Glossary
Survivorship bias
Studying only the companies that survived to today and ignoring the ones that delisted or went bankrupt. It makes past strategies look far better than they really were; our panel data is survivorship-free.
In plain terms
The bias is sneaky because it is built into convenient data: a free price API typically only knows tickers that still trade, so a 'buy losers' backtest silently skips the losers that went to zero.
Fixing it requires point-in-time membership — knowing exactly which stocks existed and were investable on each historical date. Our Japanese-equity panel was rebuilt that way, and several strategies that looked attractive on survivor-only data stopped working once the dead companies were put back in.
Related terms
Educational definitions only. Not investment advice.