There are probably some things that bankers could advise companies to do that are unequivocally bad. Obviously if I were Bank X’s Executive Director and Global Head of Lighting Money on Fire, and I went around showing companies a pitch book that was all “signalling benefits of lighting money on fire,” and I got a bunch of companies to do it, and it became a thing, and academics and industry groups did studies on it, I suspect that they would consistently report that the trade was NPV negative at a 1% level of significance. But maybe not, because, industry groups. Anyway though you can probably imagine some of these things existing outside of silly stylized examples – in hindsight, CDOs of mezz ABS look pretty close to lighting money on fire – but not too many of them. Because if a thing is always bad through the cycle then you’d quickly run out of people to do it, for some value of “quickly.”
Is it possible that mergers are such a thing?
No, it is not!
There, that was easy. Nor, obviously, are they the sort of thing that is unequivocally good – I suppose there are such things?* Instead, mergers are like, I dunno, hedge funds. You can ask the specific question of “will this merger be good for this company?,” and the answer will mostly be “maybe” but said with DCFs and stuff. Or you can ask the general question of “are mergers mostly good or mostly bad?” and the answer will be entertainingly indeterminate. Thus mergers are unequivocally good for academics, QED. Anyway this is a strange paper:
Do acquirors profit from acquisitions, or do acquiring CEOs overbid and destroy shareholder value? We present a novel approach to estimating the long-run abnormal returns to mergers exploiting detailed data on merger contests. In the sample of close bidding contests, we use the loser’s post-merger performance to construct the counterfactual performance of the winner had he not won the contest. We find that bidder returns are closely aligned in the years before the contest, but diverge afterwards: Winners underperform losers by 50 percent over the following three years.
It’s by two Berkeley professors (and one Amsterdam professor) and Stephen Gandel describes it somewhat alarmingly here. They use the close bidding contests so that they can compare the performance of two otherwise similar companies – the winner and the loser, who apparently tend to have similar stock performance leading up to the deal – and isolate the effect of the deal. And then they limit it to the quarter of deals in their sample where the bidding contest lasts more than a year because their results look better that way.** Specifically they look like this:
There are reasons to doubt the representativeness of that sample; long-drawn-out contested acquisitions are, you would think, exactly the acquisitions where (1) the acquiror is most likely to overpay (because there’s an auction), (2) the strategic fit is less obvious (because if Company X and Company Y are just made for each other, why is Company N bidding?), and (3) by the end of the deal everyone has forgotten how to run the business during their year-long sojourn into M&A strategy. The paper has a response to the overpayment concern, which is … there, anyway,*** but not the others, and all in all you are left with the impression that “winning a long-drawn-out bidding war for a company that everyone wants is likely to work out badly for you” which is not QUITE the same thing as “mergers = bad.” Not that the authors claim that it is. I dunno. Their explanation, if you care, is not operational – “The observed stock underperformance does not translate into operating underperformance” is sort of a funny thing to say – but rather capital structure; that is, acquirers lever up more than non-acquirers and that seems to go poorly. Why not.
So, TLDR: if you look at the eighty contested M&A deals that they’ve identified in the last 25 years, and then chop off sixty of them because they were contested for less than a year, you get twenty deals where, on average, the buyer performed worse than the losing bidder. Yaaaay.
This paper did not convince me to abandon my quest for a merger target for Dealbreaker, but it did bring me a sense of efficient-markets contentment. Bankers spend their days cutting the data in exactly this way, to provide evidence that everyone should do more mergers.**** Lopping off three quarters of the deals because they don’t prove your point – sorry, because those deals aren’t “merger contests where, ex ante, both bidders have a signi cant chance to win the contest” – is stereotypical investment banking behavior. But, in this case, from the other direction: the deals are being lopped off to prove that mergers are bad, not good. In general, you could have a theoretical worry that, with bankers having every incentive to push mergers, and no one having any incentive to push against mergers, there would be Too Many Mergers. So it’s nice to see league-table-dicing strategies coming from the other side.
* But they’re even harder to imagine than the unequivocally bad. No arbitrages etc.
** I … I actually think that’s what they’re doing:
We also find that the winner’s underperformance is observed only in the long-duration quartile [i.e. the 25% of deals where the contest lasted more than a year]. Auxiliary plots of market-adjusted CARs for the other duration-quartiles … show little post-merger divergence in the middle quartiles, Q2 and Q3 [contests lasting five to twelve months]. And, in the shortest-duration quartile, Q1 [deals that close in under five months, i.e. basically uncontested], both winners and losers display abnormal under-performance, with losers performing even worse than winners. Given the lack of winner-loser comparability in quartiles Q1-Q3, the latter results are hard to interpret.
So they discard 75% of deals because (1) the acquirers and non-acquirers aren’t that similar to each other before the merger and (2) the acquirers don’t actually underperform the non-acquirers after the merger. This is amazing statistical sleight of hand!
*** Basically: “long-drawn-out contested deals don’t have that much higher premia than shorter contested deals.” But why is that the question?
**** Though I’m pretty sure that no banker would recommend ex ante getting involved in a year-long bidding war, because (1) that seems intuitively bad and bankers are usually trying to give good advice in an admittedly imperfect world, and (2) bankers don’t bill by the hour.