How to read a backtest report · five things that matter more than returns

Backtest reports are full of numbers, and the most prominent is usually total return. But returns are only an outcome: they don't tell you how the result came about, how much risk it took, or under what conditions it holds.

1. Maximum drawdown

The largest peak-to-trough decline. It answers: “How bad does it get at its worst?” If you couldn't sit through that drawdown in real life, no return makes up for it.

2. Trade count and cost assumptions

Too few trades and the result may be luck; many trades and cost assumptions matter enormously. Whether fees and slippage are included — and whether those assumptions are realistic — often decides the gap between backtest and reality.

3. In-sample vs out-of-sample

Which data were the parameters tuned on? Was a separate stretch of data held back for validation? With in-sample results only, be especially wary of overfitting.

4. Performance across regimes

Split results by trending, ranging and shock regimes. A strategy that only works in one regime needs regime identification or stricter risk rules alongside it.

5. Stability

Does a small change in parameters swing the result dramatically? Is the result concentrated in a handful of trades? A robust strategy isn't overly sensitive to small changes.

One last reminder

Backtest results don't predict future performance. The point of reading a backtest report is to find a strategy's weaknesses, not to convince yourself it will succeed. This article explains a method and is not investment advice.