The Sharpe ratio measures return per unit of volatility above the risk-free rate: how much you were paid for the bumpiness of the ride. Useful reference points: bank deposit around 0.3, buy-and-hold equity index around 0.6, good equity mutual fund around 1.0, strong hedge fund around 2.0. Above 3 is rare. Above 4, the correct first reaction is to look for the measurement error, not to be impressed.
01Why return alone misleads
Suppose two strategies both return 40% over a year. The first grinds out small gains most months with a worst month of −3%. The second doubles in three months, loses 35% in one, and recovers. Same destination, entirely different instruments.
The difference matters for a practical reason rather than an aesthetic one: you have to still be holding the strategy at the end for the return to be yours. Most people exit the second strategy during the −35% month, often near the bottom, and realise a loss where the chart shows a gain. Risk-adjusted measures exist because the path determines the outcome you actually get.
Return is also trivially manipulable. Double the position size and the return roughly doubles. No skill has been added; only leverage. Any metric that rewards this cannot distinguish a good strategy from a large one.
02What Sharpe measures
The Sharpe ratio takes the return earned above the risk-free rate and divides it by the volatility of those returns. Higher is better: more excess return for each unit of variability.
Its central virtue is that it is scale-invariant. Doubling position size roughly doubles both return and volatility, so the ratio is unchanged. Leverage cannot inflate it. That is exactly the property return lacks, and it is why Sharpe is the first number worth asking for.
Because Sharpe does not change with position size, a strategy's Sharpe ratio tells you about the quality of the edge, while position sizing tells you about the risk you have chosen to take. These are separate decisions. A strategy sized down to fit a drawdown ceiling has the same Sharpe ratio as the same strategy sized up. It simply moves along one risk-return line.
03Reference points
Sharpe ratios are meaningless without calibration. Approximate long-run reference points:
| Instrument | Approximate Sharpe | Interpretation |
|---|---|---|
| Bank fixed deposit | ~0.3 | Low return, low volatility |
| Buy-and-hold equity index | ~0.6 | The market's own reward for risk |
| Good equity mutual fund | ~1.0 | Meaningfully better than the index |
| Strong hedge fund | ~2.0 | Genuinely excellent |
| Above 3 | Rare | Verify before believing |
| Above 4 | Very rare | Assume a measurement issue until ruled out |
These are rough and depend on measurement period and frequency, but the ordering is stable and the orders of magnitude are right. The useful takeaway is that a claimed Sharpe of 1.5 is a strong result, and a claimed Sharpe of 6 requires an explanation.
The backtested Sharpe ratios published on this site sit above 4. By the standard we have just laid out, that is exactly the range where you should ask what is wrong with the measurement. The honest answers: these are backtested figures rather than a live track record, they come from intraday strategies whose returns are measured at high frequency over a limited number of sessions, and they were selected from a very large search. Every one of those factors tends to raise a Sharpe ratio. We publish the drawdown, the session count and the date window alongside them precisely so the number can be discounted appropriately rather than taken at face value.
04Four ways it gets inflated
- Short or flattering measurement periods. A strategy measured across a stretch that happened to suit it will show a Sharpe ratio it cannot sustain. Always ask for the window and the number of observations.
- Gross rather than net returns. Omitting brokerage, statutory charges and slippage raises return without raising volatility, which raises Sharpe directly. This is the single most common inflation in retail strategy marketing.
- Infrequent valuation. Measuring returns monthly instead of daily smooths away volatility that was genuinely experienced. The ride was not calmer; you simply looked less often.
- Selection from a large search. Test enough configurations and some will show a high Sharpe ratio through luck alone. This is the most dangerous case, because nothing was falsified: the number is real, it just does not generalise. It is why we test whether in-sample ranking predicts out-of-sample ranking at all; in one search across 3.2 million configurations, that correlation was −0.001. See honest backtesting.
05What Sharpe cannot see
Sharpe treats upside and downside variation identically. A strategy that occasionally surges gets penalised the same as one that occasionally collapses, which is not how anyone actually experiences risk.
More seriously, it handles asymmetric return distributions poorly, and short-option strategies are the textbook case. Collecting small premiums consistently produces low measured volatility and therefore a high Sharpe ratio, right up until the rare large loss the structure was always exposed to. A high Sharpe ratio on a short-volatility strategy is partly a description of the risk that has not yet materialised.
This is not a reason to discard the metric. It is a reason never to use it alone.
06Read it with drawdown
Maximum drawdown, the largest peak-to-trough fall over the period, answers the question Sharpe cannot: how bad did it actually get?
Together the two are much harder to fake. A high Sharpe ratio with a large drawdown indicates smooth returns punctuated by something severe. A high Sharpe ratio with a small drawdown is a stronger claim, and correspondingly needs stronger evidence: a long window, many observations, and a clear account of how the strategy was selected.
The ratio of annual return to maximum drawdown is a useful companion. It expresses roughly how much return you received for the worst pain you had to sit through, and unlike Sharpe it is expressed in units anyone can feel.
07Sharpe, leverage and sizing
Because Sharpe is scale-invariant, it defines a line rather than a point. A strategy with a given Sharpe ratio can be run at many sizes; each size delivers a different return and a proportionally different drawdown, but the same risk-adjusted quality.
This is what makes drawdown-targeted sizing coherent. You choose the maximum drawdown you are willing to tolerate, and position size follows from it. The strategy is not made safer or more aggressive, only larger or smaller along a fixed line. Sharpe is what tells you whether that line is worth being on at all.
The practical discipline: decide the drawdown ceiling first, in advance, in writing. Sizing chosen after seeing an attractive return figure is not risk management, it is optimism.
08Questions people ask
What is a good Sharpe ratio?
Rough reference points: bank deposit ~0.3, buy-and-hold equity index ~0.6, good equity mutual fund ~1.0, strong hedge fund ~2.0. Sustained above 3 is rare and deserves scrutiny. Above about 4, look for the measurement issue before being impressed.
What does the Sharpe ratio actually measure?
Return per unit of volatility, above the risk-free rate: how much you were compensated for the bumpiness of the ride. A strategy returning 20% steadily scores higher than one returning 20% through wild swings, because the second took more risk to reach the same place.
Why is annual return a bad way to judge a strategy?
Because return rises with risk, and leverage scales it without improving anything. Two strategies with identical returns can carry completely different drawdowns. Return tells you where a strategy arrived, not whether you could have stayed invested through the journey, and staying invested is what determines the return you actually receive.
How does a Sharpe ratio get inflated?
Short or favourable measurement periods; gross rather than net returns; infrequent valuation that smooths away real volatility; and selection of the best result from a large search, where a high Sharpe ratio is what noise produces rather than evidence of edge.
Is a Sharpe ratio above 4 believable?
Occasionally, for short-horizon intraday strategies measured at high frequency, but it should never be accepted without the window, the number of observations, whether it is net of all costs, whether it is backtested or live, and how many candidates were searched before this one was chosen. Ours are backtested, and we publish those details alongside.
QuantNifty is an algorithm development and consulting firm. We are not a SEBI-registered Research Analyst, Investment Adviser, Portfolio Manager or Broker. This page is educational and is not investment advice. Performance figures referenced on this site are backtested, not a live track record. Options trading can lose more than your initial capital. Full risk disclosure is in our Terms.
Want the number checked, not sold to you?
We will tell you what a strategy's Sharpe ratio survives once costs, slippage and out-of-sample testing are applied, including when the answer is that it does not.
Tell us what you want tested