Merger Arbitrage Evidence of Profitability by rsf15156


									            Merger Arbitrage: Evidence of Profitability

                               Taewon Yang and Ben Branch

Taewon Yang
Visiting Assistant Professor
University of Massachusetts
Amherst, Mass. 01002
Ph: 413-253-3869

Ben Branch
Professor of Finance
Isenberg School of Management
University of Massachusetts
Amherst, Mass. 01002

       Merger arbitrage is widely considered one of the principal areas of hedge fund

investment. While the investment process of merger arbitrage is generally known, less

information exists, at least in the practitioner community, as to academic research as to

the basis for various merger activity as well as to the profitability of such merger activity.

In this article we review various approaches to merger arbitrage as well as academic

research on the profitability of various merger arbitrage strategies.

       Merger arbitrage specialists invest in companies involved in a merger or an

acquisition. In an acquisition situation the manager will usually go long the stock of the

company being acquired and short the stock of the acquiring company. The stock of the

company being acquired will in general trade at a discount since all acquisitions take time

and there always is a risk that the acquisition will not be completed. Merger arbitrage

funds make investment profits when they successfully anticipate the outcome of an

announced merger and capture the spread between the current market price and the price

at which the stock will be trading at after the merger is completed.

       When a merger is pending, uncertainty about the outcome creates a pricing

disparity between the price of the acquiring company’s stock and the price of the target

company’s stock. Merger arbitrage managers evaluate announced mergers and

acquisitions and if they find favorable risk/return characteristics they will go long the

target company’s stock and sell the acquiring company’s stock. If the deal is completed

in the way the manager anticipates, profits will be made from the long position. Since

traditional investment funds are limited to their use of short selling, the merger arbitrage

strategy can not be found in a traditional mutual fund. Merger arbitrage hedge fund

managers do not attempt to anticipate possible mergers. Instead, they analyze already

announced mergers and acquisitions to identify favorable risk/return characteristics.

       A simple example would be an acquisition where a company is being acquired

through the use of cash only. Since the outcome of the acquisition is uncertain before the

transaction has taken place, the target company’s stock will trade at a discount from what

the deal would suggest. When the deal is completed, the investor receives the discount as

an investment return. In a stock-for-stock transaction the manager of a merger arbitrage

fund will go long the acquiring company’s stock and sell the target company’s stock short

to lock in the spread. For other than cash or stock-for-stock transactions the outcome of

the deal is generally more uncertain and the discount (and potential profit) greater.

       Managers of merger arbitrage funds control the risk of the portfolio by

diversifying across different markets and by investing in both announced mergers and

acquisitions. Holding a diversified portfolio will lower the impact any one position will

have on the portfolio. Open positions are monitored constantly. If new information

regarding the merger/acquisition process is announced the manager can either increase

the exposure (if the information makes the deal less uncertain) or withdraw money from

the position (if the information makes the deal more uncertain).

       Merger arbitrage is mainly event driven rather than market driven. Therefore,

investment returns generally have a low correlation with market returns. Since the

strategy is event driven the fund can theoretically make investment returns in any market

environment. However, mergers and acquisitions situations tend to occur in times of

economic upturns and bullish markets where high stock prices make stock-for-stock

acquisitions favorable. In periods of downturn the manager can have a hard time finding

enough merger arbitrage situations to keep the portfolio diversified.

Merger Arbitrage: Academic Evidence
       Until recently, most of the risk arbitrage literature focused on two types of

mergers: Stock swap mergers and cash tender offers (Brown and Raymond [1986],

Samuelson and Rosenthal [1986], Duke et al. [1992], Karolyi and Shannon [1998] and

Jindra and Walkling [1999], Mitchell and Pulvino [2000], Huston [2000] and Baker and

Savasoglu [2000]). In addition to the above merger types, other strategies include collar

mergers. The collar merger structure is a device in which the would-be acquirer offers a

variable exchange rate (acquirer shares for target shares). The proposed exchange rate is

made variable in order to reduce the risk of over/under payment in merger deals.

Historically, collar merger has not been included in merger/risk arbitrage research.

Collar mergers seemed to be implicitly classified as a sub-category of fixed rate stock

swap mergers in the merger literature. However, in risk arbitrage, the range of exchange

rates in collar mergers might well have different effects on returns and/or success rates

from that of fixed exchange rate mergers.

Academic research n risk arbitrage followed the success stories of the 1980’s merger

wave. Initially, risk arbitrage research focused on the predictability of success and price

movements associated with the proposed mergers.

The return derived from a risk arbitrage strategy depend on three factors:

1. The locked-up initial spread1.

2. The probability that the proposed merger will succeed and

3. The return for the risk arbitrager if the merger effort fails.

       Among them, only the locked-up initial spread is likely to be known on the

announcement date. Even this initial spread may be changed later. Clearly, predicting

which mergers are likely to succeed (and over what time period), as well as plausible

results and/or price movements, especially target price movements is a major issue and

challenge for the risk arbitrage industry (Asquith [1983], Samuelson and Rosenthal

[1986], Brown and Raymond [1986], Sander and Zdanowicz [1992], Jindra and

Walkling [1998] and Huston [2000]).

        The second line of research has explored why a risk arbitrage strategy may be

profitable and which factors explain the first factor in the return-generating process – the

initial spread. This research resulted from empirical findings of high abnormal returns

from the risk arbitrage strategy. Duke et al. [1992] and Jindra and Walkling [1998]

found an annualized excess return of over 100%. Karolyi and Shannon [1998] found an

annualized excess return of 33.9%. Baker and Savasoglu [2000] found a monthly

abnormal return of 8%.

        The third line of study focuses on how to calculate the appropriate risk-adjusted

return for a risk arbitrage strategy. This line of research seeks to analyze what the risk

arbitrage return-generating-process is and to test whether the existing linear asset pricing

model can be properly used to calculate the relevant excess return. The just discussed

second line of study in the literature implicitly assumes the correctness of the linear

return process in calculating the risk-adjusted return for a risk arbitrage strategy. The

literature in the third, however, finds evidence of a nonlinear return-generating-process.

(Merton [1981], Kalay and Loewnstein [1985], Bhagat, Brickley and Loewenstein

[1987], Glosten and Jagannathan [1994] and Fung and Hsieh [1997]). This finding

suggests that a nonlinear pricing model should be used to calculate the risk arbitrage risk-

adjusted return (Mitchell and Pulvino [2000]).

