We talk about M&A performance as if it was a single process with a good or a bad outcome, but in fact it consists of multiple processes, all of which must function effectively for a successful deal
When I teach corporate strategy to MBA students, I often notice that whatever else they know or don’t know about corporate strategy before taking my class, they seem to know that Mergers and Acquisitions (M&A) have very high failure rates. How high? As high as 60-70 per cent, some estimates say. Yet, all one has to do is to look at the M&A news, and it is immediately clear that M&A activity shows no signs of disappearing from the corporate strategists toolkit. True there are merger waves and dry seasons, but nobody in their right mind would argue that M&A deals are heading towards extinction. So how does one reconcile the high failure rates with the evergreen popularity of M&A? Many leap to the conclusion that the issue is one of poor management: Either managers are conducting deals that benefit them privately but not their shareholders, or they are overconfident about their ability to make these deals work. I want to offer some alternative views here.
First, I think we should treat the data about failure with some healthy skepticism. Measuring M&A performance is notoriously difficult because of the challenge of bounding the impact in scope and time. The narrower the measure of performance (e.g. the performance of the acquired unit rather than combined entity), the lesser the contamination of other factors, but equally it is less likely to capture systemic effects. For instance, if we were to look only at the performance of an acquired division after the acquisition, we would be missing out on the consequences for other divisions in the acquirer; indeed poor performance in the acquired unit may be more than offset by good performance in other units because of the merger. Similarly, looking at short-term performance metrics (such as the reactions of the stock markets on the day of the deal announcement- a very popular measure in academic studies) can localize the effects of the deal, but cannot capture precisely the consequences that take time to unfold; yet looking at long-term performance of shares of the acquired companies confounds the impact of the deal with many other possible occurrences.
So how does one go about measuring M&A performance then? In my research and consulting work, my approach has been to a) link performance measures directly to what motivated the deal – access to products, technologies, markets etc. For instance when studying technology acquisitions, it may make sense to use data on product launches and patents. Put differently, the measurement of performance ideally begins even before the deal is done, with a clear statement of intent. A corollary is that retrospective exercises to assess performance of deals done several years in the past will yield at best noisy, at worst politically biased conclusions. b) use multiple measures where possible –studies find modest but positive correlations between accounting, financial and product based measures and c) use the same measure at different time scales (e.g. immediate market reactions, one month, one year later etc.),. Needless to say, the vast majority of reports and articles that claim the 60-70 per cent failure rates are unlikely to be this careful about measuring performance.
Second, even if it were true that as many as half of M&A deals are failing, this does not mean that on an expected value basis these deals are not worth doing. For all we know those that succeed may be creating far more value than that destroyed by those that are failing. We don’t really have good long term studies that can tell us about this.
That said, I should state that even with all the issues with the data I still do believe that M&A baseline success rates in general are unlikely to be very high. This is because I think we may have a “resolution” problem when talking about M&A performance; we may be using a telescope to look at something we should be using a microscope for. Let me explain this in a bit of detail: To conduct a merger or acquisition, there are at least five distinct stages: Target selection, due diligence, valuation, negotiation and post-merger integration. Let’s say you are very good at target selection, you get it right 90 per cent of the time; similarly, you have a fantastic due diligence team who gets it right 90 per cent of the time, and equally competent valuation and negotiation as well as post-merger integration teams, who each get it right 90 per cent of the time. Does this mean you have a 90 per cent success rate with M&A? Unfortunately no, because the joint probability of getting all these five stages right is actually less than 60 per cent (0.9 to the power 5). My point is that we talk about M&A performance as if it was a single process with a good or a bad outcome, but in fact it consists of multiple processes, all of which must function effectively for a successful deal. We shouldn’t really be expecting very high baseline success rates for such complex, multiprocess phenomena.
That brings me to my last point: If not M&A, what else? M&A is but one of the three ways in which corporations pursue new opportunities, organic growth and strategic alliances being the other two. Unfortunately, the data on the success rates of alliance and organic growth doesn’t paint a very pretty picture either; both commonly are reported to have failure rates above 50 per cent as well! Life for corporate strategists, it appears, is not easy…particularly if they want to be entrepreneurial and seek new opportunities for value creation. They shouldn’t let the mythology of M&A failure discourage them.