Costs charged on every fold
Rolling folds with next-bar-open fills, gap-aware stops, spreads, square-root impact and borrow costs.
← All products
VALIDATE
Graded only on data the strategy never saw.
Every walk-forward fold pays trading costs, the Deflated Sharpe check gets stricter on every rerun, and each result shows its universe and assumptions.
Walk-forward backtesting in the app at app.tradepolaris.com/backtests
What you get
What Walk-forward backtesting gives you.
Tearsheet with factor attribution
Equity curve vs the S&P 500, Sharpe/Sortino/Calmar, drawdown, VaR/CVaR, monthly heatmap and 6-factor attribution.
Is the edge real?
Probabilistic and Deflated Sharpe, minimum track record and overfitting probability separate edge from luck.
Overfitting suiteTearsheetPaper tradingStrategy Builder
Zero code to full control
Click or code, and face the same gates.
Use the visual builder, write Python or export the signals.
From point-and-click to production Python.
Whichever route you take, the backtest and the risk views read the same rules.
- No-code builder. Set factors, exits, optimizers and risk models.
- Python SDK. Write the strategy as a class in a sandboxed runtime.
- Export your work. Take the spec, the rule sheet or a Pine script.
strategy: momentum-x-sectoruniverse: { region: US, top: 500 }signals: - momentum(126) - momentum(21) - filter: adx(14) > 20portfolio: { long: 50, weighting: risk_parity }exits: { trailing: 8%, invalidation: signal_flip }validate: { walk_forward: 5y, costs: realistic, gates: [DSR, PBO] }
Where it fits
The loop this sits in.
Backtested / walk-forward results are hypothetical, do not represent actual trading, and are not indicative of future results.