Performance Analytics

There's no single most important metric. Use a combination: Sharpe ratio (risk-adjusted return), max drawdown (worst loss), and win rate/profit factor (trade statistics). Together they give a complete picture.

Use a benchmark matching your investment universe. For Canadian stocks: S&P/TSX Composite (XIC). For US exposure: S&P 500 (ZSP). For balanced: 60% TSX / 40% bonds. The benchmark should be something you could actually invest in passively.

Sharpe = (Your Return - Risk Free Rate) / Your Volatility. Example: 15% return, 4% risk-free rate, 12% volatility → (15-4)/12 = 0.92. For daily data, annualize both return and volatility before calculating.

Win rate alone doesn't determine success. 40% win rate with 3:1 avg win/loss ratio is very profitable. 70% win rate with 0.5:1 ratio loses money. Focus on expectancy: (Win% × Avg Win) - (Loss% × Avg Loss) must be positive.

Daily: Quick P&L check. Weekly: Performance summary and risk metrics. Monthly: Full analytics with attribution. Quarterly: Deep strategic review. More frequent for active traders.

Use Sortino when your returns are asymmetric (e.g., options strategies with capped loss but unlimited gain). Sortino only penalizes downside volatility, not upside. If your strategy has positive skew, Sortino will be more favorable than Sharpe.

Alpha = Portfolio Return - [Risk Free + Beta × (Market Return - Risk Free)]. First calculate your beta (covariance with market / market variance). Then compute expected return given beta. Alpha is actual minus expected.

Absolute: Did you make money? (Return > 0). Relative: Did you beat the benchmark? (Return > Benchmark). You can have positive absolute but negative relative (made money but underperformed) or vice versa.

Break down returns by category. By strategy: Calculate P&L for each strategy type. By sector: Group positions by sector, sum P&L. By position: Rank individual position contributions. This shows what's driving your results.

IR measures active return per unit of active risk. IR > 0.5 is good, > 0.75 is excellent, > 1.0 is exceptional. Few active managers sustain IR > 0.5 over long periods. IR of 0.5 over 2 years is statistically significant.

Report both gross and net returns. Calculate risk metrics on actual exposure, not nominal capital. Sharpe on leveraged returns will look artificially high. Also track return on margin/capital employed separately.

Take your actual trade results, randomly shuffle the order thousands of times. Plot resulting equity curves. This shows the range of possible outcomes given your trade distribution. Use 5th percentile for risk assessment.

Regress your returns against factor returns (market, size, value, momentum, quality). The coefficients show your factor exposures. R-squared shows how much is explained by factors. The unexplained portion (alpha) is idiosyncratic return.

Kelly criterion determines optimal bet size to maximize geometric growth: f* = (bp - q) / b. Track if your actual position sizing matches Kelly. Over-betting increases variance; under-betting leaves return on table. Most use fractional Kelly (25-50%).

Historical: Apply actual crisis returns (2008, 2020) to current positions. Hypothetical: Model scenarios (30% market drop, rate spike). Correlation: Test what happens if correlations go to 1. Report potential losses under each scenario.