        The fourth line of risk arbitrage research focuses on technical issues such as how

to set up an index-time series. To undertake empirical analysis, one needs to deal with

time series or cross-sectional data that represent the market well. However, constructing

a time series or cross sectional data well representing the market is difficult. Especially,

if the investment is stylized, it is more difficult to have correct data for the stylized

market such as risk arbitrage.

        In this article, we review the extant literature along the second and the fourth lines

that may of more interest to the readers of this Journal.

Performance Literature on Stock Prices of the Acquirer and the Target

The price movement or performance of the target and acquirer’s stocks has been research

may further be classified into two strands. One focus on the wealth effect and the other

on the performance of risk arbitrage. The wealth effect literature explores the abnormal

returns of the target or acquirers’ stocks around the merger periods and their

determinants. This research tends to dominate the merger and acquisition literature. The

risk arbitrage literature, on the other hand, deals with the locked up spreads from the

targets’ and acquirers’ stock price movements. Risk arbitrage has been practiced

extensively as a trading technique on Wall Street for a number of years. Interestingly, the

wealth effect and risk arbitrage literatures seem to converge. That is, the wealth effect

literature represents the movement of the underlying securities – targets and acquirers’

stocks. The risk arbitrage literature deals with the long/short trading technique over the

movement of the underlying securities. Therefore, we first briefly review the underlying

hypotheses and empirical tests in the wealth effect literature. We then proceed to review

the risk arbitrage literature.

Wealth Effects in Merger: Hypotheses and Empirical Tests.

The wealth effect is well documented in the merger and acquisition literature, relating

lots of hypotheses for underlying motives to various empirical results in finance. We

concentrate on the literature dealing with abnormal returns during a merger

announcement and consummation dates.

Hypotheses Tested in Abnormal Returns Literature

Synergy, information and wealth transfer hypotheses (Hubris) were generally tested to

explain the gains over a merger period2. If the payment method –cash or stock - tends to

impact the abnormal returns over the merger period, hypotheses relating to asymmetric

information, taxation, investment opportunities, mode, cash availabilities, ownership

structure or business cycle hypothesis, are subsequently tested. 3

   1) Synergy Theory.

This theory asserts that acquisition may increase the combined value of the target and the

acquirer. The increase in value tends to result from capitalizing some specialized or

inefficiently used pre-merger resources such as inefficient management and operating

systems of the target. Therefore, if the resources are captured by other bidders in the

bidding competition, an original bidder may suffer losses. Empirical support for this

hypothesis was shown in Dodd and Ruback [1977], Bradley [1980], Bradley, Desai and

Kim [1982] and Bradley et al. [1983].

   2) Information Hypothesis.

This view asserts that new information generated during the acquisition leads to

reevaluate the targets. The reevaluation generates higher abnormal returns for the target

stockholders. Dodd and Ruback (1977) and Bradley (1980) found that the unsuccessful

target in tender offers tends to experience significantly positive abnormal return

     3) Hubris.

This hypothesis contends that the abnormal returns to the target result from the wealth

transferring from the bidders to the target. This was theoretically developed by Roll


   4) Asymmetric Information.

This hypothesis contends that the method of payment signals the intrinsic value of

bidders to the market, because bidders with the intrinsic value information may choose

the payment method that benefits themselves most. This hypothesis is consistent with

Jensen and Meckling [1976] and Myer and Majluf [1984]. They suggest that if the

bidder’s value is undervalued, managers of the acquiring firm would use cash, while they

would use stock if the stock is overvalued. This is empirically supported by Travlos


   5) Taxation.

Mergers that are accomplished with cash payment are directly related to the issue of

taxation. Target stockholders offered a cash payment may demand a premium to offset

the tax impact of the transaction. The premium may explain the abnormal returns to the

target stockholders (Wansley, Lane and Yang [1983] and Huang and Walkling [1987]).

   6) Investment Opportunities.

Debt financing maximizes the values of firms with poor investment opportunities while

equity financing maximize the value of firms with good investment opportunities.

Therefore, the firms with valuable investment opportunities are more likely to issue

equity. And the market may welcome the equity issuing. (Jung , Kim and Stulz [1995]

and Martin [1996])

    7) Risk Sharing

Hansen [1987] models the choice of payment method under the condition of information

asymmetry. In his model, if the bidder has less information than the target, the bidder

wants to use the stock forcing the target shareholders to share any post-acquisition risks.

He implied that this problem may become more apparent, as the target size increases.

    8) Controlling.

Shareholders and managers of the target firm will have different positions vis a vis the

takeover. The shareholders want to maximize the value of their stockholders as a result

of the takeover. For example, managers will not give up their position unless the wealth

gains from merger offset the lost benefits (Stulz, Walkling and Song [1990] and Song and

Walkling [1993]). Also, managers are reluctant to use their firms’ stock to finance

acquisition if doing so will dilute their control (Amihud, Lev and Travlos [1990] and

Jung, Kim and Stulz [1995]). Therefore, their ownership position may influence the

gains from the merger.

    9) Cash availabilities.

Firms with larger free cash flows are more likely to use the cash to finance their

investments (Myer [1984] and Jensen [1986])

    10) Regulation (Mode)

Mergers accomplished with cash payment tend to take less time than those with stock

payment to compete the acquisition, meeting legal requirements. Cash payment in merger

transactions is subject to Williams act and stock payment is to Securities Act of 1933.

     11).   Economic environments

Increasing economic activity has been shown to be positively related to stock financing

(March [1982] and Choe, Masulis and Nanda [1993])

    12). Competition.

Tender offers might have more competition in the future than mergers. This competition

may drive the abnormal returns (Suk and Sung [1997]).

Empirical Tests for Abnormal Returns and Hypotheses

In the beginning, the issues were positive/negative abnormal returns in successful/

unsuccessful merger and acquisitions, depending on the estimation period. In general,

studies find a positive abnormal return to the stockholders of the target (Dodd and Bruck

[1977] and Bradley [1980]) and a negative or zero abnormal return to those of the

acquirers around an announcement date. During the merger period, the abnormal return

for successful targets tends to increase. But the abnormal return for unsuccessful targets

tends to decrease. The acquirer has mixed results –negative or zero abnormal returns

(Dodd [1980], Asquith [1983], Bradley, Desai and Kim [1983], Song and Walkling

[1993], Jennings and Mazeo [1993], Sullivan, Jensen and Hudson [1994] and Suk and

Song [1997]).

