Profitability indexes :
PI A      1.25
PI B       1.1111
PIC          1.0526
Invest in A first, then B and C.
Optimal mix : All of A (using 8 of 100 budget) plus
all of B (using additional 90 of budget, leaving 2 to
invest in C) plus       0.2105% of C
Total NPV : (10  8)  (100  90)      (1000  950)  12.11
       Economic Value Added
• EVA is a profitability measure that has become
  widely used in corporations – initially to replace
  accounting earnings or profit measures
• Accounting measures do not always measure
  economic performance (depreciation cost, for
  instance, is not a cash flow and should not be
  included in project evaluation)
• Accounting measures are therefore not directly
  consistent with NPV
• Economic Value Added is consistent with NPV
            EVA: Definition
• Three components
  – Cash flow
  – Change in asset base
  – Economic return on assets
• EVA(t) = Ct + (It – It-1) – rIt-1
• EVA(t) = Ct + It – (1+r)It-1
• Discounted sum of EVA(t) = Net Present
                 EVA, cont.
•   Investment of 100
•   The first year cash flow is 50
•   The second year cash flow is 150
•   Discount rate is 10%
•   Assets are depreciated by 50% in the first
    year and by 100% in the second year.

• NPV = -100 + 50/1.1+150/1.12=69.42
                   EVA, cont.
• EVA(0) = -100(cash flow)+(100-0)(change in assets)-
  0(0.1)(economic cost of initial assets) = 0
• EVA(1)=50(cash flow)+(50-100)(change in assets)-
  100(0.1)(economic cost of initial assets) = -10
• EVA(2)=150(cash flow)+(0-50)(change in assets-
  50(0.1)(economic cost of initial assets)= 95

• Discounted EVA = EVA(0)+EVA(1)/1.1+EVA(2)/1.12 =
  69.42 = NPV
   IRR: Internal Rate of Return
• Often managers base investment decision on
  the IRR instead of the NPV
• The rule is: if IRR is greater than the discount
  rate (i.e. the cost of capital) then adopt the
• In many cases this leads to the same investment
  decision, as IRR is greater than the discount rate
  only if the NPV is positive
• In other cases this is not true however, so to be
  safe always use NPV or EVA calculations

               C1      C2             CT
NPV   I 0                  
              1  r (1  r ) 2
                                   (1  r ) T

                 C1       C2              CT
    0  I0                     
              1  IRR (1  IRR) 2
                                      (1  IRR)T
• Investment cost = 100
• First year’s cash flow = 150
• Discount rate 10%

• NPV = -100+150/1.1=36.36
• IRR: 0=-100+150/(1+IRR) yields 50%

• Since 50% > 10% (IRR > discount rate) it is
  optimal to adopt the project
  Projects that have the cash flow
          profile of a loan
• “Investment cost” = 150
• Next year’s cash flow = -100
• Discount rate = 10%

• NPV = 150 – 100/1.1 = 59.09

• IRR: 0 = 150 – 100/(1+IRR) yields a negative
  IRR of -33.33% but this project is clearly
  profitable even though IRR < discount rate
           Problems with IRR
• IRR criterion is sensitive to the type of cash flow
  (asset or liability?)
• IRR is not unique in general (for T period
  projects there can be up to T different IRRs)
• IRR is not appropriate for mutually exclusive
  projects as small projects with high IRR and
  small NPV might then be preferred to large
  projects with low IRR and large NPV
      IRR and mutually exclusive
• Discount rate 2%
• Project A: -10, -16, +30
• Project B: -10, 2, 11

• NPV(A) = 3.149
• NPV(B) = 2.534

• IRR(A) = 10.79%
• IRR(B) = 15.36%
           Important points
• Fisher separation
• NPV definition
• NPV with mutually exclusive projects
• NPV with budget constraints
• EVA and NPV
• IRR pitfalls
• Grinblatt/Titman chapter 10
             Test for next week:
• Readings chapter 4, 5, 6 and 10
• Important formulas
  – CAPM: exp return = risk free plus risk adjustment
  – Beta-factor: covariance/variance
  – Factor models: exp return = risk free plus risk
• Risk free real investments
  –   NPV rule
  –   Profitability Index
  –   EVA
  –   IRR
       Very important formulas

CAPM : E (r )  rF   ( E (rM )  rF )
           Cov(r,r )
Beta :         M
           Var(rM )
Factor models : E (r )  rF  11     K K
(where denotes risk premium)
                                C1         CT
Net present va : NPV   I 0 
             lue                    
                               1 r     (1  r ) T
          Sample test questions
1.   The risk free return is 5% and the market index has an
     average return of 12%. What is the expected return for
     an asset with beta 1.5?
2.   An investment costs 100,000 and offers a cash flow of
     50,000 in year 1 and 150,000 in year 2. The discount
     rate is 5%. What is the net present value of the
     investment? Should you adopt the investment?
3.   In a two-factor market, the factor betas of asset A are 1
     and 0, and the factor betas of asset B are 0 and 1,
     respectively. The risk free return is 5%, and the
     average return on asset A and B are 10% and 15%,
     respectively. What are the risk premia associated with
     factor 1 and 2?
Week 6: Investing in Risky Projects
• Applying the CAPM and APT in the capital
  budgeting process
• Key problem: estimating the cost of capital
  for risky projects
  – Applying CAPM and APT
  – Using comparison firms
  – The dividend discount model
Risk Adjusted Discounting
  Compute the expected future cash flow (we do not
  know exactly wh we'll get) E (Ct ) in period t

