The Elegance Conceals the Flaws
The Sharpe ratio is simple, intuitive, and widely used — but its simplicity involves assumptions that are often violated in practice. Understanding these limitations is essential for interpreting risk-adjusted return figures correctly and avoiding the traps that have caught professional investors and regulators alike.
Penalising Upside Volatility
The Sharpe ratio uses standard deviation of returns in the denominator — the total dispersion of returns around the mean, whether that dispersion is positive or negative. This means a strategy that occasionally produces large positive surprises is penalised in exactly the same way as one that occasionally produces large losses.
An investor who holds a concentrated position in a stock that sometimes surges dramatically will have a high standard deviation and therefore a lower Sharpe ratio, even if all the excess volatility is on the upside. Most investors do not object to upside surprises and would prefer a measure that penalises only downside volatility.
The Sortino ratio addresses this directly. Named after Frank Sortino, it uses downside deviation (the standard deviation of returns falling below a target, typically zero or the risk-free rate) in the denominator instead of total standard deviation. A strategy with frequent large gains but rare losses will have a much higher Sortino ratio than Sharpe ratio, which better captures the investor's actual experience.
The Normal Distribution Assumption
The Sharpe ratio's theoretical foundations assume that portfolio returns are normally distributed — that the bell curve adequately describes how returns are spread around their mean. In practice, financial returns are not normally distributed. They exhibit two key departures:
First, fat tails: extreme events (very large gains or very large losses) occur far more frequently than the normal distribution predicts. A move of three standard deviations should occur about 0.3% of the time under normality; in practice, it occurs far more often in financial markets.
Second, negative skewness: large losses occur more often than large gains of the same magnitude. Financial markets crash faster than they rise; the distribution of returns is asymmetric.
A strategy that earns steady small gains while occasionally suffering catastrophic losses can exhibit exactly this profile: low standard deviation in normal times (producing a high Sharpe ratio), punctuated by rare but devastating drawdowns that the Sharpe ratio fails to capture. Nassim Taleb has repeatedly highlighted this pattern — "picking up nickels in front of a steamroller" — as a systematic failure mode of Sharpe ratio-based evaluation.
Period Sensitivity and the Risk-Free Rate
The Sharpe ratio is highly sensitive to the measurement period. The same strategy measured over 2019–2023 will yield a very different Sharpe ratio than one measured over 2007–2013, even if the strategy itself has not changed. Short measurement windows can amplify or distort the result based on idiosyncratic performance in the chosen period.
The risk-free rate also changes over time, sometimes dramatically. A strategy measured with a risk-free rate of 0.5% (as in 2015–2021) will show a higher Sharpe ratio than the same strategy measured with a risk-free rate of 5% (as in 2023–2024), even if portfolio returns are identical. Comparisons across time periods without adjusting for the risk-free rate level are often misleading.
Return Smoothing
Some strategies — particularly illiquid alternatives like private equity and hedge funds using mark-to-model valuations — report artificially smooth monthly returns because they do not mark assets to market prices. This artificially reduces the measured standard deviation and inflates the Sharpe ratio, giving a flattering picture of risk-adjusted performance that does not reflect the underlying economic exposure.