A forward testing protocol before risking real money
Posted: Sun Oct 25, 2026 10:05 am
Backtesting uses past data, and even careful testers cannot fully avoid bias. Forward testing, trading the rules in real time on a demo or very small live account, provides cleaner evidence. My protocol is simple. Write the rules in a document and do not change them during the test. Define the minimum sample size, for example fifty trades. Record every signal, including ones you decide to skip, with the reason. Track entry, stop, target, actual fill, and outcome, and note the market conditions. At the end, compare the results to the backtest. If the live win rate and average R are close to what you expected, that is encouraging. If they are much worse, look for causes: execution problems, rule drift, or a regime change. Do not move to larger size until the forward test is complete and positive after costs. The process takes weeks, which tests patience, but it prevents the common mistake of trusting a backtest that was never confronted with reality. A strategy that survives forward testing deserves more trust than one that only looks good on history.