Upload Your Backtest. AI Reveals the Risks You Can't See.

Your backtest shows 67% win rate and 2.1R expectancy. Looks great on paper. But you don't know if that edge survives transaction costs at scale, if the winning streaks are clustered in one market regime, or if the max drawdown happened during a period you would have actually kept trading through.

The counterintuitive truth

The single most misleading metric in any backtest is average R-multiple. A strategy with 2.0R average can still lose money if the distribution is fat-tailed — 10 trades at -1R followed by one trade at +12R produces a great average but requires surviving 10 consecutive losses first. Most traders quit before the payoff arrives.

How It Works

Upload your backtest results as a CSV or screenshot. The AI reads your trade log and runs five analyses that standard backtest reports skip entirely.

Win rate stability — your overall win rate gets broken down by market regime (trending vs. ranging). A 67% win rate that drops to 38% in ranging markets is not a 67% strategy.

Drawdown clustering — are your worst losses consecutive or spread across time? Clustered drawdowns cause account blowups. Spread drawdowns are survivable.

Regime sensitivity — identifies which market conditions your strategy depends on, and flags when those conditions are absent from your backtest sample.

Transaction cost impact — recalculates your equity curve with realistic slippage and commission assumptions at your expected trade frequency.

Psychological survivability — models whether you would realistically hold through the worst drawdown period, based on drawdown depth, duration, and recovery time.

What Changes

Regime clarity
See which market regimes your strategy actually works in — and which ones destroy it.
Drawdown pattern visibility
Know whether your max drawdown is a one-time event or a repeating pattern that will hit again.
True cost of execution
Understand the real cost of slippage and commissions at your expected trade frequency before you go live.
Plain-English risk profile
Get a clear, jargon-free explanation of your strategy's actual risk profile — what can go wrong and how likely it is.

Early Access Pricing

$15
per month
  • Unlimited backtest uploads (CSV or screenshot)
  • Regime analysis across trending, ranging, and volatile conditions
  • Drawdown clustering report with survivability score
  • Plain-English risk summary for every upload
Get Early Access

Join the Waitlist

No spam. Unsubscribe anytime.

Frequently Asked Questions

What format do I need to upload my backtest in?
CSV with columns for entry date, exit date, P&L (or R-multiple), and trade direction. If you don't have a clean CSV, you can upload a screenshot of your backtest equity curve and trade log — the AI extracts the data.
How does regime detection work?
The analyzer classifies each trade's market context as trending, ranging, or high-volatility using price action characteristics during the trade window. Your win rate, expectancy, and drawdown are then recalculated within each regime separately.
Can this analyze backtests from any platform?
Yes. As long as you can export a trade log with dates and P&L, the analyzer works regardless of platform. NinjaTrader, TradingView, MetaTrader, Sierra Chart, or a manual spreadsheet — all accepted.
What's a "psychological survivability score"?
It estimates the probability that a real trader would continue executing the strategy through its worst historical drawdown. The score factors in drawdown depth (% of account), duration (calendar days), and how many consecutive losing trades occur. A strategy with strong expectancy but a 40-day drawdown scores lower than one with moderate expectancy and a 12-day drawdown.
How is this different from just reading my backtest statistics?
Standard backtest stats give you averages across the entire sample. Averages hide regime dependency, drawdown clustering, and the gap between theoretical and executable performance. This tool decomposes your results into the specific conditions where your strategy works and where it fails — information that aggregate statistics never surface.