πŸ”¬
Honest Backtest

The Backtest That Doesn't Lie

Most strategies look brilliant on the data they were tuned on β€” and collapse live. This is how we validate a strategy before calling it tradable: out-of-sample testing, walk-forward folds, parameter sensitivity, and a blunt overfit check.

⚠️ Sample / format preview β€” the numbers below are NOT a real backtest. Every chart and metric is placeholder data generated for layout only, to demonstrate the validation methodology and how the report looks. It is not the result of running any strategy on real market data. When you commission a validation, this exact format is filled with your strategy, on your data. Past performance never predicts future returns.
Sample data

In-Sample vs Out-of-Sample Equity

The strategy is tuned only on the in-sample period (left of the split). The out-of-sample period (right) is unseen data β€” the honest test. A strategy that keeps climbing past the split is robust; one that flatlines or reverses was overfit.
In-sample (tuned) Out-of-sample (unseen) Split point
Sample data

Walk-Forward Folds

Train on a window, test on the next unseen window, roll forward. Green test bars that stay positive across folds = the edge survives regime changes.
Sample data

Parameter Sensitivity

OOS return across a grid of two parameters. A broad green plateau = robust. A lone bright cell surrounded by red = a fragile, overfit spike to avoid.
Sample data

Fold-by-Fold Results

In-sample vs out-of-sample return per fold, and the degradation between them. Low, consistent degradation is the goal.
FoldPeriodIS ReturnOOS ReturnDegradationVerdict

Overfitting Red-Flag Check

The questions we run every strategy through before trusting it.

Have a strategy you're not sure you can trust?

Send us the Pine script or the rules. We'll run this exact validation β€” out-of-sample, walk-forward, sensitivity β€” and tell you honestly whether it's tradable or overfit. No sugar-coating.

πŸ”¬ Request an Honest Validation