The Winner's Curse

The 'Winner's Curse' Hypothesis and Corporate Takeovers Hypothesis  The winner’s curse hypothesis states that in any bidding bidding situation, a party which ‘unknowingly’ overestimates the value of an asset tends to bid higher than its competitors, and thus win the auction. In case of takeovers, the magnitude of curse is the dierence between the bid premium of the winner and the maximum oerable premium conditional on the capital market’s expectation. In the paper The !"inner!s #urse! $ypothesis and #orporate Takeovers% Takeovers% by &ikhil '. (araiya published in the Managerial and Decision Economics, the author relates the magnitude of curse with) *ivergence of opinion amongst ac+uirer with respect to sie of takeover takeover gains *egree of competition for control of target -rm, and 'reac+uisition pro-tability of the winning bidder •

• •

 The main testable hypothesis that that emerge from the paper are are thus $o/) The winning 0uyer in a corporate takeover auction will, on average, pay more than the value of expected takeover gains. $12) The likelihood and magnitude of the winner!s curse, i.e. overpayment, will increase with an increase in the divergence of opinion among buyers with respect to the value of expected takeover gains $13) The likelihood and magnitude of the winner!s curse, i.e. overpayment, will increase with an increase in the degree of competition among 0uyers for control of 4eller. $15) The likelihood and magnitude of the winner!s curse 6i.e. overpayment7 is positively related to the magnitude of the preac+uisition abnormal performance of the winning 0uyer •

•

•

•

Methodology 8 sample of ac+uisitions was taken from a mergers and ac+uisitions database provided by 9idder 'eabody : #o. The time period in consideration is /;<5=3 and the value of the deal exceeded >/? million. The -nal sample set contained ;/ ac+uisitions 6?/ mergers and 51 tender oers7. The following -lters were used)  The ac+uisition was successful successful 6completed7, i.e. the identity of the winning 0uyer was known  The winning 0uyer -nanced the ac+uisition ac+uisition with an oer of cash, of common stock or a combination of cash and stock 6including preferred7 only  The ac+uisition was not a leveraged leveraged buyout@ and  The availability of suAcient information information enabled measurement of the announcement date, the winning bid premium, the degree of competition and the estimate of expected takeover gain in each ac+uisition. •

•

• •

Hypothesis 1) The magnitude was de-ned as dierence between observed premium and maximum oerable premium. The overall calculations were

inBuenced by seller’s excess return on the day of announcement and seller’s and buyer’s total e+uity value. Hypothesis 2) Cor divergence of opinion, a proxy was used based on variation in the D'4 estimate of the sample companies by the leading brokerage houses. Hypothesis 3) The degree of competition was measured as well as regression was run where 2 or more actualEpotential competitors were present. In case where there was only one bidder, a proxy was generated which would measure the increase in competition had there been more bidders. Hypothesis 4) The preac+uisition pro-tability and cash Bows of the winning bidder was measured for the /2 months preceding the deal. It was expected that these companies generated excess cash Bows and thus would fund their ac+uisitions through these.

Conclusion  The empirical results show that the magnitude of overpayment is positively related with all the determinants stated. In our sample set, F