Reading a backtest result

What the numbers mean — and how to tell a real edge from luck or overfitting.

Every backtest lands on a result page with three things: an equity curve (how the account balance moved), the list of trades, and a row of headline metrics. Here is how to read them without fooling yourself.

The headline metrics

Each metric answers a different question. Read them together — no single number tells you whether a strategy is any good.

Total return

The net profit or loss over the whole tested period, as a percent of starting capital. It's a backtest outcome, not a promise — a high return over a short window or a handful of trades is weak evidence.

Sharpe ratio

Return earned per unit of risk taken. Higher is better: below 1 is weak, 1–2 is decent, and above 2 is rare — often a sign of overfitting rather than skill.

Max drawdown (MDD)

The largest peak-to-trough drop in the equity curve. −50% means the account was cut in half at its worst point. The real question it asks: could you actually hold through that without bailing?

Win rate & profit factor

Win rate is the share of trades that closed in profit; profit factor is gross profit divided by gross loss. A high win rate can still lose money if the few losers are much larger than the many winners — always read the two together.

The Reality check

Below the metrics, every result carries a free Reality check: a plain-language read of how this specific result might mislead. It's the honest counterweight to a good-looking curve. Common flags:

  1. Too few trades — a result from a handful of trades is mostly noise; aim for 30+.
  2. Suspicious return — a big return on very few trades is likely luck or overfitting, not an edge.
  3. High or severe drawdown — a drop you probably couldn't sit through in live trading.
  4. No trading costs modeled — real fills include fees and slippage, so live results will be worse.
  5. Beats / underperforms buy & hold — did the strategy actually beat simply holding the asset?
  6. Overfit vs robust out-of-sample — whether the edge survives on data it wasn't tuned on.

Overfitting: the trap to watch for

Overfitting is when a strategy is tuned so tightly to past data that it captured noise instead of a real pattern. It looks great in the backtest and falls apart on new data. The tells: a stellar return from very few trades, an unusually high Sharpe, or results that collapse out-of-sample.

To pressure-test an edge, the Validation Lab (Pro and up) re-runs a saved strategy across consecutive time windows (walk-forward): consistent results across windows suggest a real edge; scattered ones suggest luck.