       Huang and Walking [1987] and Travlos [1987] are among the first researchers to

pay attention to the payment method in mergers and acquisitions. Huang and Walkling

[1987] and Travlos [1987] find that cash payments tend to generate higher abnormal

returns than do stock payments. This finding has been supported by other empirical

studies. However, takeover attempts involving cash payments tend to suffer greater losses

than those involving stock payment at a termination date (Chang and Suk [1998]).

       To understand the underlying motives for the wealth distribution over the merger

process, many hypotheses and related variables have been intensively tested. A

synergistic hypothesis (Bradley, Desai and Kim [1983,1988], Bhagat, Brickley and

Loewenstein [1987] and Sullivan, Jensen and Hudson [1994]), an information hypothesis

(Dodd and Ruback [1977] and Bradley [1980]) and a wealth-transferring hypothesis (Roll

[1986]) have been tested to explain the overall wealth effect between the target and the

acquirer. Empirical results tend to support the synergy hypothesis. On the other hand, an

information asymmetry hypothesis (Myer and Majluf [1984], Sullivan, Jensen and

Hudson [1994] and Chang and Suk [1998]), taxation hypothesis (Huang and Walkling

[1987]), growth opportunity hypothesis (Martin [1996]) and competition hypothesis (Suk

and Song [1997]) etc have been used to explain abnormal return patterns with a payment

method (Travoli [1987] and Martin [1996]). Empirical results tend to support the

asymmetric information hypothesis. As explanatory variables, the payment method,

bidder’s competition, target’s resistance or debt ratio of a target mainly are found to drive

the abnormal return significantly. Managerial ownership structure is found to be

significant only in the competing bids but not significant otherwise (Stulz, Walkling and

Song [1990] and Song and Walkling [1993]). Institutional ownership shows mixed

results. The target size is found to affect the gains to targets. And most recently, results

on the types of strategic acquisitions indicate that strategic acquisitions do not generally

influence the gains to bidders over an acquisition, but a diversification strategy with

potential overlap generates negative impacts on the bidder’s stock price (Walker [2000]).

       To deal with abnormal return calculation and statistical tests, many researchers

used a market model with equally weighted market or value weighted market index. A t-

or z-test with various standard deviation formulas is used to determine the statistical

significance of the abnormal returns. In the determinant analysis, a multivariate

regression model is used to examine the explanatory factors. The event periods such as t-

10 to t+10, t-5 to t+5 or t-1 to t+1, are typically used for calculating the cumulative

abnormal return.

Performance of risk arbitrage

The existing risk arbitrage literature is primarily empirical in nature. Relatively little

theoretical research has been done on the subject. Most risk arbitrage research used cash

tender offers or/and stock swap mergers to calculate returns of risk arbitrage. We adopt

the same classification scheme here.

A) Cash tender offer

Duke et al. [1992] examined 761 cash tender offers that took place during 1971 to 1985.

They reported an average abnormal return of 24.6% for 52.4 days, corresponding to an

annualized abnormal return of 171%. This abnormal return was calculated assuming that

an investor bought the target stock on the tender offer announcement day and sold the

stock on the resolution day. They found that this abnormal return tends to be inversely

related to the probability of tender success.

        Jindra and Walkling [1998] measured the excess return of a risk arbitrage strategy

without taking accounts of the input of dividends. They used a sample of 362 cash tender

offers in which a bidder seeks for 100% of target shares and the transaction value exceeds

$10 million during 1981 to 1995. The single factor market model with the CRSP value

weighted market index was used to calculate the excess returns. Parameters were

estimated during a period of t-260 to t-20. They calculated the cumulative abnormal

returns on targets over one day to eight pseudo-weeks after an announcement.

Interestingly, an arbitrageur was found to earn an excess return of 1.42 to 1.54% or an

annualized return of 102 to 115% if she or he takes a long position in a target during one

week following the announcement. Also, they found that an arbitrageur buying the target

and shorting the acquirer during one week following the announcement tended to earn an

excess return of 1.88% or the corresponding annualized return of 156%.

        Empirically, they found abnormally high volume4 before and after the tender offer

announcement, indicating significant additional trading activity. In a multivariate

regression analysis, the arbitrage spread5 was found to be, on average, positively related

to the size of bid premium, friendly managerial attitude about the offer and the existence

of rumors about the offer. But it seemed to be negatively related to the target size and

pre-offer run up. The managerial ownership was, on average, found to be only

significant in hostile takeovers. Bidder’s acquisition experience was found to be

insignificant. Also, in terms of variables unknown at the announcement date, the spread

tended to be positively related to the duration of the offer but negatively related to the

revision of the offer.

B) Stock swap merger

Karolyi and Shannon [1998] measured the return of risk arbitrage without dividends.

They used the 37 Canadian acquisitions with over $50 million during 1997 and

multivariate regression analysis. The return was calculated, assuming that an investor

buys the target and shorts the bidder one day after the tender announcement. They found

that risk arbitrage tended to generate 4.78% in excess of the TSE 300 stock index during

an average period of 57 days, or an annualized excess return 33.9%. The spread may be

significantly related to target size and pre-announcement run-up (2 week). But the excess

return was unlikely to be related to the likelihood of success (days to close), target size,

beta, price to sale ratio, price to book ratio, payment method, pre-announcement run-up

or industry sector.

C) Cash tender /Stock swap mergers

Mitchell and Pulvino [2000] calculated the daily excess return. They used a sample of

4,750 stock swap mergers, cash mergers and cash tender offers during 1963 to 1998 and a

contingent approach – selling uncovered index put options. They found that risk

arbitrage tended to generate an annual excess return of 4%. The annualized return

calculation method, transaction costs6 and a risk premium for deal-failure were attributed

to the higher returns in risk arbitrage.

        Baker and Savasoglu [2000] explored risk arbitrage, using 2,088 cash and stock

merger/acquisitions during a period of 1981 to 1996. They applied the return formulas

below to calculate the risk arbitrage returns. Event study methodology was used to test

determinants for excess returns. The risk arbitrage strategy tends to generate a monthly

abnormal return, 8%. Excess returns were found to be positively related to the firm size,

dollar trading volume and two measures of idiosyncratic risk – takeover premium and the

probability of success. And unlike Mitchell and Pulvino [2000], they found that the

transaction costs were not the major determinant of return.


