Backtesting Metrics Cheat Sheet

A single-page reference to the metrics used to evaluate a backtested strategy, each with its standard textbook formula, what it actually measures, and its blind spot.

Metrics Cheat Sheet: Backtest metrics fall into three families: return metrics (CAGR, absolute and annualised return), risk-adjusted metrics (Sharpe, Sortino, Calmar, information ratio), and risk or consistency metrics (maximum and average drawdown, recovery factor, profit factor, win rate, payoff ratio, expectancy, Ulcer index). No single figure describes a strategy: a CAGR without a drawdown is meaningless, and a high win rate can still lose money if the losers are large. Every formula below is a standard definition, and every result should be read as excluding real trading costs unless the backtest explicitly modelled them.

This cheat sheet lists the metrics that matter when judging a backtested strategy. Read each one alongside its blind spot, because a metric computed on curve-fit or look-ahead-contaminated data is worse than no metric: it looks authoritative while being false. Work through the Validation Checklist before trusting any number here, and see the Formulas Reference for every variable defined in full.

How to read a metric

Group the metrics by the question they answer. How much did it make is a return metric. How much did it make per unit of risk is a risk-adjusted metric. How much pain to get there is a drawdown or volatility metric. A serious evaluation quotes at least one from each group plus the number of trades, because a Sharpe of 3 over 12 trades is noise, not skill. The interpretation notes below are heuristics for sanity-checking a backtest, never targets to optimise toward and never a promise of future results.

Return metrics

Metric	Formula (standard)	What it measures	Blind spot
CAGR	CAGR = (End ÷ Start)^(1 ÷ years) − 1	Smoothed compound annual growth rate of equity, ignoring the path taken.	Says nothing about drawdown or the ride; meaningless quoted alone.
Absolute return	(End − Start) ÷ Start	Total percentage gain or loss over the whole period.	Not comparable across horizons; a 40% gain over 1 year and over 5 years are very different.
Annualised return	(1 + total return)^(1 ÷ years) − 1	Return rescaled to a one-year basis for comparison.	Annualising a short sample overstates how reliable the figure is.
Expectancy	E = (Win% × AvgWin) − (Loss% × AvgLoss)	Average profit or loss per trade; the engine of the equity curve.	Must be positive after costs; negative expectancy cannot be fixed by sizing.

Risk-adjusted return metrics

Metric	Formula (standard)	What it measures	Blind spot
Sharpe ratio	Sharpe = (Rp − Rf) ÷ σp, annualised × √periods	Excess return per unit of total volatility; the default yardstick.	Treats upside and downside alike; assumes near-normal returns; inflated by short samples and autocorrelation.
Sortino ratio	Sortino = (Rp − Rf) ÷ σdown	Excess return per unit of downside deviation only.	Ignores upside volatility; sensitive to the choice of target return and to few downside observations.
Calmar ratio	Calmar = CAGR ÷ \|max drawdown\|	Return earned per unit of worst peak-to-trough loss, usually over 3 years.	Dominated by the single worst drawdown in the sample.
Information ratio	IR = (Rp − Rb) ÷ tracking error	Consistency of active return over a benchmark.	Depends entirely on the benchmark chosen; irrelevant if the benchmark is inappropriate.

Risk and consistency metrics

Metric	Formula (standard)	What it measures	Blind spot
Maximum drawdown	MaxDD = max over t of (Peak − Equity) ÷ Peak	Largest peak-to-trough equity decline; the core survivability number.	One realised path only; the drawdown you could have seen is usually worse.
Average drawdown	mean depth of all drawdown episodes	The typical decline rather than the single worst one.	Can mask a rare catastrophic drawdown behind many shallow ones.
Recovery factor	RF = net profit ÷ \|max drawdown\|	How many times the strategy earned back its worst decline.	Like Calmar, dominated by the single deepest drawdown.
Profit factor	PF = gross profit ÷ gross loss	Currency won per unit lost across all trades.	A single large winner inflates it; unreliable on a small sample.
Win rate	Win% = winning trades ÷ total trades	How often the system is right.	Says nothing about the size of wins versus losses.
Payoff ratio	Payoff = average win ÷ average loss	How large winners are relative to losers.	Meaningless without the win rate; both are needed for expectancy.
Ulcer index	UI = √(mean of squared % drawdowns)	Depth and duration of drawdowns combined into one number.	Only downside; harder to interpret and less widely reported than drawdown.

The one-line reading for each

CAGR tells you the growth rate, but only means something next to maximum drawdown.
Sharpe is return per unit of total risk; annualise daily figures by ×√252.
Sortino is Sharpe that only counts downside volatility; compare the two to see how skewed returns are.
Calmar and recovery factor reward avoiding deep drawdowns, but both hinge on one worst episode.
Profit factor around 1.0 is break-even before costs; very high values on few trades usually signal curve-fitting.
Expectancy per trade times the number of trades is the total edge; keep it positive after realistic Indian costs.

Common ways these numbers lie

No cost model. Brokerage, STT, exchange fees, GST and slippage can turn a positive expectancy negative. A frictionless backtest overstates every return metric.
Too few trades. Sharpe, win rate and profit factor need a meaningful sample; on 20 trades they are noise.
Single-path drawdown. Maximum drawdown is one realised path; Monte Carlo reshuffling shows the drawdown you could plausibly have seen.
In-sample only. Metrics computed on the data you optimised on are upper bounds; trust out-of-sample and forward-test figures more.
Non-normal returns. Sharpe assumes roughly Gaussian returns; a strategy that sells tail risk can show a lovely Sharpe right up to the loss that ends it.

Frequently asked questions

Which single metric best summarises a backtest?

There is not one. The most defensible minimal set is CAGR paired with maximum drawdown, or the Calmar ratio that combines them, plus Sharpe or Sortino for risk-adjustment and the trade count for statistical weight. Any lone number can be gamed by overfitting.

What is the difference between Sharpe and Sortino?

Both divide excess return by a measure of variability. Sharpe uses total standard deviation, penalising upside and downside swings equally. Sortino uses only downside deviation, so it rewards strategies whose volatility is mostly to the upside. Comparing the two indicates how skewed the returns are.

Why is maximum drawdown so important?

Drawdown is what actually forces traders and allocators to abandon a system. A strategy with excellent long-run CAGR is worthless if its worst drawdown exceeds your tolerance, because you will exit at the low and never realise the recovery. It defines survivability.

Is a high win rate a sign of a good strategy?

Not by itself. Win rate ignores the size of wins and losses. A system that wins 80% of the time but loses far more on its 20% losers can have negative expectancy. Always read win rate together with the payoff ratio and profit factor.

Do these formulas assume any trading costs?

No. The formulas are standard textbook definitions computed on a return or trade series. Unless the backtest that produced the inputs explicitly modelled brokerage, STT, exchange fees, GST and slippage, the resulting metrics exclude real costs and are optimistic.

What profit factor should I look for?

A profit factor of 1.0 is break-even before costs. Many practitioners treat something in the 1.3 to 1.6 range as workable on a realistic, cost-inclusive backtest with enough trades. A profit factor above 3 on a few dozen trades usually indicates curve-fitting rather than a durable edge.

Last reviewed 11 July 2026. Educational content only — not investment advice.