Sharpe Ratio

Sharpe Ratio plots a rolling Sharpe ratio as a live pane — the definitive clean build of the standard risk-adjusted score. Excess returns over the last 252 daily periods are averaged, divided by their standard deviation, and scaled to annual terms. The line prints green above zero and red below, framed by dashed guides at 1, 2, and 3 and a dotted target level, while a dashboard ranks it against up to three comparison symbols computed identically.

How to Trade the Sharpe Ratio?

• Zero crossings: excess returns flipping positive or negative, each with an alert.
• Target monitoring: set Target Sharpe to a deployed system's backtest expectation; the below-target alert flags live performance decaying against it.
• Cross-symbol ranking: dashboard rows share one measurement recipe — readings only rank cleanly when window, frequency, and annualization match.

Optional detail rows expose the arithmetic: annualized mean excess return and volatility, the window and frequency, and the annualization factor.

Sharpe Ratio Settings

• Window Length (default 252): return periods per reading.
• Returns Timeframe (default D): sampling timeframe; higher-timeframe requests use confirmed bars only, so values do not repaint.
• Returns Type (default Simple): simple or logarithmic returns.
• Risk-Free Rate (% / Year) (default 0): prorated per period and subtracted before averaging.
• Periods per Year (0 = Auto) (default 0): auto resolves 252, 52, or 12 for daily, weekly, or monthly data, and assumes 24/7 trading intraday.
• Target Sharpe (default 1) with Show Target Level (enabled).
• Comparison rows: Symbol 1 (AMEX:SPY, shown), Symbol 2 (NASDAQ:QQQ, hidden), Symbol 3 (NASDAQ:TLT, hidden).
• Guide Levels, the dashboard, and Show Calculation Details are on by default.

Frequently Asked Questions

When is the Sortino ratio the better score?

When the return stream is asymmetric by design. The Sortino Ratio divides by downside deviation only, so a strategy whose volatility is mostly upside is not penalized for it, whereas standard deviation charges both directions equally.

Why does my intraday reading look off?

Auto annualization assumes continuous 24/7 trading below daily timeframes, which suits crypto but not session markets. Set Periods per Year manually — for example 1638 for hourly bars across a 6.5-hour equity session — and the scale corrects.

Is a high reading proof of quality?

No. Short windows are statistically noisy, and a heavily skewed return stream can print a flattering number until a tail event. Read the level together with the window size and the drawdown behavior behind it.