Glossary · Economic ideas
Survivorship bias
Survivorship bias is the mistake of judging by the winners you can see while forgetting the losers that dropped out of sight.
Also called: survivor bias, survivorship, selection bias
In plain words
You read about college dropouts who became billionaires and think: maybe school doesn’t matter. But you never read about the many dropouts who didn’t. Looking only at the survivors makes the odds look far better than they are. That is survivorship bias.
It sneaks in everywhere: “old buildings were built better” (the badly built ones fell down long ago), “music was better back then” (we only remember the hits), “this fund has beaten the market for ten years” (the funds that failed were closed and are no longer on the list).
The question to ask is always: what am I not seeing? Who started out with the winners but is no longer in the picture?
Example
The planes that didn’t come back
During World War II, US analysts studied bombers returning from missions and found bullet holes clustered on the wings, the tail and the body. The obvious idea: add armor where the holes are.
- The statistician Abraham Wald, of the Statistical Research Group at Columbia University, pointed out what the data were missing: the planes that had been shot down.
- Holes on the returning planes showed where a plane could be hit and still make it home.
- The places with no holes on returning planes, such as the engines, were where hits brought planes down.
So the armor belonged where the survivors had no holes. Looking only at the survivors gives exactly the wrong answer.
One level deeper
In investing, survivorship bias is everywhere. Fund companies close or merge their worst funds, so a list of today’s funds looks better than the funds investors actually had. Studies of US mutual funds find that a large share of the funds that existed 20 years ago are gone, mostly poor performers.
Stock indexes have it too. The S&P 500 drops companies that shrink and adds ones that have grown, so today’s list is a list of survivors. That is fine for the index itself, which you could have owned all along, but a study of only today’s members over the past 50 years would make stocks look better than they were. Lists like the dividend aristocrats are survivors by definition: a company that cuts its dividend drops off.
The bias goes far beyond money: medical studies that count only patients who finished a treatment, business books that study only successful firms, advice from lottery winners. The cure is to start from everyone who began, not from those who are left.
Why it matters
It makes risky paths look safe and average results look great, because the failures are out of sight.
Investment records are especially prone to it: fund track records, “best stocks” lists and success stories leave out what didn’t make it.
It is one of the easiest mistakes to fix: ask what happened to all the ones that started.
A common mistake
Myth: If many successful people did something, it must lead to success.
In fact: Only if those who did it and failed were rare. If 1,000 people quit their jobs to start a company and 10 became rich, interviewing the 10 tells you nothing about the odds: you would need to hear from the other 990 too.
Questions people ask
What is an example of survivorship bias?
Judging mutual funds by the ones still open today, when the worst were closed; or concluding that old houses were built better, when the badly built ones are gone.
Why is it called survivorship bias?
Because the data include only the survivors of some process (funds that stayed open, firms that stayed in business, planes that came back), and that skews the picture.
How do you avoid survivorship bias?
Start from everyone who began: all the funds that existed at the start of the period, all the companies that were in the index then, all the people who tried. Then see how all of them did.
Read more on BigCapital
About the numbers
- This page explains a term. It is not investment advice.