Backtesting, optimisation and gates

How results are produced with real costs and how a strategy earns promotion.

Real exchange fees and costs

Backtests apply each venue's published maker and taker fees, funding payments on perpetual futures, and a market-impact model on every fill. Equity is marked to market bar by bar, and leveraged positions can be liquidated at the venue's maintenance margin. Nothing is reported gross of costs.

Position sizing and leverage

Strategies size positions relative to account equity rather than in fixed amounts. The requested leverage is plumbed into the engine and clamped to the venue's cap: spot and options venues stay at one, futures venues allow more. Spot accounts cannot short, and the strategy base class enforces it.

Three-phase optimisation

Sensitivity analysis varies each parameter by twenty percent to find cliffs. Walk-forward optimisation runs Bayesian search across rolling windows, tracking in-sample and out-of-sample results separately. The final holdout, the last fifth of the data, is tested once and must hold up against the walk-forward out-of-sample result.

Overfitting detection

A strategy is rejected when out-of-sample performance is negative, when the in-sample to out-of-sample ratio is too high, when the probability of backtest overfitting is above one half, or when the deflated Sharpe ratio, which accounts for how many trials were run, no longer supports the result.

Promotion gates

Promotion floors are enforced in code from a published criteria file. Hard gates cover overfitting, non-positive out-of-sample results, failed holdouts and zero trades and cannot be waived. Soft gates cover Sharpe ratio, profit factor, drawdown and win rate and can be waived only with a logged reason that the Risk Officer confirms or rejects.

Reading a backtest honestly

Compare the equity curve with the drawdown, check how many trades the result rests on, and read the average return per trade against the round-trip cost. A high return from a handful of trades, or a Sharpe ratio that disappears out of sample, is a warning rather than a result.