The risk arbitrage literature generally found annualized returns in excess of 100% for a

risk arbitrage strategy applied to the cash tender offers during 1971 to 1985 and 1981 to

1995 (Duke at al. [1992] and Jindra and Walkling [1998]). Karolyi and Shannon [1998]

find an annualized return of 26% in stock mergers in the Canadian market during 1997.

And recently, Mitchell and Pulvino (2000) report that the risk arbitrage in stock merger,

cash merger and cash tender offers during 1963 to 1998 tended to generate annual excess

returns of 4% after controlling for nonlinear return profile and transaction costs. Baker

and Savasoglu [2000] document a monthly abnormal return of 8% in the risk arbitrage

strategy during 1981 to 1996.

       The possible explanations that have been suggested in the literature for these high

returns include: an anomaly, the impact of liquidity or transaction costs, and/or a risk

premium. Shleifer and Vishny [1997], using a theoretical model, suggest that there is an

inverse relationship between the arbitrageur’s liquidity and potential arbitrage profits in

the market. Mitchell and Pulvino [2000] suggest transaction costs, other practical

limitations and the risk premium for deal-failure as reasons to expect the higher return for

a risk arbitrage strategy. However, Baker and Savasoglu (2000) find that the transaction

cost was not the major determinant of risk arbitrage returns.

       The literature also explores firm/deal-specific factors and economic circumstance

as potential explainers of the initial spread, applying a multivariate regression analysis.

Jindra and Walkling [1998] classify the factors into two groups: known and unknown

factors as of the announcement date. They find the initial spread to be positively related

to the known factors such as bid-premium, target managerial ownership, target

managerial attitude towards the offer, rumors and run-ups and negatively related to the

unknown factors such as revision of exchange ratio and duration. Brown and Ryngaert

[1992] find the tendering rate to be positively related to the bid-premium. And in terms of

economic environment, Moor [1999] argues that economic factors might not influence

risk arbitrage but Mitchell and Pulvino [1999] find risk arbitrage returns to be positively

related to the level of market returns in severely depreciating markets but uncorrelated in

flat or appreciating markets. Baker and Savasoglu [2000] contend that the excess returns

are positively related to the firm size, trading volume, takeover premium and probability

of success. The wealth effect and risk arbitrage literature are summarized in Appendix 1.

Passive/Naïve Indices

Historically, setting up the indices for the underlying markets seemed to receive less

attention than dealing with biases in the current stylized indices. Interestingly, both

indices tend to deal with the same issues in the performance literatures. That is, finding

right indices/time series that reflect the stylized/underlying markets. However, the

stylized indices are likely to be limited in reflecting the markets. These current stylized

indices in the markets are structured to represent the selection and timing skills of the

fund managers, not the underlying stylized markets such as risk arbitrage markets. In

other words, the stylized indices tend to show how the funds (and their managers) are

doing, but not the markets. These characteristics also cause survivor biases in the

performance estimation.

Fund Indices and Survivorship Bias

Kalay and Loewnstein [1985] show how to test the survivor biases problem in calculating

abnormal return around the dividend payment announcement, using the sub period

analysis –sensitivity analysis. In exploring the performance of mutual funds, Brown,

Goetzman, Ibbotson and Ross [1992], and Brown and Goetzman [1995] conclude that

surviving funds generated higher returns than all funds, suggesting that there is a

survivorship bias in performance measurement. Malkiel [1995] explored the performance

of equity mutual funds during 1971 to 1991. He tested all equity mutual funds existing

each year, and reported that survivorship bias was substantial. The performance

persistency is likely to be influenced by survivorship bias. However, persistency may not

be robust as it is found that there was strong persistency in the 1970s but not in the 1980s.

Fung and Hsieh [1997] explored the survivorship bias in CTAs during 1980 to 1995. The

survivorship measurement was defined as the difference in average returns between

surviving funds and all funds. They find that the style analysis is not influenced by the

survivorship bias. Ackermann et al. [1999] examined the performance of 906 hedge

funds during 1988 to 1995. They find that hedge funds tend to outperform mutual funds.

Hedge funds tended to be more volatile than mutual funds and market indices. In a

subsection, they explored data conditioning biases such as survivorship bias, liquidation

bias, backfilling bias and multi-period sampling bias, etc. They found no systematic bias

in their conclusions.

Risk arbitrage indices/time series from the underlying assets

In the risk arbitrage literature, research on how to set up time series data that truly

represent the risk arbitrage market is sparse. Jindra and Walkling [1998] and Baker and

Savasoglu [2000] used cross-sectional data to analyze the initial spread, while Mitchell

and Pulvino [2000] developed time series indices: passive risk arbitrage portfolio

(PRAM) and calendar time value weighted average of returns (VWRA). Both are

monthly indices. PRAM includes the transaction costs – brokerage fee and price impact

from illiquidity – and trading constraints. PRAM seems to simulate a risk arbitrage

portfolio. The portfolio starts with $1 million in 1963. The investment for $1 million has

three constraints: 1) any investment on a merger can’t exceed 10% of the total portfolio’s

value. 2) the amounts invested in any single deal are decided such that the price impact

on stocks is less than 5%7. 3) PRAM doesn’t allow leverage in investments. In VWRA,

monthly returns are obtained by calculating a weighted average of each monthly return of

deals. And the weight was the total market value of each target. VWRA assumes that

risk arbitrage will be set up for every merger transactions. Also, no transaction costs are

assumed. They find empirical results from using PRAM similar to those from using

actual risk arbitrage hedge funds during 1990 to 1998.


The empirical results are, on average, influenced by the characteristics of data.

Especially if the data are rare or hard to be taken, the empirical results will be dominated

by available data. The stylized investment markets seem to have the similar situation.

Though some stylized indices are available in the markets, they, as active indices, tend to

represent the selection and timing skills of the fund managers rather than the underlying

stylized markets. Also, the haunting bias problems in the existing indices have been

known to cause other problems in testing hypotheses.

Ackermann. C, R. Mckenally and D. Ravenscraft , 1999 “ The Performance of Hedge
Funds: Risks, Return and Incentives.” Journal of Finance, Vol.54, No.2, pp. 833-874

Amihud. Y, Lev. B and N. G Travlos, 1990 “ Corporate control and the choice of
investment financing: the case of corporate acquisition” Journal of Finance Vol.1990,

Asquith. P 1983 “Merger bids, Uncertainty and Stockholder Return” Journal of Financial
Economics Vol 11, pp.51 – 83.