  Compute the beta risk associated with this cash
  flow (the beta is the covarianceof the return wit
  the market return, over the varianceof the market
  return 

  Compute the expected market return of the cash
  flow r  rF   ( E (rM  rF )

                                E (Ct )
  Discount : PV ( E (Ct )) 
                               (1  r ) t
 Fundamental problem: Estimating
        the beta factor
• Betas for traded equity are easy to estimate –
  we simply regress equity returns on the index
  return, and possibly adjust to take into account
  estimation error (e.g. Bloomberg adjustment)
• Betas for projects are much more difficult to
  estimate as there simply does not exist a trading
• Possible solution: use comparison firms (firms
  we imagine has similar risk profile to the project
  in question)
      Using comparison firms
• Asset base needs to be sufficiently similar
  to the planned investment
• We need to adjust for leverage effects (the
  comparison firm may have debt)
  – In general, it is only the equity beta of the
    comparison firm we can estimate but we are
    really interested in the asset beta
  – The more the firm borrows, the higher the
    equity beta (even though the asset beta
    remains the same)
    Adjusting for leverage
Asset beta  value - weighted sum of debt and equity

     D    E
 A  D  E
     V    V
V  DE

 E   A  ( A   D )

Estimated equity beta equals the asset beta plus a
leverage term
Value assets100, value debt 40, and value equity 60.

Estimated equity beta 1.5
Estimated debt beta 0

1.5   A  (  A  0)

implies  A         0.9
    Implementing risk adjusted
discounting with comparison firms
       A project has the average beta of Church's Chicken,
       McDonald's and Wendy's. The equity betas of these
       three companies are 0.75,1.00, and 1.08 respectively.
       The debt and equity values of these companies are
       0.004 and 0.096 (Church's Chicken), 2.300 and
       7.700 (McDonalds) and 0.210 and 0.790 (Wendy's),
       respectively, and the beta of the debt of these
       companesis assumedequal to zero.

       The asset betas are
       Church's Chicken 0.72          0.75
       McDonalds 0.77       1.00
       Wendy's 0.85       1.08

                                     0.72  0.77  0.85
Average beta  project beta  0.78 

Risk free rate 4% and market risk premium 8.4%.

Cost of capital (discount rate) for theproject is
0.1055  0.04  0.78(0.084)
               Applying APT

The APT model estimates the cost of capital
                               lues are given by
by a factor model, so present va

                            E (Ct )
PV ( E (Ct )) 
                (1  rF  11     K K ) t
   APT and CAPM vs Alternative
• A drawback with the APT and CAPM
  models is that they require a number of
  estimates: the risk free rate of return, the
  beta factor(s), the market risk premium
  and the factor risk premia.
• It can in some circumstances be better to
  work with simpler model. The dividend
  growth model is an alternative to the APT
  and CAPM.
Dividend Discount Model
     div1  S1
S0 
      (1  r )
     div2  S 2
S1             , etc...
      (1  r )

      div1      div2           div1     div (1  g )
S0                                 1           
     (1  r ) (1  r ) 2
                              (1  r )   (1  r ) 2

S0 

rg       growthin dividends dividend yield
   What if comparison firms don’t
• In general there is little we can do
• However, if there exist firms where one
  division is similar to our project we may be
  able to identify the relevant betas.
• For instance, if you want to estimate the
  beta of the network division of television
  companies you can use the fact that these
  divisions play a varying role in generating
  the asset beta for these companies
  Network division example
General Electric : asset beta 0.99, 25% of value from
network division, 75% from non - network divisions.
Viacom : asset beta 0.78, 50% of value from network
division, 50% from non - network divisions. If non -
network divisions are sufficiently similar, we know
 GE  0.25 Network  0.75 Non  network  0.99
Viacom  0.5 Network  0.5 Non  network  0.78

  Network  0.36
  Pitfalls in using the comparison
• Project betas not the same as firm betas:
  mature projects generally lower beta than
  R&D projects etc
• Growth opportunities are usually the
  source of high betas: company value often
  significantly linked to future growth
  opportunities as opposed to current
• Investment cost 100,000
• Annual running cost 5,000 for 5 years
• Expected revenue stream 50,000 for 5
• Beta-risk of revenue stream 1.2
• Risk free return 5%
• Expected market return 12%
                    Example cont…
The discount rate for running costs: 5% (as the costs
can be assumed risk free?)
The discount rate for the revenue stream: 5% 
1.2(12%  5%)  13.4%
PV  5(1.05) 1  5(1.05)  2  5(1.05) 3  5(1.05)  4  5(1.05) 5
 50(1.134) 1  50(1.134)  2  50(1.134) 3  50(1.134)  4  50(1.134) 5
 21.65  174.16  152.51

