…dead right about this. At XXX Firm, the Financial Restructuring Group was part of the M&A Group for a number of years because I had done some work for the M&A guys very early on and they liked me. In consequence, for those years I had a front row seat to a lot of M&A activity. It amazed me how the M&A guys would toss together a proforma combination analysis of two companies (without having ANY in depth comprehension of either company’s business), convince the CEO(s) to do the deal, then arrogantly extend their hand and demand 10s of millions of dollars in fees. For what? What value did they add? How has society improved? How has commerce been made more efficient? How is anyone better off?

I am not sure that my reading of the literature is quite the same. My understanding of the empirical literature is that, if anything, the event-study results suggest a “slightly” negative effect of acquisitions on returns (in some cases insignificant in others statistically significant but weak in statistical magnitude). But then there are the studies of firm performance that demonstrate reduced ex post profitability and increased likelihood of ex post divestiture.

I actually even remember a study in SMJ (can’t recall the authors right now) that demonstrates that deals in which advisers were not involved performed significantly better than those in which advisers were involved. Consistent with your latter points about the “typical” deal.

The empirical basis for saying that acquisitions “fail” is that the stock market reaction to the announcement of acquisitions is, on average, indistinguishable from zero. Sometimes the market likes the deal and the stock price goes up, sometimes it doesn’t and the stock price goes down — which averages out to zero. Imagine the counterfactual in which 100% of acquisition annoucements were greeted with a positive market bump. Everyone would see that acquisitions are surefire winners and would rush to do more of them. They would do the best deals first and then engage in increasingly marginal transactions until the average mkt annoucement effect was… what? Zero? There are many different ways one could model this but I think there are many plausible models that result in an average zero expected value for acquirers without having to impugn anyone’s motives.

That said, we ought to be able to do better than a zero average. Saying that the _marginal_ deal should have a zero expected value is different from saying the _average_ deal should have a zero expected value. The only way for rational, share-holder value-maximizing managers to get an _average_ zero return on their acquisitions is for the managers to be really bad at distinguishing good and bad deals. If the distribution of possible deals has a zero mean and they choose from targets from that distribution at random like numbers from a bingo cage and can’t throw out bad draws with negative expected values, then you wind up where we are now.

Perhaps I should point out though that others in the marketplace aren’t doing any better. The average mutual fund underperforms indexes. The average motion picture barely breaks even. etc. I’m sure you would find that the average new product launch fails as well. That’s life in a highly competitive environment.

But even if you believe managers are doing deals at random, they still aren’t doing harm when they do these deals. If the expected value to aquirer shareholders is zero, but target shareholders get a big premium, then a diversified shareholder is actually better off because of the M&A deal.

Finally, it’s important to note that the typical deal is not one you read about in the WSJ. Only unusual deals are newsworthy. According to data from Thomson Financial’s SDC, the typical deal is under $200M and is done without outside advisers on either side.