Baker. M and Savasoglu. S, 2000 “ Limited arbitrage in mergers and acquisitions”
Working paper, Havard University.

Bhagat. S, J Brickley and U. Loewenstein, 1987 “ The Pricing Effects of Interfirm Cash
Tender Offers” The Journal of Finance Vol 42, pp.965 – 986.

Bradley. M , 1980 “ Interfirm tender offers and the market for corporate control” Journal
of business Vol.53, No.4, pp. 345 –376.

Bradley. M, A.Desai and E.H. Kim, 1988 “Synergistic gains from corporate acquisitions
and their division between the stockholders of target and acquiring firms” Journal of
Financial Economics Vol.21., pp. 3-40.

Breen. W, L.S. Hodrick and R.Korajczyk, 1999 “ The determinants of equity illiquidity,
Working paper, Northwestern University

Brown, Keith and Raymond, Mitchael, 1986 “ Risk Arbitrage and the Prediction of the
successful Corporate Takeovers” Financial Management, Autumn, pp.54.

Chang.S and D.Suk 1998 “Failed takeover, method of payment and bidder returns” The
Financial Review Vol.33, pp.77-84.

Choe. H, R.W Masulis and V. Nanda, 1993 “ Common stock offerings across the
business cycle” , Journal of empirical finance Vol.1, pp. 21-33.

Cosh, Andy, 1990 “ Predicting Success; Pre-Merger Characteristics and Post-Merger
Performance” Working paper 6 of University of Cambridge Small Business Research
Centre. pp. 24, August.

Cosh,Andy, Hugh,Alan and Kambhampati, Uma, 1993 “ Takeover Success or Failure?
The Experience of Large and Small U.K.Companies” Working paper 10 of University of
Cambridge Small Business Research Centre. pp. 10, August

Cornelli, F and Li, D, 1998 “ Risk Arbitrage in Takeovers” Rodney L. White Center for
Financial Research. The Wharton School, University of Pennsylvania

Dodd. P and R. Ruback, 1977 “ Tender offers and stockholder returns” Journal of
Financial Economics Vol(5) No 4, pp. 351-373.

Dodd. P , 1980, “ Merger proposals, Management discretion and Stockholder Wealth”
Journal of Financial Economics. Vol. 8 No 2, pp.105 –138.

Duke, William, Cheryl, Frohlich and Ma, Christopher, 1992 “ Risk Arbitrage in Tender
offers” Journal of Portfolio Management, Vol. 18, pp. 47-55.

Ebeid, Fred Joseph “ The Inetr-firm corporate cash tender offer: operating, marker and
bid characteristics of target firms. Ph.D Dissertation. University of Illinois at Urbana-
Champaign (1974)

Fung. W and D. Hsieh, 1997, “ Empirical Characteristics of Dynamic Trading Strategies:
The case of hedge funds” Review of financial studies, Vol.10, pp. 275-302.

Fung, William and David Hsieh, 1999 “ A Risk neutral approach to valuing trend
following strategies” Review of Financial Review.

Glosten, L and R, Jagannathan, 1994 “ A contingent claim approach to performance
evaluation” Journal of Empirical Finance, Vol.1, pp. 133-160.

Hansen. R, 1987 “ A theoty for the choice of exchange medium in mergers and
acquisition” Journal of Business. Vol. 60, pp. 75 –95.

Hoffmeister, J. Ronard and Edward A Dyl, 1980 “ Predicting outcomes of cash tender
offers” Financial Management. Vol.9, pp. 50-58.

Jennings R H and M A Mazeo 1993 “ Competing Bids, Target Management Resistance
and the structure of Takeover Bids” Review of Financial Studies Vol 6 (4), pp. 883-909.

Jensen. M and W. Meckling, 1976 “ The theory of firm: Managerial Behavior, Agency
costs and Ownership structure” Journal of Financial Economics. (October), pp. 305-360.

Jensen. M, 1986 “ Agency costs of free cash flow, corporate finance and market for
takeovers, American Economic Review Vol. 76, pp. 323-329.

Jindra, Jan and Walkling Ralph, 1998 “ Arbitrage Spreads and the Market Pricing of
Proposed Acquisitions”, Working paper, The Ohio State University

Jung. K, Y. Kim and R. Stulz, 1995 “ Investment opportunities, managerial discretion
and the security issue decision, Working paper. The Ohio State University

Kalay Avner and Uri Loewenstein , 1985 “ Predictable Events and Excess Returns: The
Case of Dividend Payment Announcements” Journal of Financial Economics, Vol 14 pp.

Karolyi G A and J Shannon, 1998 “ Where’s the Risk Arbitrage?” Working paper, The
Ohio State University.

Larker, David and Lys, Thomas, 1987 “An empirical analysis of the incentives to engage
in costly information acquisition: The case of risk arbitrage” The Journal of Financial
Economics Vol. 18, pp.111-126.

Leland Hayne, 1999 “ Beyond mean-variance: Performance measurement in a
nonsymmetrical world” Financial Analysts Journal. Jan/Feb

Malkiel. B , 1995 “ Returns from investing in equity mutual funds 1971 to 1991” Journal
of Finance Vol.L(50) No(2), pp. 549 – 572.

March. P, 1982 “ The choice between equity and debt: An empirical study” Journal of
Finance. Vol. 37, pp. 121-144.

Martin. K , 1996 “ the method of payment in corporate acquisitions, investment
opportunities and managerial ownership” Journal of Finance Vol.51 No.4 , pp. 1227 –

Mitchell, Mark and Pulvino, Todd, 2000 “ Characteristics of Risk in Risk Arbitrage ”
Working paper

Moore, Keith M, 1999 “Risk Arbitrage: An Investor’s Guide” Wiley

Myer. S, 1984 “ The capital structure puzzle” Journal of Finance Vol. 39, pp. 575 –592.

Myer. S and N. Majiluf, 1984 “ Corporate financing and investment decisions when firms
have information that investor doesn’t have , Journal of Financial Economics Vol.13.,
pp.183 –221.