NPV  100  152.51  52.51  0

Therefore,adopt project.
 Comparison method, example
• A firm with equity currently valued at 100,000
  and outstanding debt worth 50,000 holds 25%
  cash and 75% of a risky asset on its balance
• The equity beta is 1.5
• You consider investing in a project very similar
  to the risky asset owned by this firm
• The risk free rate is 5% and the expected return
  on the market is 12%
• Work out the project beta and the cost of capital
  for your project
Comparison method cont…
Assume debt beta is very close to 0, and also assume
that the cash balance has a beta close to 0

The total asset beta  A          1
Total asset beta  A  0.25(0)  0.75 RiskyAsset
 RiskyAsset        1.33

Cost of capital  5%  1.33(12%- 5%)  14.33%
• Grinblatt & Titman chapter 11
• I have not emphasized the certainty
  equivalent method
 Week 7: Taxes and Financing
• Irrelevance in the absence of transaction
  costs and taxes (Modigliani-Miller)
• Financing choices not neutral to taxation:
  – Level: corporate vs private tax rates
  – Timing: dividends can be deferred whereas
    interest payments on debt cannot
• The operating cash flow is divided into two
  – Cash flow to debt holders
  – Cash flow to equity holders
• Fundamental question: Does it matter how
  the split is made?
• If it does we can create value also through
  financing choices (not only through
  investment choices)
                 MM cont…
• Modigliani-Miller proved that capital structure
  choices are irrelevant – the split does not matter
• This proof rests on the absence of transaction
  costs of any kind: taxes, trading costs, and
  bankruptcy costs
• The proof of the MM theorem uses a “no
  arbitrage” argument – financial markets do not
  admit “free lunches”, or trading strategies giving
  you a positive cash flow with no prior investment
              MM cont…
• Consider two “versions” of the same firm –
  one version is U for unlevered (with no
  debt) and the other version L for levered
  (with debt)
• The firms have otherwise the same
  operating cash flow X
• The unlevered firm has value VU and the
  levered firm value VL
                MM cont…
• The fundamental question is whether VU and VL
• The cash flows of firm U’s equity holders is
  simply X
• The cash flow of firm L’s debt holders is (1+r)D
  to the firm’s debt holders and X-(1+r)D to the
  firm’s equity holders, in total a cash flow of X
• The value of L is the combined value of the debt
  and the equity
                    MM cont…
• Suppose VL is smaller than VU
• Then an investor can buy a 10% holding of L’s debt and
  a 10% holding of L’s equity, which entitles the investor to
  a 10% share in the total cash flow X. He would then go
  to the market and sell 10% of the cash flow X, which is
  valued at 10% of the value of U. This leaves him with
  zero future liability.
• His trading gains are 10% of the difference between VU
  and VL, which we have assumed is positive
• This cannot be possible in an arbitrage free market, so
  we can conclude that VL must be equal to or greater than
                    MM cont…
• Now suppose VU is smaller than VL
• An investor buys 10% of the cash flow X and sells 10%
  of a claim that promises the cash flow (1+r)D. The net
  cash flow is 10% of a claim that pays X-(1+r)D at
  maturity, which is priced at 10% of the equity in L
• The net future liability is zero, and the trading gains
  equal 10% of the difference between VL and VU, which
  we have assumed positive
• Again, this is not consistent with arbitrage free markets
• In conclusion, it must be the case that VU = VL and that
  capital structure is irrelevant
       What about risky debt?
• When the corporate debt contract is risky it may
  be difficult to find a “synthetic” corporate debt
  contract if a real one does not exist
• We must assume, therefore, that the markets
  are sufficiently complete in order to conclude
  that financing does not matter
• Complete market = a market where the
  dimensionality of the asset structure equals the
  dimensionality of the uncertainty structure
• If there are two states of nature (e.g. “good” and
  “bad”) then it suffices with two distinct assets to
  make the market complete
            Bankruptcy costs
• The Modigliani-Miller theorem also assumes that
  there are no deadweight costs of bankruptcy
• The debt holders may not get all their money
  back if the firm defaults, but this is not in itself
  enough to jeopardise the MM-theorem
• There must also be deadweight costs or
  liquidation costs (i.e. the value of the assets in
  default is less than the value of the assets as a
  going concern)
Taxes: Another important factor
• The tax system is generally fairly complex
  with different tax rates for different
  individuals and institutions, and for
  different types of income
• Therefore, it may be scope for “tax
  arbitrage” profits in financing
    After tax cash flow analysis
• A constant after tax discount rate r
• Tax rate for personal income from debt tD
• Tax rate for personal income from equity tE
• Corporate tax rate tC
• Earnings before taxes and interest payments X
• Earnings before taxes (X – kD) (k coupon rate, D
  nominal amount borrowed)
• After tax personal income from debt kD(1-tD)
• After tax earnings (X-kD)(1-tC)
• After tax personal income from equity (X-kD)(1-tC)(1-tE)
     After tax cash flow from investor perspective
  C  ( X  kD)(1  tC )(1  t E )  kD(1  t D )
                                           (1  tC )(1  t E ) 
     X (1  tC )(1  t E )  kD(1  t D )1 
                                                               
