How to verify a trading backtest is real

The core test of any backtest is reproducibility. If you cannot independently regenerate the result, you are trusting a picture. A real strategy lets you run it on your own platform and get the same trade list — same entries, same exits, same numbers.

Check for look-ahead bias

The most common way backtests lie without lying is by using information that wouldn't have been available in real time. Signals that repaint — that change after the fact — produce gorgeous, impossible results. Confirm every signal is based on confirmed candle close, not on data that finalizes only later.

Demand the losses

An honest backtest shows you its worst case as prominently as its best. If the only number on offer is a win rate or a total return, the drawdown is being hidden because it's unflattering. Ask for the maximum drawdown, the blow rate, and the full percentile spread, not just the median.

Beware the single window

A strategy tuned to look perfect on one carefully chosen date range will fall apart outside it. Reproducibility plus disclosed tail risk plus a sample large enough to mean something — that combination is hard to fake. That's the bar to hold any vendor to, including us.