Pelligrino, Joseph Charles, 1972 “ Causes of Inter-Firm tender offers: An Empirical
study, 1962-1968,” Ph.D Dissertation, Northwestern University

Raad E, R. Ryan and J.F.Sinkey, 1999 “ Leverage , Ownership Structure and Returns to
Shareholders of Target and Bidding firms” Quarterly Journal of Business and Economics,
Vol 38. No.2, pp. 37 –53.

Roll.R , 1986 “ The hubris hypothesis of corporate takeovers” Journal of Business.
Vol59, pp. 197-216.

Samuelson, William and Rosenthal, Leonard, 1986 “ Price Movements as indicators of
tender offer success” The Journal of Finance, Vol. 41, pp. 481-510.

Sanders. Ralph W and John S. Zdanowicz, 1992 “ Target firm abnormal rerturns and
trading volume around the initiation of change in control transactions” Journal of
Financial And Quantitative Analysis Vol 27 No.1

Shleifer, Andrei and Vishny, Robert, 1997 “ The Limit of Arbitrage,” Journal of Finance,
Vol. 52, pp. 35-55.

Sullivan. M.J, M.R. Jensen and C. D Hudson, 1994 “ The Role of Medium of Exchange
in Merger Offers: Examination of Terminated Merger Proposals” Financial Management,
Vol. 23, No(3), pp. 51-62.

Song.M and Ralph. A Walkling, 1993 “ The impact of Managerial Ownership on
Acquisition Attempts and Target Shareholder Wealth” Journal of Financial and
Quantatitive Analysis Vol 28 No.4.

Stulz R M, R A Walkling and M H Song, 1990 “ The distribution of Target Ownership
and Division of Gains in Successful Takeovers” Journal of Finance. Vol 45, pp. 817 –

Suk. D.Y and Sung H. M, 1997 “ The effect of the method of payment and the type of
offer on target returns in merger and tender offer” Financial Review Vol.32 No.3, pp.

Travlos.N.G, 1987 “ Corporate takeover bids, method of payment and bidding firms’
stock return,” Journal of Finance, Vol. XLII, No4, pp. 943 – 963.

Walkling Ralph, 1985 “ Predicting tender offer success: A Logistic Analysis” Journal of
Financial and Quantitative Analysis, Vol. 20(4), pp. 461-478.

Wansley. J.W, William. R.L and H.C Yang, 1983 “Abnormal returns to acquired firms by
type of acquisition and method of payment” Financial Management Vol.12, pp. 16-22.

                                           Appendix 1

                         Summary Table for Literature Reviewed

a) On the Abnormal Returns During a Merger Period
1. Synergy hypothesis.

Authors           Subject                 Data, Model and tested              Results & Supporting
                                          Hypotheses                          Hypothesis.
Asquith [1983]    To test abnormal        1).211 targets and 196 bidders in   1).CER (cumulative excess
                  returns around an       successful mergers and 91           return) during a period from one
                  announcement date.      targets and 89 bidders in           day after the press day until two
                                          unsuccessful mergers during         days before the outcome day.
                                          1962 to 1976
                                                                              - Successful target: Positive
                                          2).Market model (daily).
                                                                              - Successful bidder:
                                                                              Insignificantly Negative.

                                                                              - Unsuccessful target: Negative

                                                                              - Unsuccessful bidder: Negative.

                                                                              2) Supporting the synergy
Bradley, Desai    To the information      1). 241 successful targets, 112     1). Abnormal returns of target
and Kim [1983]    and synergy             unsuccessful targets and 94         and bidders on the
                  hypotheses to explain   unsuccessful bidders during         announcement date of
                  the wealth effect of    1963 to 1980                        termination.
                  unsuccessful tender
                  offers                  2). Market model (monthly)          - Target: Continuously positive.

                                          3) Synergy and information          - Target without the subsequent
                                          hypothese.                          offer: Decreasing abnormal

                                                                              - Bidder lost competition:

                                                                              2).Supporting synergy
Bhagat,           To test the wealth      1).295 interfirm cash tenders       1).Abnormal returns
Brickley and      effect for interfirm    offers during 1962 to 1980
Loewenstein       cash tender offer       Market model (daily)                - Target : Positive
                                          2).B/S option model                 2).Supporting synergy
                                                                              hypothesis, not wealth transfer
Bradly, Desai     To examine              1). 263 successful tender offer     1). The average synergistic gain
and Kim [1988]    synergistic gains to    contests during 1963 to 1984        in samples is $117 million
                  target and acquirer                                         dollar, 7.4% increase in the
                  stock holders           2). Market model                    combined wealth of the

                                                                                stockholders of the target and
                                           3). Dollar synergistic gain (-5 to   acquirer
                                                                                2) Supporting the synergy
                                           4) Weighted least square             hypothesis

                                           5). Variables: two periods,
                                           multiple bidders, fraction of
                                           target purchased by bidder

2. Taxation hypothesis in the payment method

Huang and          To test the influence   1).204 target firms during 1977      1).The cumulative abnormal
Walkling [1987]    of payment,             to 1982.                             return declines 1.8% over the
                   acquisition form and                                         50day post announcement
                   managerial resistance   2).Market model (daily)              period.
                   on the target
                   abnormal returns        3).Regression                        2).A tender offer generates
                                                                                higher yield than a merger
                                           4).Variables: tender offer,
                                           undisclosed, cash, mixed,            3).Cash offer generates higher
                                           neutral/undisclosed, resistance      return.
                                           and undisclosed
                                                                                4) Supporting the taxation
                                           5) Hypotheses: personal tax,         hypothesis to explain 3).
                                           account treatment,
                                           compensation effects, agency
                                           effect, regulation effect.

