Sharpe Ratio

The Sharpe ratio divides excess return (portfolio return minus the risk-free rate) by the standard deviation of those returns. The numerator measures how much the portfolio earned above a risk-free alternative -- the compensation for bearing risk. The denominator measures how volatile that compensation was -- the risk taken to earn it. A Sharpe ratio of 1.0 means the portfolio earned exactly one unit of excess return per unit of volatility. Ratios above 1.0 are generally considered good; above 2.0, exceptional; above 3.0, extraordinary and warranting scrutiny for strategy decay or selection bias in the reporting period.

William Sharpe developed the metric in 1966 to evaluate mutual fund managers -- comparing their excess returns against the volatility they generated to achieve those returns. In this original context, comparing similar equity managers over the same time period, Sharpe ratio is a genuinely useful ranking tool. The problems emerge when it is applied across different strategy types, different time periods, or -- most dangerously -- when it is used as the sole measure of portfolio risk quality.

One of the Sharpe ratio's least obvious properties is that it penalizes upside volatility equally with downside volatility. Standard deviation does not distinguish between positive surprises and negative surprises -- a portfolio that oscillates between months of +8% returns and months of +2% returns has the same volatility as one that oscillates between +5% and -3% returns. The first portfolio has high volatility that should not concern an investor; the second has lower average returns with real downside risk. Both would receive identical Sharpe penalties for the same standard deviation value. This is why the Sortino ratio, which penalizes only downside deviation, often provides a more investor-relevant risk-adjustment.

Sharpe ratios are highly sensitive to the measurement period. A portfolio with a Sharpe of 2.0 over the three-year bull market of 2019-2021 may have a Sharpe of 0.4 over the full ten-year period including 2022. Single-period Sharpe ratios extracted from favorable market environments tell you almost nothing reliable about future performance. Academic literature suggests that a minimum of three years of monthly returns is required for a Sharpe estimate to have statistical significance, and five years is preferred. Yet most fund marketing materials cite one or two year Sharpe ratios with no regime context.

Sharpe ratios can be gamed in ways that are technically accurate but economically misleading. Selling out-of-the-money options (collecting premium income) produces steady small gains that inflate both the numerator (return) and reduce the standard deviation (because the put premium income smooths returns). Until the market falls sharply and the options get exercised, this strategy looks like a high-Sharpe opportunity. Long-Term Capital Management had extraordinary Sharpe ratios for years before its 1998 collapse. Private equity and hedge fund strategies that mark positions infrequently show artificially smooth returns and therefore artificially high Sharpe ratios -- the volatility is hidden, not eliminated.

High-Sharpe strategies that get crowded become self-undermining. When enough capital chases a strategy because of its historical Sharpe ratio, the entry prices improve (more buyers raises prices) and the exit conditions deteriorate (more sellers compete when conditions turn). Risk parity strategies, which weight assets by inverse volatility and therefore mechanically pursue high Sharpe exposures, experienced significant simultaneous drawdowns in February 2018, March 2020, and 2022 as the crowded correlation assumptions broke down. A historically high Sharpe ratio can be a warning of crowding, not a confirmation of persistent edge.

The Sortino ratio replaces standard deviation with downside deviation -- only periods where returns fall below a target (typically zero or the risk-free rate) contribute to the denominator. This aligns the penalty with actual investor preferences: losses hurt, but gains do not, regardless of their magnitude. For positively skewed strategies -- those with frequent modest gains and occasional large wins -- Sortino produces a more favorable and investor-relevant assessment than Sharpe. For negatively skewed strategies -- option selling, carry trades, risk parity -- Sortino is more honest about the real risk profile.

For long-term investors focused primarily on avoiding catastrophic loss, the Calmar ratio (annualized return divided by maximum drawdown) is more psychologically and practically relevant than Sharpe. A portfolio with a Sharpe of 1.5 that experiences a 45% maximum drawdown is a very different risk proposition from one with a Sharpe of 1.0 and a 15% maximum drawdown. Most individual investors' actual decision-making is dominated by drawdown tolerance, not volatility comfort -- they can tolerate a bumpy path if it does not go sharply down, but they cannot emotionally sustain large portfolio losses even if volatility statistics look fine.