Survivorship and Related Biases

Ever see one of those Facebook posts that shows kids riding in the back of a pickup truck, maybe back in the 1950s, with commentary something like “We didn’t wear seat belts and we turned out just fine.”? Such posts are supposed to speak to an over-reaching nanny state, I suppose.

But you know what? The kids who were killed because they were not wearing seat belts are obviously not posting on Facebook. This example is a literal illustration of survivorship bias—only the survivors are around to post on Facebook.

What does this have to do with finance? Plenty. It is fairly well known that there is survivorship bias in financial databases. As a simplified example, if you look at the past performance of all hedge funds that existed on 31 December 2019, you would not see those that had failed before that time. Instead, you only see the survivors, which have tended to be successful. Consequently, you cannot answer questions like “Do hedge fund managers provide positive alpha?” accurately using the database. Of course, you probably cannot go back and identify all the non-survivors; thus, it might be impossible to overcome completely this statistical bias.

A closely related issue is “backfill bias,” in which a hedge fund’s performance is not added to a database until it has had a period of positive performance. Fortunately, it is possible to combat this bias; see “The Fix Is In: Properly Backing out Backfill Bias” by Philippe Jorion and Christopher Schwarz in The Review of Financial Studies, Volume 32, Issue 12, December 2019.

Survivorship bias is not limited to databases. Suppose you want to study successful investors to determine the keys to their success. For example, Warren Buffett and his style of decision-making have been studied extensively. But he has been a success. What you do not (and generally cannot) observe is those investment managers who tried to emulate the Buffett style but failed. The lesson: In addition to what you are seeing (successful Warren Buffett), you should ask yourself, “What am I not seeing?” 

The question also alerts you to a related bias—the availability bias—your tendency to focus on information that comes readily to mind. So articles that you have read recently can play an outsize role in your thinking. Memory is also a source of information that comes readily to mind and here the so-called “recency effect” plays a reinforcing role—more recent information is better remembered. Obviously, biases interact with each other, sometimes leading you in the same direction, sometimes working in opposition.

Awareness of survivorship bias can help you make better decisions in arenas other than investing. For example, you can observe that a number of very successful entrepreneurs dropped out of college, e.g., Mark Zuckerberg and Bill Gates. I think by now you are ahead of me: What you do not see are the entrepreneurs who dropped out of college and did not succeed, and there are certainly many of those! One could push faulty reasoning even further: Both Zuckerberg and Gates dropped out of Harvard. So the key to success is to get admitted to Harvard and then drop out?!

Finally, what you observe and do not observe (or think about) says something about our perception of the roles of luck and hard work in success. Entrepreneurs who have worked hard are keenly aware of that (because of the availability bias and perhaps the recency effect), but might not think about the lucky breaks that have helped them succeed such as being born into a family that values education, eats well, exercises, etc. Robert Frank has written cogently on this topic and I recommend his book, Success and Luck. Whether examining your own life or those of others, it is important to disentangle the contributions of hard work versus luck. It is attribution analysis on a high level.

In sum, the existence of survivorship bias should prompt people to ask, “What am I not seeing and what are the implications of that?” The existence of the availability bias and the recency effect should prompt people not to rely too heavily on evidence (including memory) that is readily at hand. Clear thinking can be hard work!

About the Author

John S. Howe, PhD, CFA, is Professor of Finance and Missouri Bankers Chair at the University of Missouri–Columbia. He is the co-author of The Foolish Corner: Avoiding Mind Traps in Personal Financial Decisions, available on Amazon. Connect with John on LinkedIn.

John Howe