3. Asymmetric information hypothesis in the payment method

Travlos (1987)     To explore the          1).167 successful mergers and        1) Abnormal returns.
                   method of payment in    tender offers during 1972 to
                   explaining the          1981.                                - Bidder with stock payment:
                   abnormal return of                                           Negative
                   bidders                 2).Market model and
                                           multivariate regression              - Bidder with cash payment:
                                           3).Variables: payment method,
                                           bid premium, relative size of the    - Successful bidders with stock
                                           acquisition and type of              exchange: Negative during one
                                           acquisition                          or two days after an
                                           4) Hypotheses: asymmetric
                                           information, taxation and co-        2) Supporting the asymmetric
                                           insurance                            information hypothesis

Sullivan, Jensen   To examine the          1).84 targets and 123 bidding        1).CARs during a period
and Hudson         relationship between    firms during 1980 to 1988            between a day following the
[1994]             the medium of                                                merger announcement and two
                   exchange and            2).Market model                      trading days prior to the
                   valuation effect with                                        termination date.
                   terminated merger       3).Multivariate regression

                  proposal                                                     - Targets with cash offers have
                                           4).Variables: medium exchange,      higher CARs than those with
                                           occurrence of subsequent bid,       stock offers.
                                           the party terminating the bid,
                                           occurrence of acquisition           - Bidders with cash offers were
                                           program, foothold in the target,    found to have higher CARs than
                                           relative size of the acquisition    those with stock offers
                                           and presence of competing bid
                                                                               2).Stock payment is negatively
                                           5) Hypothesis: Synergy, taxation    related to the abnormal return
                                           and financing and investment.
                                                                               4)Supporting information
                                                                               hypothesis and synergy
Chang and Suk     To examine the           1). 279 failed takeovers during     1) CARs around the termination
[1998]            abnormal returns of      1982 to 1990                        date.
                  bidders around the
                  announcement of          2).Market model                     - Bidders with the stock
                  merger termination (-                                        payment: Positive.
                  1 to 0) and explaining   3).Weighted least square
                  hypotheses               regression                          - Bidders with cash payment:
                                           4)Hypothesis: information
                                                                               - Bidders initiating termination:
                                           5) Variables: Stock offer, Stock    Positive.
                                           offer*Bidder initiated
                                           termination, Stock*Target           2) Supporting the asymmetric
                                           initiated termination, Stock        information hypothesis.
                                           offer*other termination, Merger,
                                           Log(target size/bidder
                                           size),Multiple bidder, Hostile
                                           bidder, Bidder managerial
                                           ownership, long term
                                           debt/market value of equity,
                                           log(market value of equity)

4. Ownership hypothesis in the payment method

Stulz, Walkling   To test relationship     1). 104 successful tender offers    1).If multiple bidders exist, the
and Song          between the              during 1968 to 1986.                target gain increases with target
[1990]            distribution of target                                       managerial ownership and
                  ownership and            2).CAR/Synergistic gains with a     decrease with institutional
                  division of the          market model                        ownership.
                  takeover gains
                                           3).Multivariate regression          2).The gains to value weighted
                                                                               portfolio of the bidder and target
                                           4).Variables: market value of the have nothing to do with the
                                           target, market value of the bidder, target ownership distribution.
                                           bidder acquiring percentage,
                                           managerial ownership, institutiona 3) Supporting the ownership
                                           ownership, bidder ownership and hypothesis
                                           total gains

                                           5) Hypothesis: Ownership

Song and          To examine the            1).112 firms during 1977 to           1).The cumulative abnormal
Walkling          impact of managerial      1986.                                 return (-5 to +5) to the target is
[1993].           ownership on target                                             higher in successful acquisition
                  shareholder’s wealth      2).Market model                       (29.5%) than in unsuccessful
                                                                                  2).The target shareholder returns
                                            4).Variables: Contested,              are positively and significantly
                                            successfully acquired,                related to managerial ownership
                                            contested*acquired, managerial        in contested but successful
                                            ownership,                            acquisitions
                                            ownership,                            3) Supporting the ownership
                                            successful*managerial                 hypothesis

                                            5) Hypothesis: Ownership
Raad, Ryan and    To test for the effects   1).81 targets and 81 bidders in       1).Total gains of target shares
Sinkey [1999]     of leverage and           successful takeovers during 1980      increase with leverage and
                  ownership structure       to 1990                               institutional ownership in the
                  in target firms on                                              target firms.
                  returns to                2).Market model (Weighted
                  shareholders of target    Average Excess Return –dollar         2).Debt ratios of target firms are
                  and bidders               gains)                                associated with positive
                                                                                  abnormal returns for target firms
                                            3). Multivariate regression           and negative abnormal returns
                                                                                  for bidders.
                                            4).Variables: total dollar gains to
                                            targets and bidders, debt ratio,      3).Institutional ownership in
                                            log(target size), log(managerial      targets has positive effects on
                                            ownership) and log(percentage         the excess dollar return.
                                            owned by financial institution)
                                                                                  4).Managerial ownership in
                                                                                  target firms has no effect on
                                                                                  such gains.

                                                                                  5).Size of target firm has
                                                                                  nothing to do with gains

                                                                                  6) Indirectly supporting
                                                                                  Ownership hypothesis.

5.Regulation (Mode) and Investment opportunities hypothesis in the payment method.

Martin [1996],    To examine the            1).846 corporate acquisitions         1).The higher the acquirer’s
examine the       underlying motives        during 1978 to 1988.                  growth opportunities, the more
underlying        for the payment                                                 likely the acquirer is to use stock
motives for the   method.                   2).A logistic regression analysis     to finance an acquisition.
method in                                   3) To test 7 hypotheses:              2) Only middle range of
acquisition                                 Investment opportunities, Risk        ownership is negatively related

                                            sharing, Control, Cash              to the probability of stock
                                            availability, Outside monitoring,   financing.
                                            Mode of acquisition and
                                            Business cycle hypotheses.          3)The likelihood of stock
                                                                                financing increases with higher
                                                                                pre-acquisition market and
                                                                                acquiring stock returns but
                                                                                decrease with higher cash
                                                                                availability, higher institutional
                                                                                shareholdings and block-

                                                                                4) Supporting the Mode and
                                                                                Investment hypotheses.

