In case you haven’t come across it yet, I highly recommend the blog Overcoming Bias, where a number of academics write about epistemology. Given how we are bombarded with new factoids, studies, and commentary on a daily basis, it’s helpful to consider the views of those who think about the nature of knowledge and the limits of information more rigorously than most of us do.
One recent post that caught my attention was “I Defy the Data!” by Eliezer Yudkowsky. He argued that when an experimental result contradicts a standard model, the scientific community should have a way to challenge the experiment (by defying the data), rather than, say, engaging in character attacks on the researchers. He sees defying the data as both a recognition that the new findings are important, and a demand to either confirm or refute them.
Well, I’m going to defy some data, but I may not be as high minded about it as Yudkowsky.
Deal Blogs tells us that the Boston Consulting Group has produced groundbreaking research, touted as the biggest non-academic study of M&A ever done, that “shatters” myths.
Although I will confess to having read only the press release, I’m far from convinced. There’s some slippery logic involved in the development of these supposedly novel conclusions:
Although 58.3 percent of deals between 1992 and 2006 destroyed value for acquirers, with a net loss of 1.2 percent for all transactions, the average deal produced a net gain to shareholders of 1.8 percent when returns of the targets are taken into account. Moreover, the majority of deals (56 percent) created value for the combined set of shareholders. In addition, acquirers in several industries, including automotive and retail, created value, on average, as did acquirers in the Asia-Pacific region.
First, this study broadly confirms the findings of every academic ever done, namely, that most deals fail (failure defined as destroying value for the acquirer). The failure rate is typically between 60% and 75%, so BCG’s 58.3% is barely outside the normal range, which might be due to the selection of a timeframe whose starting point just happens to be at the beginning of the 90s bull market.
Second, the inclusion of returns to the target shareholders is specious. And even if it wasn’t, a mere 1.8% gain (over what time frame? It has to be several years, and I suspect the figure is expressed as gross gain, rather than on an annual returns basis) is almost certain to be inadequate given the disruption, risks, and management effort involved.
And the report sets up straw men:
It’s commonly assumed that PE firms have gained an increasingly large share of the M&A market by using their huge reserves of capital to pay over the top for targets. But BCG’s analysis indicates that, on average, PE firms pay lower multiples and lower acquisition premiums than “strategic” buyers.
Commonly assumed by whom? Most people I know recognize that a strategic buyer will usually bid more aggressively than a financial buyer. The strategic buyer has whatever value he ascribes to synergies on top of the straightforward financial returns that a PE pro considers. The only time a PE firm can be expected to bid up a property beyond its stand alone economic value is in a consolidation play. Then it has considerations similar to a corporate buyer.
The reason corporate buyers have been comparatively quiet is that companies have become acutely risk averse. Cutting costs, outsourcing, and buying back stock have seemed more attractive than doing deals.
Of course, to learn all you really need to know about the study, consider this section of the press release:
“We are seeing a return to normalcy, which is healthy,” said Jeff Gell, a Chicago-based partner and coauthor of the report, upon its release. “Prices and leverage will come down slightly, but volumes will remain high as the strategic need for most deals is still present. Companies are still sitting on excess cash that they need to deploy, and private equity funds still have large war chests that they need to put to work.”
Interesting how one man’s normalcy is another’s twenty-five standard deviation event.