Trading Concept
Survivorship Bias
The distortion that occurs when analysis only includes assets that still exist, ignoring those that failed, delisted, or went to zero.
Backtesting on today's index members produces inflated results because the index only contains survivors. The companies that went bankrupt, delisted, or merged away are absent from the dataset — but they were tradeable at the time. A strategy that would have bought some of them takes losses that the backtest never sees.
Survivorship bias affects every asset class. Crypto tokens that went to zero disappear from exchange listings. Prediction markets that resolved at zero expire. Collectible card sets that fell out of favor stop being tracked. Any backtest that uses only currently available data is contaminated.
botwir3 does not maintain historical datasets (that is a platform responsibility). The backtesting module runs against whatever data the user provides or the platform offers. The user is responsible for evaluating whether the data includes or excludes failed assets. The library entry exists so the user knows to ask the question.