Advanced Metrics
These metrics cover period-level statistics, stagnation analysis, and efficiency measures that provide additional context for strategy evaluation.
Return & Efficiency Metrics
| Metric | Formula | Unit |
|---|---|---|
| CAGR | ((End Equity / Initial Capital)^(1/Years) - 1) × 100 | % |
| Avg Monthly Return | CAGR / 12 | % |
| AANP (Average Annual Net Profit) | Net Profit / Backtest Years | $/yr |
| ARR | (AANP / Initial Capital) × 100 | % |
| Profit Rank % | (Exposure-Adjusted Annual Profit / Perfect Profit) × 100 | % |
| Profit Per Bar | Net Profit / Total Bars | $/bar |
| Efficiency Ratio | CAGR / Exposure % | ratio |
Key Metrics Explained
CAGR (Compound Annual Growth Rate)
The annualized growth rate, accounting for compounding. This is the standard way to compare returns across different time periods.
| CAGR | Tier |
|---|---|
| ≥ 30% | Excellent |
| ≥ 18% | Good |
| ≥ 10% | Caution |
| < 10% | Failed |
Profit Rank %
A normalized efficiency metric that compares your strategy's actual annual profit (adjusted for exposure time) against a theoretical "perfect" strategy. Values above 50% indicate exceptional efficiency.
Efficiency Ratio
CAGR divided by Exposure %. A strategy earning 20% CAGR with only 30% market exposure has an efficiency ratio of 0.67 — your capital is deployed efficiently.
Stagnation Metrics
| Metric | Formula | Unit |
|---|---|---|
| Stagnation | Max bars without a new equity high | bars |
| Stagnation % | (Stagnation / Total Bars) × 100 | % |
| Longest Flat Days | Stagnation converted from trading days to calendar days | days |
Stagnation measures the longest period without making new highs. Even profitable strategies have flat periods — a stagnation of 30% means the strategy spent nearly a third of its time going nowhere.
Period Statistics
| Metric | Description |
|---|---|
| Best Year | Maximum yearly return % |
| Worst Year | Minimum yearly return % |
| Positive Years | Count of profitable years |
| Total Years | Count of all years |
| Best Month | Maximum monthly return % |
| Worst Month | Minimum monthly return % |
| Positive Months | Count of profitable months |
| Total Months | Count of all months |
| Positive Months % | (Positive Months / Total Months) × 100 |
| Time Underwater % | (Days below equity peak / Total Days) × 100 |
Key Stats
Positive Months % is a strong indicator of return consistency. Strategies with 65%+ profitable months demonstrate reliable, steady returns — the kind of profile that tends to earn a healthy Profitability Score.
Worst Month / Worst Year shows the worst-case calendar period. If you couldn't stomach a single month of that loss, the strategy may be too volatile for you.
Time Underwater % shows what fraction of the backtest was spent in drawdown. Even a strategy with a small max drawdown can have high time underwater if drawdowns happen frequently.
Tip
Stagnation and Time Underwater are often overlooked — a strategy with small drawdowns but 60% time underwater is harder to trade psychologically than the numbers suggest.
Outlier Analysis
These metrics measure how much your strategy's performance depends on a few exceptional trades.
| Metric | Description | Unit |
|---|---|---|
| Outlier Dependency Score | How much total profitability relies on statistical outlier trades (0-100) | score |
| Outlier Positive Count | Number of winning trades classified as outliers | count |
| Outlier Positive PnL | Total P&L from positive outlier trades | $ |
| Outlier Negative Count | Number of losing trades classified as outliers | count |
| Outlier Negative PnL | Total P&L from negative outlier trades | $ |
| Outlier Net PnL | Net P&L from all outlier trades combined | $ |
Outlier Dependency Score is the headline metric, and it reads the way the other scores do: higher is better. AlgoChef removes your largest winning trades and recalculates whether the strategy is still profitable without them. A high score means the edge survives — it does not depend on a handful of exceptional trades. A low score is the warning: strip out the outliers and the profit largely disappears, which means the edge may not repeat.
Compare Outlier Net PnL to total Net Profit — if outlier trades account for most of your profit, your strategy's apparent edge is concentrated in a few trades rather than distributed across many.
Stability Signals
| Metric | Description | Unit |
|---|---|---|
| Chrono Stability Score | Measures how consistent strategy behavior is across time periods | score |
Chrono Stability Score splits your backtest at its midpoint and compares the first half against the second — profit factor, win rate and average trade. A low score means the two halves behaved very differently, which is a sign of possible regime dependency.
Tip
Ready to analyze your own strategy? Start your 14-day free trial.