Shai Dardashti, managing director of The Manual of Ideas, had the pleasure of sitting down for an interview with Michael Mauboussin, head of Global Financial Strategies at Credit Suisse.

The following transcript has been edited for space and clarity.

MOI Global: How do you think about survivorship bias [in multi-year studies] and how do you try to account for it?

Michael Mauboussin: It’s a really important topic. One of the researchers I like on this is a guy named Jerker Denrell, who’s a professor at University of Warwick in England, and he talks about this concept of undersampling of failure. Here’s one way to think about this. Let’s say there are two strategies – low-risk strategy and high-risk strategy. Let’s say two companies take on a low-risk strategy. One does a little bit better than the market, the other does a little bit worse than the market, but they both survive. The other is a high-risk strategy and one company does absolutely fantastically well and the other one goes out of business. When we go back to figure out which strategy is better, what do you see?

Risk is when we don’t know what the outcome is going to be, but we do know the definable distribution… Uncertainty, by contrast, is when we don’t know what the outcome is, but we actually don’t know what the underlying distribution looks like.

Well, you tend to see the two middle guys and then you see the guy with a high-risk strategy that did really well and you fail to sample the other high-risk strategy that went out of business, so you naturally perceive the high-risk strategy looks better because you have failed to consider the failure in your sample. Whenever you’re assessing a strategy, it’s very important to consider, of all the companies that embrace the strategy: those that succeeded, those that failed, and why. The natural problem is we only see the successful company. This is especially true with businesses that have strong network effects. There’s typically a long wake of companies that have failed in trying to achieve the same objective as the winning company. We tend to dwell on the winning company and its story and its success and we tend to forget about the other companies that tried to do the same thing and failed. It is actually a really, really important issue.

There’s another book I’d recommend, which I love, by Phil Rosenzweig called The Halo Effect. The idea is that when we see success we attach impossibly great attributes to those companies and when we see failure we attach impossibly bad attributes. You have to be very, very careful about extrapolating too much from performance, especially when the role of luck is important.

MOI: If we could speak about sample size. How do you extrapolate relevant outcomes with the small universe of successful names?

Members, log in below to access the restricted content.

Not a member?

Thank you for your interest.  Please note that MOI Global is closed to new members at this time. If you would like to join the waiting list, complete the following form: