Backtesting replays trade idea models, option strategies, flow signals, and chain-metric rules against historical data so you can test an idea before relying on it.

Use it for research and comparison. It is not a guarantee that the same setup will work in live markets.

Backtest types

• Trade Idea Models - replays registered model signals over a date range and symbol universe. Use this to compare model signal quality, score bands, and watchlist behavior.
• Options Strategies - tests multi-leg strategy templates such as spreads, condors, covered calls, and straddles on one stock.
• Unusual Flow - tests whether following unusual options activity categories historically produced useful trades.
• Chain Metrics - tests entry and exit rules built from option-chain metrics like IV Rank, Put/Call ratios, and Greek exposures.

Setting up a useful run

1. Start with a clear hypothesis, such as "high-score volume spikes on AAPL resolve faster than whale activity" or "30-45 DTE iron condors outperform 45-60 DTE".
2. Pick a date range long enough to produce a meaningful number of trades or signals. Short ranges are useful for quick checks, but they can be noisy.
3. Change one setting at a time when comparing runs. This makes it easier to know which setting improved or weakened the result.
4. For Trade Idea Models, start with all symbols to understand signal volume, then rerun on a watchlist you actually trade.
5. For strategy and flow runs, repeat the same setup on another stock before assuming the edge is broad.

Reading results

• Total Return - overall simulated return for strategy, chain, and flow backtests.
• Win Rate - percentage of resolved trades or signals that produced a win.
• Max Drawdown - worst peak-to-trough decline. A high-return run with a large drawdown may be hard to trade.
• Sharpe Ratio - return adjusted for volatility. Higher is generally better, but compare it with trade count and drawdown.
• Profit Factor - gross profit divided by gross loss. Values above 1 mean winners outweighed losers.
• Score - for Trade Idea Models, the model's confidence in a signal from 0 to 100. Strong models should improve at higher score bands.
• Move - best favorable price change observed within the evaluation window.
• Days to Win - how many trading days it took for the signal to reach the win threshold.
• Evaluation Policy - how signal outcomes are determined, such as checking whether the underlying reached a favorable move threshold.

What to trust

• Prefer runs with enough trades or evaluated signals to survive a few outliers.
• Compare return, win rate, drawdown, and trade count together. No single metric is enough.
• If a strategy only works on one symbol or one month, treat it as fragile.
• If skipped or unpriced counts are high, review the details before trusting the summary metrics.
• Use Rerun to test the same idea with adjusted dates, thresholds, DTE, holding period, or symbol universe.

Tips

• Start broad, then narrow. Broad runs reveal volume; focused runs reveal whether the result applies to names you actually trade.
• Save mental notes about each run: what changed, what improved, and which metric got worse.
• Compare at least two neighboring settings. For example, test 30, 45, and 60 DTE instead of only one value.
• Review individual trades or signals when summary metrics look surprisingly strong.
• Treat backtests as decision support, not trading instructions.