                                                1  tD         
    Discounted after tax cash flow
     X (1  tC )(1  t E ) kD(1  t D )  (1  tC )(1  t E ) 
DC                                    1 
                                                             
           1 r              1 r             1  tD         
     Value unlevered firm discounted tax benefits
• If there is a positive discounted tax benefit firms
  choose to borrow more, and investors with
  higher personal tax rate on debt income is
  encouraged to enter the market. This implies a
  reduction of tax benefits of borrowing.
• Reverse effect is there is a negative discounted
  tax benefit of borrowing
• In equilibrium, we expect the tax benefit from
  borrowing to be equal to zero
• This is the so-called “Miller’s equilibrium”
  described in Appendix 14A in the textbook
             Preferred stock
• Preferred stock: dividends on preferred stock are
  not tax deductible at the corporate level as are
  interest payments on debt
• This implies that corporate junior debt may be
  tax efficient relative to preferred stock
• However, the US tax code allows a 70% tax
  exclusion for preferred dividends paid to
  corporate holders, so the yield on preferred
  stock is often lower (before tax) than on junior
  debt even though the debt has seniority over the
  preferred stock
             Investor conflicts?
• Tax exempt equity holders prefer in general to reduce
  the borrowing of the firm so as to transfer income from
  debt repayments to dividend payments
• High-tax bracket investor prefer the opposite
• Often tax-exempt municipal bonds (or similar
  investments) offer yields that are greater than the after
  tax yield on corporate bonds for high-tax bracket
• Thus, the firm can give these investors an advantage by
  increasing the firm’s borrowing, as this frees capital that
  the investors can use to invest in tax-exempt municipal
• We expect to see a one-to-one relationship
  between inflation and nominal interest rates - if
  inflation increases by one percentage point then
  so do nominal interest rates
• Higher inflation, therefore, leads to higher
  nominal borrowing costs that yield in turn greater
  tax deductions
• Therefore, the tax effect has greater bite in
  periods of high inflation
              Empirical evidence
• Do firms with greater taxable earnings borrow more?
   – No, but this may be because firms in general rarely issue equity
   – Firms that perform poorly, therefore, tend to accumulate debt to
     meet their investments
• Tax code changes that affect the relative tax benefit of
  borrowing should have an impact on corporate financing
   – Yes, US tax reform of 1986 which reduced the tax benefits of
     other things than debt (such as depreciation rules and
     investment tax credits) gave rise to an increase in borrowing
     among firms most affected
   – The firms less affected did not increase their borrowing to the
     same extent
• Taxes matter but don’t explain everything
• Grinblatt/Titman chapter 14, including the
• 14.10 Are There Tax Advantages to
  Leasing not so relevant
1.    A firm has assets valued at 100, and debt valued at
      50. It plans to restructure its liability side by increasing
      its borrowing to 70 and paying a dividend of 20 to its
      shareholders. The debt has zero beta before and
      0.001 beta after the recapitalization. The beta of the
      equity is 2 before the recapitalization.
     a)   What are the values of the equity before and after the
     b)   What is the beta of the assets of the firm?
     c)   What is the beta of the equity after recapitalization?
     d)   The recapitalization has increased the beta of the debt (and
          therefore the cost of debt capital). Has it also increased the
          beta of the equity? Does this mean that the total cost of
          financing has increased? Explain.
 Week 8: Taxes and Dividends
• In frictionless markets dividends don’t
• Why do firms nonetheless pay dividends?
• Taxes and dividends
• Stock returns and dividend yields – what is
  the connection?
• Investment distortions caused by taxes in
     Cash flow to shareholders
• Shareholders earn money through holding equity that
  earns a cash flow (such as dividends) and capital gains
  (which can be realized through selling stock)
• The cash distribution to shareholders is normally
  discretional – the company can decide how much cash
  flow to give their shareholders
• Cash distribution comes in two forms – dividend
  payments and share repurchase schemes
• Dividend payments do not affect the number of shares
  but will reduce the value of each share
• Share repurchases do normally not affect the value of
  each share but will reduce the number of shares
How much of earnings is cash flow
       to shareholders?
• Dividend payout ratio: the ratio of dividends to
• In the US, this ratio has declined from about
  22% in 1980 to about 14% in 1998
• Over the same period, the ratio of share
  repurchases to earnings increased from 3% to
  about 14%
• The total ratio of cash flow to earnings has been
  relatively stable at about 25% of earnings
            Dividend yields
• Dividend yield is the ratio of dividends per
  share over price per share
• Typical pattern is that high-tech growth
  firms have low dividend yield and dividend
  payout ratios (Microsoft paid its very first
  dividend this year)
• Stable, old economy companies such as
  mining, oil and manufacturing pay about
  half their earnings as dividends
What is the optimal dividend payout
• Assumption: frictionless economy (no
  transaction costs, taxes, or other frictions)
• Investment policy unaffected by dividend
• Modigliani-Miller Dividend Irrelevance
  – The choice between paying dividends and
    repurchasing shares is a matter of
    indifference to shareholders
   Modigliani-Miller Irrelevance
• Consider two identical equity financed
  firms, the only difference is dividend policy
• Firm 1 pays 10m as dividends
• Firm 2 repurchases stock worth 10m
• After the end of the year, the firms are
  worth X
• In the beginning each firm has 1m shares
                       MM cont…
• Each share eventually sells for X divided by the number of shares
• Firm 2 buys back 10m worth of stock
• If share price is p, and firm 2 buys back n shares, we know that
• We also know that p=X/(1m-n)
• Suppose X = 150m
• Solving both equations gives us n = (10m1m)/(X+10m), so we get n
  = 62,500, and p = 150m/(1m-62,500) = 160
• Firm 1: stock price is p = 150m/1m = 150, but each stock gives a
  dividend worth 10m/1m = 10, so the total value of each stock is
  150+10 = 160
• Since shareholders get the same cash flow eventually, the shares
  must sell at the same price initially, i.e. dividend policy does not
    Taxes and cash distribution to
• Classical tax system
   – Dividends taxed as ordinary income and capital gains at a lower
     rate than ordinary income
   – Dividends are not tax deductible at corporate level, so dividends
     are also subject to corporate taxation
• Imputation system
   – Dividends are taxed as ordinary income but investors get a
     partial tax credit for corporate taxes (to offset personal taxes)
   – Dividends are not tax deductible at corporate level
• Systems that eliminate double taxation
   – Dividends are tax deductible at corporate level and taxed as
     ordinary income at investor level
         Classical tax system
• The classical tax system implies a tax
  disadvantage of dividend payments
• Dividend $100, 35% tax implies an immediate
  tax liability of $35
• Share repurchase scheme: an investor sells
  $100 worth of shares. Suppose original cost was
  $76. This implies a taxable capital gain of $24.
  Taxed at 20%, this implies an immediate tax
  liability of $4.8
• Share repurchase scheme much cheaper than
  paying dividends
      Tax avoidance schemes
• In theory, investors can often invest in a scheme
  that gives an immediate tax relief against a
  deferred future tax liability
• In practice, investors do not take advantage of
  these schemes but instead choose to pay taxes
  (or are unable to invest in tax avoidance
  schemes) on the received dividends
• The question is, therefore, why corporations
  continue to pay dividends when they are so tax
           Dividend clienteles
• Some investors do not pay taxes
• These investors will, everything else being
  equal, prefer high dividend yield firms to low
  dividend yield firms as they do not pay tax on the
• Firms might adopt different dividend policies to
  attract different investor clienteles
• Empirical evidence suggests that investors’
  portfolios have dividend yields that are related to
  their tax status (high tax bracket investors
  choose low dividend yield stocks and vice versa)
    Dividend payments and stock
• Do stocks with high dividend yield compensate
  investors for the tax disadvantage?
• Higher returns should then lead to lower values,
  reflecting the higher discount rates applied to
  future cash flows
• Research has focused on two returns effects
  – Ex-dividend day behaviour of stock prices
  – Whether cross-sectional dividend yield differences
    affect expected returns
      Ex-dividend day price drop
• If you buy the stock on the day before the ex-dividend day, you are
  entitled to the future value of the stock and the current dividend
• If you buy the stock on the ex-dividend day, you are entitled only to
  the future value of the stock
• The stock price should, therefore, drop on the ex-dividend day to
  reflect the dividend payment

