Return per unit of risk: three ways to measure it.

Why risk-adjust at all?

Two portfolios that returned 10 % in a given year can be very different. One bounced between +5 % and +15 % across the year — smooth, predictable, low-stress. The other ranged from −20 % to +35 % — same end point, very different journey. The volatility itself matters, both psychologically (investors panic-sell during drawdowns) and structurally (a drawdown to −20 % requires +25 % to break even, not +20 %; the asymmetry compounds against high-volatility portfolios).

Risk-adjusted return measures attempt to make portfolios with different volatility profiles comparable.

Sharpe ratio

The original risk-adjusted measure. (Portfolio return minus risk-free rate) divided by portfolio standard deviation:

Sharpe = (R_p − R_f) / σ_p

Higher is better. Reference values:

• Sharpe < 0: portfolio underperformed the risk-free rate. Bad.
• Sharpe 0–0.5: mediocre. Most actively managed mutual funds fall here after fees.
• Sharpe 0.5–1.0: decent. The S&P 500's long-run Sharpe is roughly 0.4–0.5; managers above 0.7 are doing real work.
• Sharpe 1.0–2.0: strong. Persistent Sharpe above 1.0 over a multi-year period is uncommon.
• Sharpe > 2.0: exceptional. Verify the calculation is correct, the time period is meaningful (3+ years), and the strategy has not benefited from a single fortunate position.

Sortino ratio

A refinement of Sharpe that penalises only downside volatility. Upside volatility is good (you don't mind big positive months); downside is bad. Sortino divides the excess return by the standard deviation of negative returns only:

Sortino = (R_p − R_f) / σ_downside

Sortino is always higher than Sharpe for the same portfolio (because the denominator is smaller). The ranking of portfolios can differ between the two measures: a portfolio with skewed-positive returns will look better under Sortino than Sharpe.

Information ratio

Measures excess return relative to a benchmark, normalised by tracking error (the standard deviation of the excess returns):

IR = (R_p − R_benchmark) / σ_{(Rp − Rbenchmark)}

The right measure for active managers. An IR of 0.5 over a multi-year period is materially good. An IR above 1.0 in active management is rare; an IR above 1.5 sustained over a decade is essentially world-class.

Worked example

Portfolio A returned 12% with volatility of 15% in a year when 3-month T-bills paid 4%. Portfolio B returned 12% with volatility of 22% over the same year.

• Portfolio A Sharpe: (12 − 4) / 15 = 0.53
• Portfolio B Sharpe: (12 − 4) / 22 = 0.36

Same headline return, very different risk-adjusted picture. Portfolio A delivered the same return with materially less volatility — the better outcome on a Sharpe basis.

What the calculator on this site does — and doesn't

The main calculator computes simple ROI, CAGR, IRR, and real return. It does not compute Sharpe, Sortino, or information ratio — those require a return time-series, not just an initial value and a final value. Risk-adjusted measures need monthly or daily return data; the calculator's input is a snapshot.

For Sharpe-ratio analysis, export your portfolio's monthly return series from your brokerage and run the calculation in a spreadsheet: standard deviation of monthly returns × √12 to annualise, divided into the annualised excess return.

Common misuses

• Short time periods. Sharpe ratios over less than 12 months are heavily influenced by the specific months sampled. Use 3+ years for meaningful comparison.
• Survivorship bias. Funds that exist today are the survivors of a much larger pool that included failures. Reported Sharpe ratios for “the universe of mutual funds” are systematically biased upward by the failures that no longer exist.
• Non-normal distributions. Sharpe assumes returns are approximately normal. Tail-risk strategies (selling deep out-of-the-money options) produce attractive Sharpe ratios that mask catastrophic-loss exposure. Sortino partially addresses this; for skewed strategies, look at the actual return distribution, not the headline Sharpe.