RiskRatios<T>
Models & Types · AiDotNet.Finance.Risk
Risk-adjusted performance ratios from a periodic return series: Sharpe, Sortino, and Calmar.
Implements: IRiskRatioCalculator<T>
For Beginners
Sharpe divides average return by total volatility; Sortino only counts downside volatility (it doesn’t punish upside swings); Calmar divides annualized return by the worst peak-to-trough drawdown. Higher is better for all three.
How It Works
AiDotNet exposes Sharpe on portfolio optimizers and trading agents, but Sortino (downside-only) and Calmar (return-over-max-drawdown) were missing as standalone, series-based utilities. These let any forecaster/strategy backtest be scored on risk-adjusted return without a portfolio or agent object.
Constructors
| Constructor | Summary |
|---|---|
RiskRatios() |
Properties
| Property | Summary |
|---|---|
IRiskRatioCalculator<T> Default { get; } |
Shared stateless default instance for injection as an IRiskRatioCalculator. |
Methods
| Method | Summary |
|---|---|
T Calmar(IReadOnlyList<T> returns, int periodsPerYear) |
Calmar ratio: annualized return / maximum drawdown, computed from a periodic return series by compounding an equity curve. |
T Sharpe(IReadOnlyList<T> returns, double riskFreePerPeriod, int periodsPerYear) |
Annualized Sharpe ratio: (mean(excess) / std(excess)) · √periodsPerYear. |
T Sortino(IReadOnlyList<T> returns, double riskFreePerPeriod, int periodsPerYear) |
Annualized Sortino ratio: (mean(excess) / downsideDeviation) · √periodsPerYear. |