• Empirical results from the 1960s indicate that the ex-dividend day
  price drop is about 77.7% of the dividend payment on average, but
  was higher (90%) for dividend payments greater than 5% of the
  stock price, and lower (50%) for the smallest dividends.
• These results indicate a tax effect (investors discount a tax rate of
  around 22.3% on dividends), and a clientele effect (investors with
  different tax rates hold portfolios with different dividend yields)
           Ex-dividend day cont…
• Transaction cost argument
    – Consider buying a stock at $20 before the ex-div day, receive a $1
      dividend, then sell the stock for $19.20. This yields $1 taxable profits
      and $(20-19.20) = $0.80 tax deductible losses. The net profit is $0.20
      less taxes, but it is still arbitrage profits. The stock needs to drop by the
      full amount to preclude arbitrage profits.
    – If there is a $0.10 per share transaction cost, the investor receives
      taxable profits of $1 in dividends, and incur $0.80 in tax deductible
      losses. The net profit is $0.20, but the investor must also pay $0.10 in
      transaction costs, so the net profit is only $0.10 less taxes. If the stock
      drops to $19.10, therefore, there are no arbitrage profits to be made.
    – If the dividend payment is only $0.40, the necessary price drop is $0.30
      to prevent arbitrage profits. That is, the price drop is greater for high
      dividend yielding stocks in percentage terms (as the clientele effect
• Price drop less than the dividend payment is also observed in
  countries that do not have a classical tax system, suggesting this is
  not a tax driven phenomenon at all
  Cross-sectional relation between
  dividend yield and stock returns
• If dividends are more heavily taxed than capital
  gains, the expected return must be greater for
  high dividend yield stocks.
• Empirically, stocks with high dividend yields
  have higher returns, but the relationship is not
• The relationship is U-shaped, with zero dividend
  yield stocks have higher expected return than
  stocks with low dividend yield, but for stocks
  paying dividends, the expected return increases
  with the dividend yield
How dividend taxes affect financing
    and investment decisions
• Marginal tax rate of 50%
• Company has a choice between paying $1m in dividends or retain
  the earnings
• Retained earnings yield 6% after corporate taxes (alternative II)
• Dividends yield 7% before personal taxes in corporate bonds
  (alternative I)
• Alternative I yields $500,000 to invest at 7%, which after tax yields
  $17,500 per year
• Alternative II yields $60,000 in extra dividend payments per year,
  which yields $30,000 after tax to the investor
• If you are a zero tax payer, however, alternative I yields $1,000,000
  to invest at 7%, which equals $70,000, and alternative II only
  $60,000 in additional dividends per year.
• Investors with different tax rates are likely to disagree with regard to
  the dividend policy the firm should pursue
         The general principle
• Investors prefer retained earnings if (1-corporate
  tax rate) x (pretax return internally at corporate
  level) > (after tax return at investor level)
• This has implications for investment policy as
  – Tax-exempt and tax-paying investors agree on
    externally funded projects but may disagree on
    internally funded ones (tax exempt investors require
    higher return on internal investment than tax-paying
• Grinblatt/Titman chapter 15
1.    A stock trades at 100p per share (prior to ex-dividend
      day) and the firm will pay a dividend of 10p per share.
     a)   Work out the ex-dividend day price if investors pay 40% tax on
          dividends and the ex-dividend day price equals the initial price
          less after-tax dividend payment
     b)   Work out the minimum transaction cost per share that
          prevents tax-arbitrage by a tax-paying investor
     c)   Suppose the dividend payment was 50p per share. What is
          your answer to a) and b) now?
     d)   Suppose the actual transaction cost is 2p per share. What are
          the arbitrage free price drops in a) and c) above now?
     e)   What are the “implied” tax rates on dividends in d)?
Week 9: Managerial Incentives and
       Corporate Finance
• Manager – shareholder conflicts
  – Occidental Petroleum and founder/CEO
    Armand Hammer case in the textbook
  – Maxwell Communications and Robert Maxwell
• How such conflicts affect investment,
  financing, and ownership structure
• How such conflicts can be mitigated by
  executive compensation schemes
     Separation of ownership and
• The separation of ownership and control is beneficial in
  terms of diversification and optimal investment while
  keeping a stable management team in control of the firm
• But it can be harmful if the management team is more
  interested in pursuing their own interest as opposed to
  their shareholders’ interests
• In what way do their interests differ?
   – Managers represent investors, customers, suppliers, and
     employees – not just investors
   – Managers get utility from non-financial benefits such as status,
     perks, job-security etc and are willing to spend corporate
     resources on these even though they are likely to be negative
     NPV projects
     Factors that determine the
    manager-shareholder conflict
• Proportions of stock owned by the manager
• Managerial entrenchment and lack of means to
  control managers
  – Diffuse ownership structure (no individual manager
    benefits enough to take action)
  – Proxy fights (shareholder revolt at general meeting)
    are very expensive and difficult to organize
• Bonus schemes not performance sensitive
• Changes in corporate governance have made
  managers more accountable in recent years
               Ownership structure
•   Ownership structure is on the whole more concentrated than we would
    expect (CAPM advocates diversification), particularly outside the US/UK
•   Ownership concentration a response to weak legal protection of
    shareholders’ interests
•   UK/US have the strongest protection and the most diffuse ownership

