Performance Metrics Glossary — Sharpe, Calmar, Profit Factor, and More

Performance metrics are the quantitative language of trading-system evaluation. A strategy without measurable performance characteristics is just a hunch; metrics let you compare strategies, identify edge, and recognise when 'good performance' is actually evidence of hidden risk. This cluster collects the metrics most relevant to algorithmic forex trading.

The foundational metrics are risk-adjusted return ratios (Sharpe, Sortino, Calmar) that normalise return by volatility or drawdown. These let you compare a high-return high-volatility scalper against a steady-return low-volatility trend-follower on the same axis — risk-adjusted productivity per unit of risk taken.

Distribution metrics (profit factor, recovery factor, win/loss ratio, expectancy) characterise the shape of a strategy's returns. A 60% win rate strategy with 0.8 win/loss ratio is fundamentally different from a 30% win rate strategy with 3.0 win/loss ratio, even if both have identical expectancy. Understanding the distribution matters for psychological tolerance and tail-risk awareness.

Diagnostic metrics (MAE, MFE, R-multiple) provide trade-level analysis that aggregate metrics hide. MAE distributions reveal whether stops are too tight or too wide; MFE distributions reveal whether take-profit logic captures available profit. R-multiples normalise across position sizes for consistent comparison.

Position-sizing metrics (Kelly criterion) provide the mathematical framework for optimal bet sizing. Full Kelly is rarely used in practice because of its psychological aggressiveness, but understanding Kelly clarifies why position sizing matters — being right about edge isn't enough if you bet too much per trade.

For evaluating any commercial EA: insist on multi-year metrics computed on verified live trading data. A backtest Sharpe of 3.0 means little; a live Sharpe of 1.5 over three years is genuine evidence of edge. The trader's challenge is not finding 'high Sharpe' EAs but finding ones whose live metrics support the marketing claims.