How does survivorship bias inflate historical returns?

Question

AcadiFi · Accepted Answer

Survivorship bias occurs when failed or discontinued entities are excluded from a dataset, leaving only survivors — which by definition performed well enough to still exist. The average return of 'funds alive today' overstates the return an investor would have earned by randomly picking a fund N years ago. Example: an analyst, Mira Kaltenbach, studies US equity mutual funds existing in 2026. Looking at 10-year returns, she finds average 9.8% annualized. But 34% of funds that existed in 2016 have since closed or merged — typically after underperforming. Elton-Gruber-Blake showed that correcting for this bias reduces measured fund returns by roughly 0.8-1.5% per year. Similar biases affect: (1) hedge fund databases (self-reporting + ceased funds dropped), often overstating by 3-4% annually; (2) stock indices where bankrupt companies were removed; (3) published academic strategies — failed strategies never get published. The fix: use point-in-time databases that include dead funds (CRSP survivorship-bias-free mutual fund DB; TASS + HFR merged for hedge funds). For CFA exam: survivorship bias is one of the 'big three' biases along with look-ahead and data mining.

How does survivorship bias inflate historical returns?

Master Level II with our CFA Course

Related Questions

Practice Questions