Sharpe Ratio
Category
Quantitative & Statistische Methoden
Sub-category
Backtesting-Metriken
Curated by
Last reviewed
The Sharpe ratio measures a strategy's risk-adjusted return: excess return over the risk-free rate divided by the volatility of returns. If a strategy earns 12 % per year with a 2 % risk-free rate and 8 % volatility, the Sharpe ratio is 1.25. Higher values mean more return per unit of risk; values above 1 are historically considered solid.
Context & Mechanics
Definition and calculation
The Sharpe ratio – named after Nobel laureate William F. Sharpe – relates a strategy's excess return to its risk. Formula: (return − risk-free rate) ÷ standard deviation of returns. Example: 12 % annual return, 2 % risk-free rate, 8 % volatility → (12 − 2) ÷ 8 = 1.25. The metric answers the question: how much return does the strategy deliver per unit of fluctuation risk?
Interpretation
Historically common classification: values below 1 are considered weak risk-adjusted, 1 to 2 solid, above 2 very good – and values well above 3 in backtests are often a warning sign of overfitting (backtesting with curve fitting). The benchmark matters: a strategy with 8 % return at 4 % volatility (Sharpe 1.5) is better risk-adjusted than one with 20 % return at 18 % volatility (Sharpe 1.0), although the second earns more in absolute terms.
Limits of the metric
The Sharpe ratio treats upside and downside fluctuation equally – a strong winning month raises volatility just like a losing month. Strategies with rare, large losses (fat tails) can therefore show deceptively good values. Complementary metrics are the Sortino ratio (downside volatility only), the maximum drawdown and the profit factor. The value also depends heavily on the measurement interval: daily, weekly and monthly returns yield different results.
Why it matters for traders
For active traders the Sharpe ratio is above all a comparison and progress measure: it shows whether more return came from better trading or just from more risk. The GlanWick simulator tracks a strategy's risk-adjusted development via the equity curve before real capital or a challenge fee is at stake – GlanWick is a training and simulation tool and not a prop firm itself.
Execution Example
A trader compares two strategies on a $100,000 account: strategy A earned a 12 % annual return at 8 % volatility, strategy B 20 % at 18 % volatility. The risk-free rate is 2 %.
- Strategy A: (12 % − 2 %) ÷ 8 % = 1.25.
- Strategy B: (20 % − 2 %) ÷ 18 % = 1.0.
- Comparison: A delivers more return per unit of risk, although B would have earned $8,000 more in absolute terms ($20,000 vs. $12,000).
- Consequence: running B at half size to match A's volatility would leave only ≈10 % return – A remains superior risk-adjusted; for a prop-firm account with tight loss limits that is the more relevant yardstick.
Execution Risk & Errors
Comparing only absolute returns and ignoring the risk taken
Not recognising very high backtest Sharpe values as an overfitting warning
Equating upside and downside volatility although only losses hurt (check Sortino as complement)
Directly comparing Sharpe ratios from different measurement intervals
Calculating the metric on samples that are too small
Frequently Asked
What is a good Sharpe ratio?
Historically values above 1 are considered solid and above 2 very good. Values well above 3 in backtests are frequently a hint of overfitting rather than genuine quality.
How does the Sortino ratio differ from the Sharpe ratio?
The Sortino ratio considers only downside volatility. Strategies with many small wins and rare large losses are assessed more realistically that way.
Why is a high return alone not meaningful?
Because it can be bought with any amount of risk. The Sharpe ratio normalises return to fluctuation risk, making strategies comparable.
Can I determine my strategy's Sharpe ratio without real money?
Yes. Every run in the GlanWick simulator produces an equity curve from which return and volatility – and thus the Sharpe ratio – can be read.