Mergers are among the largest and most disruptive events in a corporation’s lifetime. The proper assessment of their value implications has been of foremost interest to policy-makers and academic researchers alike. Much of the research on mergers and acquisitions aims to assess which transactions create, or destroy, how much shareholder value, including a recent debate about “massive wealth destruction” through mergers (Moeller et al. (2005)).
Empirically, the measurement of the causal effect of mergers is challenging. The standard approach in the literature is to use stock-market reactions to merger announcements and to interpret the combined change in target and acquirer values as the expected total value created. This approach builds on a number of assumptions, including the assumptions that markets are efficient, that mergers are unanticipated and unlikely to fail, and that merger bids reveal little about the stand-alone values of the merging entities. Various studies document a small positive combined announcement return of targets and bidders, and interpret this finding as evidence in favor of value creation.
In our recent NBER working paper, Cash Is King — Revaluation after Merger Bids, my co-authors (Marcus Opp of UC Berkeley and Farzad Saidi of New York University) and I argue that a large portion of the announcement effect reflects target revaluation rather than value created through mergers, and that this portion varies with the type of payment: Targets of cash offers are revalued by +15%, but there is no revaluation of stock targets. We also find significant negative revaluation effects for stock bidders, but no effect for cash bidders. Our results imply that the widespread use of announcement effects significantly distorts the assessment of mergers.
Our empirical analysis is motivated by the existing body of theories that predict that acquirers’ private information is reflected in their bids and their choice of payment. While the negative information effect of stock bids on the acquirer’s stand-alone value is well understood at least since Myers and Majluf (1984), we focus on information revelation about the target. Our identification strategy exploits deal failures to measure information revelation: we compare target values prior to the announcement and after failure of a bid. To see the intuition for our approach consider the following stylized example: A target currently trades at $100 and receives a cash offer of $125, increasing its price to $125 per share. Shortly afterwards, the deal fails for exogenous reasons, and the target then trades at $115. After deal failure, the information the bid revealed about the target’s stand-alone value is no longer confounded with the merger effect. Hence, we can attribute $15 of the original $25 announcement effect to the revision of beliefs about the stand-alone value of the target.
We collect a new data set of all unsuccessful merger bids in the US between 1980 and 2008. We document initial announcement effects of 15% for stock deals and 25% for cash deals (including the 25-trading days run-up). We then show that, by the time of deal failure, the value of stock targets falls below the pre-announcement level while the value of cash targets remain at 15% cumulative abnormal returns, relative to the pre-announcement level. All results hold controlling for a host of deal- and firm-level characteristics, including the target size, relative deal size, offer premium, and hostility of the bid. The magnitude of our estimates implies that, for the average cash deal completed during our sample period, approximately $78m (in 2010 dollars) should be attributed to target revaluation rather than the merger.
We also show that, consistent with previous studies, stock acquirers trade at significantly lower prices post-failure (-17%) while cash acquirers return to their pre-announcement level. Taken together, our findings suggest that cash bids are “all about the target,” and indicative of prior target undervaluation, while stock bids are “all about the acquirer,” and indicate prior acquirer overvaluation.
An important issue for the interpretation of our results is sample selection: deal failure is generally not exogenous to target and bidder valuations. To address this endogeneity concern, we seek to identify the subsample of deals where failure was exogenous to the stand-alone value of the target (for the analysis of target revaluation), and the subsample where failure was exogenous to the stand-alone value of the acquirer (for the analysis of acquirer revaluation). We perform an extensive news search for the failure reasons for all observations in our sample. For the target analysis, we isolate all regulatory interventions and negative shocks to the bidder that can be reasonably interpreted as exogenous to the target’s stand-alone value. For the bidder analysis, we isolate negative shocks to the target, such as the uncovering of negative information in the due-diligence process, that are exogenous to the bidder’s value. We then replicate and confirm our results on the respective subsamples. In addition, we find that, for every failure category other than market-wide shocks (“September 11,” “October 87,” etc.), we obtain very similar estimates (though statistically weaker, given the small subsamples). This indicates that our result is robust to alternative classifications of exogenous failure reasons.
The robustness of the revaluation effect across categories is also consistent with an auxiliary analysis on the determinants of failure: We show that the choice of cash versus stock payment does not predict failure, whereas other bid and firm characteristics, such as hostility of the bid or relative deal size, do predict deal failure. In other words, within the (partly self-selected) sample of failed bids, there is no additional sorting by means of payment. Taken together, the robustness of our results in the exogenous- and the endogenous-failure samples as well as the finding that cash payment does not predict failure are most easily explained if there is no differential bias in the cash-stock difference in revaluation between successful and unsuccessful bids.
We also show that the revaluation effect is not explained by more lucrative future takeover offers: Failed cash deals are no more likely than failed stock deals to be followed by another takeover attempt in subsequent years, nor are future offer premia higher.
Why, then, do cash bids induce the market to positively revalue the target? One explanation is that the acquirer’s cash bid reveals positive private information about the target. A related explanation is limited attention: a cash bid draws investors’ attention to the target and induces them to process information that was already available. In both cases, the market learns about the target’s value from the bidder’s action, and we will dub both interpretations information revelation in a broad sense.
To ease the interpretation of our empirical results, we develop a theoretical framework, to show that the data can be understood through one simple learning channel.Our model incorporates rational market updating into the mispricing framework of Shleifer and Vishny (2003). The market updates its beliefs about the stand-alone values of the bidder and target upon observing an offer. Due to the presence of uninformed bidders, the updating is imperfect, and informed bidders can benefit from their information. Informed bidders make cash offers when the combined entity is undervalued as to prevent target shareholders from participating in the long-run price appreciation. They make stock offers when the combined entity is overvalued as to share the long-run price decline with target shareholders.
In this framework, rational market updating predicts that firms involved in cash deals are revalued upwards relative to firms engaging in stock deals, for both the target and the bidder. On an absolute scale, a cash offer is always positive news about the stand-alone value of the target, and a stock offer is always negative news about the bidder. Interestingly, and consistent with our empirical findings, a stock offer does not necessarily reveal positive information about the target — an informed bidder would have chosen cash if his information about the target had been sufficiently positive.
Our findings have significant implications for the welfare assessment of mergers, i.e., for the question whether mergers create surplus, as in the Q-theory model of Jovanovic and Rousseau (2002), or whether they merely have distributional consequences, as in the misvaluation-driven model of Shleifer and Vishny (2003). Our findings imply that, while stock deals may be associated with declines in stock-market capitalization, a large fraction is due to revaluation rather than the merger. Hence, naive announcement-effect based estimates of the causal effect of stock-financed mergers are downward biased. In contrast, stock-price increases at the announcement of cash deals reflect, to a large extent, target revaluation rather than surplus creation.
The full paper is available for download here.