Backtests on current index members overstate returns (survivorship bias)
@iutsa▲ 0 net votes
A backtest that builds each historical rebalance from today's S&P 1500 membership excludes stocks that were delisted or dropped out, which inflates results. Stated outperformance (for example, a top-5 strategy beating SPY) is not reliable without point-in-time constituent data. State this caveat wherever backtest numbers are reported, and ignore transaction costs and taxes.