Recomputed Profit Factor
We replay the trades to compare the claimed PF with the real, net-of-fees one.
1.47vs 2.10 declared
Independent audit · Pine & Python
Audit your strategy,
Audit your strategy,
not its storytelling.
Upload a strategy. BacktestProof recomputes its metrics, detects biases and measures robustness — proof before promise.
Pine · Python · Notebook
Fees & slippage included
Your code stays private
+41% net
Verification report
Illustrative example
Verified
BTCUSDT · 4H · 2021–2026
Declared
Recomputed net
Profit Factor
1.28
declared 2.10
Net after fees
+41%
declared +212%
Drawdown
−23.5%
robust
12 assets · 3 timeframes · walk-forwardrobustesse 74/100
+41%
Verified net return
1.28
Recomputed Profit Factor
74/100
Robustness score
16/18
Positive assets
Illustrative example · net of fees · walk-forward
What BacktestProof verifies
Claimed performance, put to the evidence.
Every metric is recomputed on your data — fees and slippage included — then tied to the evidence behind it.
Net return after fees
The gross promised curve, brought back to reality: commissions, spread and slippage removed.
+41%gross +212%
Measured drawdown
The real maximum loss — the one you actually feel — across the full test window.
−23.5%max compounded
Robustness score
Out-of-sample, multi-asset and multi-period stability, summarised in one score.
74/100holds in OOS
Multi-asset × TF matrix
The same strategy tested across a broad basket and several timeframes.
16/18assets positive
HIGH
MED
MED
Detected divergences
Gaps between the declared backtest and the recalculation: look-ahead, missing fees, overfitting.
3gaps flagged
Method
Four steps, one defensible conclusion.
A transparent protocol, from upload to report — no black box.
01
Upload
Pine strategy, Python script, notebook or folder. Your code stays private.
02
Recompute
Metrics are replayed on your data, fees and slippage included.
03
Robustness
Walk-forward, multi-asset, multi-TF and out-of-sample stress tests.
04
Report
Each number tied to its evidence, with divergences and improvement leads.
Declared vs verified
The gap nobody shows you.
Side by side: the displayed performance and the one that survives recomputation.
VERIFIED
Backtest comparison
3 divergences · illustrative example
Generalisation
Robust across a basket, not on a lucky ticker.
Asset × timeframe matrix
Illustrative example
pass
fragile
rejected
74robustness / 100
Fees ×3
PF 1.24
OOS 2026
+9.1%
Slippage 2×
PF 1.11
Research & biases
The biases that inflate a backtest, named one by one.
Our engine hunts the most common errors — the ones that turn a losing curve into a spectacular promise.
High bias
Look-ahead
Decisions taken with information not yet available at bar close.
Fees
Missing costs
Commissions, spread and slippage absent: the apparent edge vanishes once deducted.
Overfitting
Single window
Performance carried by one favourable period, invisible in walk-forward.
Your report
A full deliverable, unlocked after validation.
BacktestProof — BTCUSDT · 4H · Wolfe+RSI
Report ready — action required
Blurred preview of a demo report. The full report unlocks after payment and human validation.
Unlock an example report
Pricing
An audit, not a subscription to promises.
Launch pricing. Indicative rates — final details are shown in the configurator.
FAQ
Frequently asked questions
Is my strategy code shared or stored?
No. Your code powers the audit computation and is never published or resold. You keep ownership of your strategy.
Which formats can I upload?
Pine Script, Python scripts, notebooks or a compatible folder. You can also provide your exported trades.
Are the site numbers real client results?
No. All figures shown here are illustrative examples. An audit guarantees no future performance.
How is the report unlocked?
After payment, your request goes through human validation before the full report becomes accessible via a temporary link.
Ready to put your strategy to the evidence?
Upload it. You will know what holds — and what was only storytelling.