Survivorship bias: why success statistics mislead

Success stories are available; failure stories are not. From that asymmetry comes a systematically wrong picture.

Where it appears in markets

In indices: companies that ceased to exist have been removed. An index's long-term return is calculated on the survivors.

In funds: the unsuccessful ones close and vanish from the statistics. The average return of those remaining looks better than reality — Active and passive management: what the fee buys.

In backtests: a strategy tested on an index's current constituents is tested on companies already known to have survived.

In people's stories

A successful investor tells the story of a concentrated bet that worked. Those who made the same bet and went broke write no books and give no talks.

What follows is not that concentration works but that we only see its favourable outcomes.

How to defend against it

Ask where the others are. How many participants used the same strategy and what became of them. Whether the calculation covered the full set, including the ones that disappeared.

Demand data as of the moment of decision rather than from the vantage point of today's knowledge.

Related: The benchmark: what to compare your result against honestly and How a bubble works: the general pattern.