Long-run post-merger stock performance of UK acquiring firms: a stochastic dominance perspective

Abstract

Using the idea of stochastic dominance, the long-run post-merger stock performance of UK acquiring firms is studied. Performance is compared by using the entire distribution of returns rather than only the mean as in traditional event studies. The main results are as follows: First, it is found that, in general, acquiring firms do not significantly underperform in three years after merger since no evidence of first- or second-order stochastic dominance relation between acquirer and benchmark portfolios is observed. Second, it is found that acquirers paying excessively large premiums are stochastically dominated by their benchmark portfolio implying that overpayment is a possible reason for post-merger underperformance. Consistent with previous studies, it is found that cash financed mergers outperform stock financed ones. Finally, no evidence is observed that glamour acquirers underperform value ones as no stochastic dominance relations between the two. In general, the results underline the importance of examining long-run post-merger stock performance from alternative perspectives.

Notes

A non-exhaustive list of US studies includes for example: Asquith (1983), Jensen and Ruback (1983), Malatesta (1983), Magenheim and Mueller (1988), Agrawal et al. (1992), Loderer and Martin (1992), Anderson and Mandelker (1993), Loughran and Vijh (1997), Rau and Vermaelen (1998), and Agrawal and Jaffe (2000). Examples using UK data are: Firth (1979), Franks and Harris (1989), Limmack (1991), Limmack and McGregor (1995), Kennedy and Limmack (1996), Gregory (1997), Chatterjee (2000), Aw and Chatterjee (2004).

Most standard finance textbooks (e.g., Huang and Litzenberger, 1988) include sections on the concept of stochastic dominance. However, one can see few empirical applications in recent finance literature. Some early exceptions include Porter and Gaumnitz (1972), Porter (1973), Joy and Porter (1974), Tehrenian (1980), and more recently Post (2003) among others. It is noted that the methodology in Post (2003) focuses on portfolio diversification issues by comparing a given portfolio to a set of assets. In the present study, only two return distributions are compared and therefore the study does not use the linear programming method in Post (2003). Comparisons of income, wealth and earning distributions using tests for stochastic dominance are however common in empirical economics (see for example Anderson, 1996 and Davidson and Duclos, 2000).

It is noted here that other studies, for example, Bradley and Jarrell (1988), Franks et al. (1991) do not find significant underperformance up to three years after the merger.

The earliest year for which UK M&A data is available in the Securities Data Corporation (SDC) database is 1985.

The literature is followed (see for example Fuller et al., 2002 and Moeller et al., 2004, 2005) and a one million dollars cut-off point is used so as to exclude very small deals.

Evidence shows that target firm share prices only change significantly at merger announcement date and the day before. Thus the use of one-month merger premium can reflect the true difference between offer price and target's normal price. See, for example, Dodd (1980), Asquith (1983), Dennis and McConnell (1986), Huang and Walkling (1987) and Bradley and Jarrell (1988). The one-month merger premium equals to the difference between the initial bid price and target market price four weeks before the initial merger announcement divided by the same target price four-week prior to the announcement.

The stochastic dominance approach allows for a more general framework than one that uses only the mean and the variance as measures of comparative risk since these imply that the utility function is quadratic or the distribution of payoffs is normal.

Several tests proposed earlier (for example Anderson, 1996 and Davidson and Duclos, 2000) compare the distribution functions only at a fixed number of arbitrarily chosen points. In general, comparisons using only a small number of arbitrarily chosen points will have low power if there is a violation of the inequality in the null hypothesis on some subinterval lying between the evaluation points used in the test.

Ho (2003) is a recent example of an application of stochastic dominance tests to evaluate IPO long-run performance.

Alternately, if one fails to reject in the second step that the merger portfolio stochastically dominates the benchmark portfolio but can reject in the first step that the benchmark portfolio stochastically dominates the merger portfolio, it can be concluded that there is a stochastic dominance relation of merger portfolio over the benchmark portfolio.

See for example, Barber and Lyon (1997), Kothari and Warner (1997), Fama (1998), Lyon et al. (1999), and Viswanathan and Wei (2004).

Hansen (1987), Fishman (1989), and Eckbo et al. (1988) also argue that the choice of method of payment in takeovers may be partially driven by agency cost considerations.

See for example, Wansley et al. (1983), Travlos (1987), Huang and Walkling (1987), Eckbo (1988), Eckbo et al. (1988), Asquith et al. (1988), Franks et al. (1988), Limmack and McGregor (1995), Loughran and Vijh (1997) and Rau and Vermaelen (1998).