Survivor Bias is an important idea to understand as an investor. It refers to the human predilection for focusing on visible results while ignoring those outcomes that were possible, but did not occur. This frequently happens when investors consider fund managers. It, of course, makes sense to look at those managers and funds available to us, as they are our only investable options. The mistake is to believe that just because those managers are available, just becasue they have somehow made money in the past, they possess special skill and will continue to make money. It is entirely possible that these managers are equivalent to or worse than those managers who did not make it to the present. Our survivor bias pressures us to consider only the available managers, and believe that their very survival bespeaks ability. We should resist this bias, and instead rationally consider managers and their strategies using tools and metrics available to us, such as Sortino and Sharpe Ratios.
Q. Now, this is not exactly a buzzword you hear every day. Why is it important?
A. It’s like the famous Sherlock Holmes story where the dog didn’t bark in the night. You don’t hear this often, but you should. In fact, it’s a big reason why you do hear many of our other buzzwords, like Sortino, Sharpe Ratios, R squared and the rest.
Q. So this idea is responsible for our need to know all those? OK, tell us.
A. Let’s do this by example. Say I start with 10,000 investment managers, and say to each of them that their profit or loss for the year will be determined by a single coin flip. So at the end of the first year, half make money, half don’t. Now, repeat that for five years. At the end of the period, about 313 will have made money every single year. So in year 5, we’ll see them on the cover of magazines, read about their secrets of success in best selling books, and, I’m afraid to say, see them on certain business TV channels a lot! But in my example, it was all, 100%, luck, nothing else.
Q. So they’re the survivors, and some managers had to have that luck. So the bias is thinking there’s something special about them.
A. Right, we love to attribute reasons to random things, and investment performance can be top of that list. So you cannot simply look at a manager’s track record of raw returns, even for a relatively long period, and conclude definitively that he’s a great manager. He might just be lucky.
Q. And so things like Sharpe Ratios and the rest of the measurements we talk about in this segment help us get a better idea about that?
A. Yes. Lots of them, in some way, measure performance versus volatility, which is just another word for risk. And if you take a lot of risks, you might well be successful… but that success will probably be based on luck. And luck, of course, normally comes to end. The mathematical reason for that is reversion to the mean; but we don’t really need math to understand that idea.
Q. And of course Survivor Bias is also relevant to other claims in the financial world, like the folks who are selling how successful their trading systems are.
A. Definitely. Those can be even worse, because a lot of them are generated through “backtesting,” where folks look in the rear view mirror to find patterns. Remember, if you torture the data long enough, it will confess to anything.