survivorship bias
An error that comes from looking only at funds or companies still alive, which flatters the average because the failures dropped out.
Survivorship bias is the error you make when you look only at the funds, companies or strategies that are still around. The ones that failed, closed or merged quietly drop out of the data, so the average of what remains looks better than the average of what people actually owned.
In investing it shows up everywhere. Fund tables list today's funds and their records, not the ones shut down after a bad run. Lists of great stocks start from companies that made it. Success stories about a strategy come from the people it worked for. Each time, the losers are missing, and the result flatters the choice being sold.
The fix is to ask what the starting group was. To judge active funds, count every fund that existed at the start of the period, including the ones that closed, and compare them with a cheap index fund over the same years. That one question often turns an impressive average into an ordinary one.
Example
Zapas tracks 20 synthetic active funds over two 7-year periods. By the end of the second period 5 have closed. The 15 survivors averaged 5.13% a year in years 8–14, which looks close to the index fund's 6.00%.
Add the 5 closed funds back and the average of all 20 drops to 4.20%: dropping the dead flattered the record by 0.93 points a year. Only 3 funds beat the index in the second period, and none beat it in both.
Common mistakes
- Comparing today's fund list with an index without asking how many funds closed along the way.
- Learning a strategy from people it worked for, without hearing from those it did not.
- Reading a provider's "our funds' average return" as if it included the funds they shut down.
- Assuming that because a fund survived, its past record shows skill rather than luck.