•   Managers tend to keep a significant ownership stake in firms where the
    incentive conflict with the shareholders is the greatest
•   In many internet IPOs, the managers kept a large share of their holding in
    order to get a higher price in the IPO (lock in clauses)
•   Eg. – Martha Lane-Fox and Brent Hoberman (founders –
    Hoberman still manager) were still large owners after IPO and were
    prevented from selling their share for a given time period after the IPO
•   Firms with higher concentration of management ownership have higher
    market values relative to their book values, provided management share is
    not too big. If it gets above 5%, managers become “entrenched” which
    allows them to pursue own interests more
 How managers distort investment
• Managers prefer investments that fit the manager’s
   – Makes him (her) more indispensable
• Investments in visible/fun industries
   – Raising the manager’s external profile (and his potential future
     job opportunities and wages)
• Investments that pay off early
   – Financial success in the short run can increase bonus, reduce
     the risk of losing job, increase the possibility of raising more
• Investments that reduce risk and increase the scope of
  the firm
   – To avoid bankruptcy the manager seeks relatively safe
     investments and may take a portfolio approach to investments
  Capital structure and managerial
• Managers are likely to prefer equity to debt because they
  are interested in minimizing the probability of default
• Shareholders may, therefore, prefer debt financing as
  debt is a good way to discipline managers (the fear of
  losing job is a good motivator)
• Empirical investigations show there is a positive
  relationship between leverage and
   –   Percentage of executive pay tied to performance
   –   Percentage of equity owned by managers
   –   Percentage of investment bankers on the board of directors
   –   Percentage of equity owned by large individual investors
• Debt is a good way to curb overinvestment
• Debt engages often a bank who is a good monitor of
       Executive compensation
• The problem of incentivizing managers is often called a
  principal-agent problem
   – Tenant farmer works the land of a land-owner. If compensated
     too much in terms of output, the tenant farmer must bear all the
     risk influencing output (weather etc). If compensated too little in
     terms of output, the tenant farmer doesn’t put in the required
   – Compensation is a matter of balancing the two concerns: Called
     the problem of designing the optimal incentive contract
   – Effort (input) cannot be observed, otherwise compensation could
     be tied to effort instead of output
   – Design objective is to minimize the agency costs of delegated
       Performance based executive
•   Jensen and Murphy (1990) found that a $1000 increase in firm value is
    associated with a $3 increase in CEO bonus (a $10m jet costs the CEO
    $30,000 just in lost bonus payments)
•   Some disagreement about this result, as it may have underestimated the
    real sensitivity by ignoring longer term impact on bonus payments
•   Substantial differences in pay-for-performance sensitivity across firms
     – Some explained by the agency costs of delegated control
     – Some explained by the risk of the firm
•   Over time, the pay-for-performance sensitivity has been increasing
•   Adoption of performance-based pay is generally a positive signal to the
•   What about relative performance sensitivity (pay linked to the position of the
    company relative to the average for the industry)? Relative performance-
    pay is rarely observed, but can be costly to investors in terms of price wars
    and overly aggressive competition.
•   Stock-based performance versus earnings-based performance. Stock
    based performance is much noisier than earnings-based performance, but
    in return earnings can be manipulated by the manager
 Mergers, Spin-offs, Carve Outs
• It may be easier to design an optimal compensation contract for a
  small, single-unit, firm than for a multi-divisional conglomerate
• Solution may be a spin-off (a division set up as an independent firm
  by distributing shares in the new firm to the existing investors) or a
  carve-out (do an IPO of the division and sell to new investors)
• Spin-offs and carve-outs are positive signals
• Mergers create the opposite effect, and in particular conglomerate
  mergers can be seen as a negative signal to investors as they affect
  managerial incentives negatively (conglomerate mergers are
  relatively rare now but were popular in the 1960s and 70s)
• Many spin-offs and carve-outs are reversing prior conglomerate
• Grinblatt/Titman chapter 18
1. The manager of a firm considers investing £1m
   of free cash flow (earnings currently held in a
   bank account) in a project that has private
   value £10,000 to the manager but NPV of -
   £200,000 to the investors. What is the optimal
   decision for the manager if
  a) He has fixed pay?
  b) He has in addition a bonus scheme where an
     increase of £1000 in the stock value leads to an
     increase of £10 to the manager?
  c) What is the optimal bonus scheme for the manager
     in this case?
Week 10: Information and Financial
• Key premise: managers have better information
  than investors
• What managers do, therefore, conveys
  information to the market
• Managers can
  – Distort accounts to manipulate the information flow
  – Reveal information through dividend policy, capital
    structure choice, and investment decisions
• Empirical evidence: how stock prices react to
  various financial decisions
         What can better informed
             individuals do?
• Signals: they act in a way that conveys their information
   – Difference between “cheap talk” and “credible action”
   – Signals need to be costly
• Pooling: they act in a way that everybody else act in
  order not to reveal information
   – It is too expensive to send a signal
• Manipulation
   – Actions: Investors overestimate the true cost of signalling
   – Reporting: “Bad” reports attract attention – it may be easier to
     disguise bad outcomes by submitting an “average” report
Distortions to managerial incentives
• Managers seek to maximize the share price
• The share price may, however, deviate from the “intrinsic
  value” (the full information price)
• Long term investors prefer that managers maximize the
  intrinsic value (which eventually transpires)
• Short term investors prefer that managers maximize the
  current share price (which may be distorted due to lack
  of information)

