Survivorship Bias: Why You See Only Survivors

Survivorship bias is a distortion in which we draw conclusions based only on the 'survivors,' without seeing those who didn't make it to the end. In trading it distorts both the evaluation of strategies and the perception of traders' success. Let's look at what the survivor's error is and how it deceives.

What survivorship bias is

Survivorship bias is a systematic error in which analysis is built only on the objects that 'survived' some selection, while the ones that 'perished' are invisible and unaccounted for. A classic historical example: analyzing planes that returned from combat missions and reinforcing the spots where they were damaged, one overlooks that the vulnerable areas are precisely those whose damage appeared on the planes that didn't return (the invisible ones). The essence of the error is that we see only the survivors and draw conclusions from them, forgetting those we can't see, which distorts the picture. In trading this shows up in several important forms.

The survivor's error in evaluating strategies

In testing and evaluating strategies, survivorship bias arises when the data includes only 'surviving' instruments. For example, if you test a strategy on the current set of assets while ignoring those that disappeared (went bankrupt, were delisted), the result is inflated: the failed cases are invisible, only the survivors remain. The same goes for databases cleaned of vanished instruments. Such a test result is deceptively good because it doesn't account for the failures that would have been real in real time. Although on forex (where the currency pairs of major economies are traded) this distortion is weaker than in stocks, the principle matters: testing should account for the full picture, not just the 'survivors.'

The survivor's error in perceiving success

Survivorship bias is even more insidious in the perception of traders' success. We see the stories of successful traders (those who 'survived' and got rich) and conclude that their methods work and that success is attainable. But we don't see the huge number of people who traded the same way or similarly and lost (didn't 'survive'); they're invisible, and no one tells their stories. This distorts perception: success seems more attainable and inevitable than it is, and the role of luck is underrated. Among the many people trading, some will succeed purely by statistics (even trading randomly), and it's precisely their stories that are visible, creating the illusion that their approach is reliable. The survivor's error makes us learn from the winners without knowing how many losers did the same thing.

How to account for survivorship bias

Accounting for the survivor's error requires remembering the invisible 'perished.' When testing strategies, use full data that includes vanished instruments, not just the 'survivors,' so the result isn't inflated. When evaluating someone else's success, remember that you see only the survivors: behind every success story stand many invisible failures with a similar approach, and the success may reflect luck and the statistics of large numbers rather than just the reliability of the method. Be skeptical of 'proof' that an approach works based on successful examples; ask how many people did the same thing and lost. Don't overrate the attainability of success or the reliability of methods based on the visible winners. Understanding survivorship bias protects you both from inflated evaluations of strategies (with incomplete data) and from illusions about success, helping you see the full picture, including those who usually go unseen.

Practical takeaway

Survivorship bias is a distortion in which conclusions are built only on the 'survivors,' while the 'perished' are invisible and unaccounted for, distorting the picture (the classic example: reinforcing the surviving planes instead of the vulnerable spots of the ones that didn't return). In trading it appears in two ways. In evaluating strategies: testing only on 'surviving' instruments (without the vanished, bankrupt, delisted ones) inflates the result because the failed cases are invisible; on forex this is weaker than in stocks, but the principle matters, the test should account for the full picture. In perceiving success: we see the stories of successful traders but not the many losers with a similar approach, so success seems more attainable and inevitable while the role of luck is underrated (some succeed purely by statistics, and it's their stories that are visible). Account for the survivor's error: test on full data with vanished instruments, remember the invisible failures with the same approach when judging others' success, be skeptical of 'proof' based on successful examples, and don't overrate the attainability of success or the reliability of methods based on visible winners. Understanding survivorship bias protects you from inflated evaluations of strategies with incomplete data and from illusions about success, helping you see the full picture, including those who usually go unseen, and therefore judge both strategies and the attainability of success more soberly.

This material is for educational purposes and is not individual investment advice.