Moeller, Schlingemann, and Stulz examine the returns to acquiring firms in takeovers. They find that "from 1991 to 2001 (the 1990s), acquiring firms' shareholders lost an aggregate $216 billion, or more than 50 times the $4 billion they lost from 1980 to 1990 (the 1980s), yet firms spent just 6 times as much on acquisitions in the latter period."
However, before you say that all takeovers are bad, there are several key points that can get buried in the headline. For instance, most firms did not lose money. The large losses seem to be concentrated in the 1998-2001 time period and further concentrated in a relatively few large deals.
In the author's words: "the average acquisition creates wealth for acquiring-firm shareholders. Instead, this large loss is caused by an increase in the size of the dollar losses of acquisitions with the worst dollar returns that is not offset by an equivalent increase in the size of the dollar gains of acquisitions with the best dollar returns. Statistically, the distribution of dollar returns in the late 1990s exhibits substantially more skewness compared to earlier years"
They continue: "Out of the 4,136 acquisitions from 1998 through 2001, 87 are large loss deals. The aggregate wealth loss associated with these acquisitions is $397 billion, while all other acquisitions made a total gain of $157 billion. The large loss deals represent only 2.1% of the 1998 to 2001 acquisitions, but they account for 43.4% of the money spent on acquisitions." This negative return is both large and significant: "The excess return of the large loss deal portfolio over the matching-firm portfolio is -39%."
Prior to making the bad acquisition, the acquiring firms had been a. active in the acquisition market and b. outperforming the market. Thus, "the evidence is consistent with the view of Jensen (2003) that high valuations give management more discretion, so that management can make poor acquisitions if it [management] values growth more than shareholder wealth." However, the authors do not feel this high valuation " is not sufficient to explain the change in returns associated with acquisition announcements, since these firms have comparable valuations when they announce previous mergers or acquisitions that are associated with positive abnormal returns." Rather it is hypothesized that "the acquisitions led
investors to reconsider the extremely high stand-alone valuations of the announcing firms."
Attempts to explain this negative returns using regression analysis are only marginally useful and can only explain about 20% of the negative abnormal return.
Moreover it does not appear to be due to diversification: "Large loss deals are more likely to be hostile and more likely to be tender offers than other transactions, but the fraction of large loss deals that are tender offers or hostile is small enough that these deal characteristics seem unlikely explanations. The acquisitions in our large loss deal sample are more likely to be within the acquirer's industry than are the other acquisitions, but the difference is not significant. The large loss deals cannot be attributed to diversification attempts."
To summarize: most deals are slightly positive or at least neutral, however many of the larger deals (especially in the later 1990s) appear to be losers and the size of these deals is enough to make the overall average return negative. This is consistent with Jensen's overvaluation problem.
Of course what can not be measured is how far the firms would have fallen had they not done the deal (see Shleifer and Vishny 2003--yeah I know this version is a 2001 version, but I cannot find a link to their 2003 JFE article.
This paper is forthcoming in the Journal of Finance:
A previous version of this is available through FEN.