• The conflict is, therefore, essentially one of short-
  termism versus long-termism
 Why do managers care about the
      current share price?
• New issues or the managers may plan to
  sell private stock
• Low prices attract bidders in takeovers
• Managerial compensation directly linked to
  stock price
• Customers or employees may flee the
  company if the stock price goes too low
          Earnings manipulation
• The same underlying profits can be reported in different
  ways as earnings
   – Depends on the choice of depreciation method
   – Choice of inventory valuation method (FIFO LIFO)
   – The estimates of the economic value of assets, the estimates of
     the cost of guarantees or warranties issued, the estimates of the
     pension liability of the firm, the discount rates used for valuation
     of leases and pensions etc.
• There is a tendency to inflate reported earnings to
  increase the current stock price
• But managers may also find it useful sometimes to
  deflate reported earnings
   – For instance when the manager has just been hired
   – When applying for government subsidies or tariff protection
     against foreign competitors
    Short-termism in investment
• Bias towards short term projects because these makes it clear very
  quickly whether the investment is a good one
• Example:
    – Project A: yr 1 cash flow 40; yr 2-11 cash flow 80 per year; PV 840
    – Project B: yr 1 cash flow 60; yr 2-11 cash flow 50 per year; PV 560
    – Project C: yr 1 cash flow 40; yr 2-11 cash flow 40; PV 440
• Investors think C is much more realistic than A or B
• If company chooses A, the stock price is close to 440 after yr 1
  earnings are revealed, why?
• If company chooses B, the stock price is close to 560 after yr 1
  earnings are revealed, why?

