survivorship bias
Judging a group by the winners that remain while the failures that disappeared are left out.
Survivorship bias is the mistake of judging a group using only the members that remain, while ignoring those that disappeared or failed. The remaining examples can make the past look more successful than it was.
In market data, a list of current funds or companies may exclude ones that closed, merged, or were removed. Looking only at the survivors leaves out losses and changes the question being measured.
A fair comparison needs the full group as it existed at the time, including members that later vanished. Even complete records describe the past and do not guarantee similar future results.
Example
Suppose a Pebble directory once listed 10 fictional energy funds. Four later closed after each lost $10, and six remain, each showing a $5 gain over the same period. Looking only at today's survivors suggests gains of 6 × $5 = $30.
Working with the original group: survivor gains $30 minus closed-fund losses (4 × $10 = $40) gives a total of −$10, or −$1 per original fund. The surviving list tells a very different story because it omits the four closed funds.
Common mistakes
- Assuming a current list represents all members that existed in the past.
- Counting only winners when evaluating a strategy or market segment.
- Comparing groups built from different dates or membership rules.
- Most people who trade lose money.
Learn it in the course
- Backtest basics and traps · Test a rule on past prices and see the traps: one lucky period, hindsight, too many tweaks and missing failures.
- Case: three promises that sound too good · Judge three ads that promise easy money by what you can check, not by what they promise; find the hidden losses, the guarantee, the pressure and the request for access.