6. Competition hypothesis.

Suk and Sung      To re-examine the tax     1).205 successful tender offers     1).Cash offers yield higher
[1997].           hypothesis, the           during 1974 to 1987.                target return(33.9%) than stock
                  information                                                   exchange offers(19.6%).
                  asymmetry effect          2).Market model with CRSP
                  hypothesis and            equally weighted market index       2).Institutional ownership has
                  competition                                                   nothing to do with the abnormal
                  hypothesis,               3).Weighted least square            return in cash offers
                  incorporating the         regression
                  institutional                                                 3).Supporting the competition
                  ownership of the          4) Hypotheses: Tax, Asymmetry       expectation hypothesis, which
                  target firms to           information and Competition         suggests that the likelihood of
                  explain the target        hypotheses.                         future competition might be
                  cumulative abnormal                                           greater in tender offer than in
                  returns (-5 to 5)         5) Variables: cash payment,         mergers
                                            tender offer, two periods,
                                            institutional holding, net
                                            operating loss + unused
                                            investment + foreign tax credit.
                                            Stock payment and depreciation
                                            in targets

7. Others

Larcker and Lys   To calculate excess       1). 111 13-D filings during 1977    1). Average excess return for an
[1987]            return of firms owned     to 1983                             arbitrageur is 3.75% during a
                  by arbitrageurs                                               merger period.
                                            2).Market model (daily).
                                                                                2). Arbitrageur has superior
Fabozzi, Ferri,   To test failed targets’   1).21 targets in failed tender      1). The average weekly excess
and Tucker        weekly returns during     offers which don’t receive          returns of unsuccessful targets
[1988]            an announcement           another offer after the failure     during that period are not
                  date to the withdrawn                                         statistically different from zero
                  date                      2).Market model (daily)

Jennings and      To examine the            1). 647 acquisitions during 1979    1). The successful acquisitions
Mazeo [1993]      structure of initial      to 1987                             have much higher return to

                   takeover bids,                                                 targets than unsuccessful
                   competing bid and         2).Equally weighted average of       acquisition during an
                   management                the compounded returns               announcement date to the date
                   resistance.                                                    of outcome

                                                                                  2). Returns to targets on the
                                                                                  resisted offers are, regardless of
                                                                                  outcomes, higher than those on
                                                                                  the unresisted offers
Walker [2000]      To investigate the        1).278 acquisitions during 1980      1).Diversification strategy with
                   strategic objectives      to 1996                              potential overlap generated
                   and stock price                                                negative impact on the bidder
                   performance of            2).Cumulative market adjusted        stock price
                   acquiring firms           return (CAR) and matched firm
                   around the                adjusted return
                   announcement date (-
                   5 to +5)                  3) Strategies tested: Geographic
                                             expansion, production line,
                                             increase market share, vertical
                                             integration and diversification.

b) On Risk Arbitrage Returns

Duke et al.        To examine the 761                                             1) The average abnormal return
[1992]             cash tender offers                                             was 24.6% for 52.4 days,
                   during 1971 to 1985                                            corresponding annualized
                                                                                  abnormal return of 171%
Shleifer and       To examine                                                     1). The investors’ control over
Vishny [1997]      theoretically limits of                                        arbitrageurs’ liquidity produces
                   arbitrage                                                      limitations in taking advantage
                                                                                  of an arbitrage.
Karolyi and        To examine the            1) 1.37Canadian acquisitions         1) Risk arbitrage generates
Shannon [1998]     profitability of risk     with over $50 million during         4.78% in excess of the TSE 300
measured the       arbitrage and risk        1997 and multivariate                stock index during average 57
return of risk     measure.                  regression analysis                  days and annualized excess
arbitrage                                                                         return 33.9%.
without                                      2)Regression
dividends, using                                                                  2) The spread may be
the                                          3) Variables: market                 significantly related to target
                                             capitalization (logarithm), beta,    size and pre-announcement run-
                                             price to sale ratio, price to book   up (2 week)
                                             ratio, oil & gas dummy, days to
                                             close, cash offer dummy, pre-        3) The return has nothing to do
                                             announcement run-up(2 weeks)         with the likelihood of success
                                             and pre-announcement run-up(1        (days to close), target size, beta,
                                             year)                                price to sale ratio, price to book
                                                                                  ratio, payment method, pre-
                                                                                  announcement run-up or
                                                                                  industry sector
Jindra and         To test profitability     1)362 cash tender offers in          1) An arbitrageur was found to
Walkling [1998]    of risk arbitrage and     which a bidder looks for 100%        earn an excess return of 1.42 to
measured the       determinants to the       of target shares and the             1.54% or an annualized return of

excess return of   initial spread        transaction value exceeds $10      102 to 115%, if she or he takes a
risk arbitrage                           million during 1981 to 1995        long position on a target during
without                                                                     one week following the
dividends, using                         2) Multivariate regression         announcement.
a sample of
                                         3) Variables: Log (market          2) An arbitrageur buying the
                                         capitalization), ownership,        target and shorting the acquirer
                                         friendly acquisition, hostile      during one week following the
                                         acquisition, blockholders,         announcement earned an excess
                                         premium, toeholders, attitude,     return of 1.88% or the
                                         rumor, run-up, abnormal volume     corresponding annualized return
                                         growth rate, option on target,     of 156%.
                                         previous acquisition,
                                         experienced bidder and multi-      3) The arbitrage spread8 is found
                                         financial advisors                 to be positively related to the
                                                                            size of bid premium, friendly
                                                                            managerial attitude about the
                                                                            offer and existence of rumors
                                                                            about the offer but negatively
                                                                            related to the target size and pre-
                                                                            offer run up.

                                                                            4) The managerial ownership is
                                                                            found to be only significant in
                                                                            hostile takeover and bidder’s
                                                                            acquisition experience to be

                                                                            5) The spread is positively
                                                                            related to the duration of the
                                                                            offer and negatively related to
                                                                            the revision of the offer.
Mitchell and       To examine the risk   1).4750 stock swap mergers,        1).Risk arbitrage generated the
Pulvino [2000]     of risk arbitrage,    cash mergers and cash tender       annual excess return of 4%.
                   using a contingent    offers during 1963 to 1998 and a
                   approach.             contingent approach                2).Return is correlated with
                                                                            market return only during
                                         2).Piecewise linear regression     market downs


    The locked-up initial spread refers to the price difference between the offered price and the actual market
price on the announcement date. Here, however, we start with the spread on the day after the
announcement .
   Taxation was researched in Dertouzos and Thorpe [1982]. They find that the increase of depreciable
assets tended to lead to the bidding competition.
   These hypotheses related to the payment method were well summarized in Martin [1996]. We shall
follow his hypothesis classification in this review article.
   They followed Lakonishok and Vermaelen [1990] and Schwert [1996] to examine the abnormal trading
volume for the targets around the tender offer announcements.
  It was defined as the percentage difference between the initial bid price and the target’s closing price after
the acquisition announcement
  Their transaction costs are calculated, using brokerage commissions and the price impact associated with
trading illiquid securities.
   Price impact model from Breen et al. [1999] was used.
  It was defined as the percentage difference between the initial bid price and the target’s closing price after
the acquisition announcement


To top