• Company has a disincentive to choose the best project which is A
  because it is too similar to C in the first year
• If managers seek to maximize the intrinsic value they should choose
  A regardless
Dividends and Stock Repurchases:
      Announcement Effects
• An announcement of a dividend increase normally
  increases the stock price by about 2%
• If a company announces it is to cut its dividend
  completely, the stock price decreases by about 9.5%
• Is paying dividends therefore a good decision?
   – Dividends may be a costly signal conveying information that is
     hidden from investors
   – Paying dividends is, in effect, a cost to the shareholders to
     ensure that current information is reflected in current prices
   – The alternative: long term savings in signalling costs against the
     cost of deviations between the current stock price and the
     intrinsic value of equity
        Dividends and Investment
• News may be
   – Increased cash flow
   – Increase in investment opportunities
• An increase in dividends signals increased cash flow (as
  dividends then are more affordable) but is not consistent
  with an increase in investment opportunities (as they are
  then needed for investments)
• An increase/cut in dividends is, therefore, a more
  complex signal than is suggested in previous slides
• Empirical evidence suggests that cuts are viewed more
  favourably when the firms experience an increase in
  investment opportunities
 Capital Structure and Information
• Borrowing can also be thought of as a
  costly signal:
  – If mangers are convinced that future cash flow
    is high then the most credible way of
    communicating this information is to borrow
  – If the manager is “lying”, the firm is going to
    default on its debt liability and the manager
    will be out of a job
• Firms with poor prospects find it hard to
  “mimic” the same borrowing decisions
            Empirical Evidence
• Event study methodology
• Leverage increasing transactions (debt-for-equity swaps)
  have positive stock price response
• Leverage neutral transactions (debt-for-debt) have zero
• Leverage decreasings (equity-for-debt) have negative
  stock price response
• Security sales (equity, debt) have negative stock price
  response, and more so for equity than for debt
• Empirical evidence is consistent with information
  theories (this week) but is also consistent with incentive
  theories (last week)
              Adverse Selection
• Sick people tend to see health/life insurance as cheap –
  consequently they will be over-represented in the group
  of buyers of this type of insurance
• Example: very expensive insurance that covers 100% of
  all costs – or – cheap insurance that covers only 80% of
  all costs
   – In this case the sick people might migrate to the expensive type
     of insurance and the healthy ones to the cheap type
• This is called adverse selection – buyers or sellers do
  not always select themselves randomly but rather
  according to their “type”
• This also plays a role in the sale of corporate securities
 Managers have inside knowledge
 and at the same time sell or buy
       corporate securities
• Corporation can be expected to sell equity when
  the stock is overvalued and buy back equity
  when the stock is undervalued
• This makes sell transactions a bad signal and
  buy transactions a good signal
• This makes equity a bad source of capital for
  new investment, since it must be sold at a
  discount to the current stock price (why?)
• Pecking-order theory: firms prefer retained
  earnings to external capital, and external debt to
  external equity, when financing investments
• Grinblatt/Titman chapter 19
• A firm has already made an investment and is considering an
  additional investment opportunity
   – State of nature is good or bad, equal probabilities. Assume risk neutral
     valuation with zero discount rates. Manager knows the true state of
   – Current investment has value 150 (good) or 50 (bad)
   – NPV investment opportunity is 20 (good) or 10 (bad)
   – Currently the firm is financed by equity only
   – It plans to issue equity to finance the new investment, which costs 100
   – To do:
       • Set up the balance sheet before and after investment @ expected values
       • Work out how much of the existing equity the firm needs to sell in order to
         finance the investment
       • Compare the value of the existing (old) equity with investment and without
         investment in the good and the bad state
       • If the manager acts in the interests of the existing shareholders, should he
         always go ahead with investment. Explain.