Quant Strategies

Rules-based trading strategies, evaluated without the sales pitch.

Sharpe Ratio Myths: What the Sharpe Ratio Does and Doesn't Tell You

September 3, 2026

TL;DR

What the Sharpe ratio actually measures

The formula fits on a napkin: subtract the risk-free rate from the strategy’s annualized return, then divide by the annualized standard deviation of its returns. What comes out is how much return you were paid per unit of fluctuation you had to tolerate. The numerator is the reward for taking risk at all — the excess over what you would have earned parking the money in short-term government paper. The denominator is the price of that reward in wobble, and every deviation from the average counts, in both directions.

The Sharpe ratio became the default yardstick because it is easy to compute, easy to quote, and built from price data anyone can download. For a roughly symmetric return stream — a long-only index fund, say — it is a legitimate one-number summary of the risk-reward deal. But the moment you use it to judge rules-based strategies, asymmetric return streams, volatility-targeted sleeves, or decade-old backtests, it quietly stops doing what people assume it does. Everything below is about where it breaks.

Myth one: A high Sharpe is a promise

The most dangerous phrase in strategy marketing is “Sharpe of 2.” A backtested Sharpe is an estimate drawn from one specific slice of history, and it carries error bars you will never see printed. Over three years of monthly data, a strategy whose true Sharpe is 1.0 can easily measure anywhere from 0.4 to 1.6 by luck alone. Run forty variations of a rule and the best one will look spectacular even when all forty are worthless — that is what people mean by backtest overfitting, and no ratio printed afterward tells you it happened. Markets also get crowded, volatility regimes shift, and a strategy’s own success changes the opportunity it trades. A backtested 2.0 is best treated as an upper bound on what to expect, never a lower one.

That is why every strategy card at the publisher I point readers to, Kairos Trading, carries the line “Based on backtest; not a guarantee.” That sentence is not legal decoration; it is the correct epistemic stance, printed where you will see it before you pay. When a vendor buries that framing in a terms page or omits it entirely, treat the headline ratio accordingly.

Myth two: Sharpes from different windows and assets don’t compare

A Sharpe ratio has no meaning without its measurement window, its frequency, and its asset. The same index can post a 1.3 over a calm two-and-a-half-year stretch and 0.5 over a choppy decade. Monthly versus daily annualization changes the number. The risk-free rate you subtract changes the numerator: subtracting 0% versus subtracting 5% is not a rounding error. Compare across asset classes — a bond fund, a Nasdaq momentum system, a commodity trend follower — and you are comparing return distributions that have almost nothing in common beyond the name of the ratio.

So when a pitch shows “our Sharpe 2.1, market 0.9,” the first questions are: same calendar window? Same measurement frequency? Same benchmark, measured over the strategy’s own window? Comparing a two-year-old strategy against a “ten-year SPY” line is comparing two different experiences. Reports that do it properly — such as those at kairostrading.net, which line each system’s Sharpe and Sortino up against its benchmarks over the same reported period — are the ones from which you can actually learn. The honest checklist for any published ratio: same window as the benchmark, stated frequency, stated risk-free assumption, and the raw returns printed right next to it.

Myth three: Upside gets punished — and the Sortino shows the gap

Standard deviation does not care about direction. A spectacular winning month inflates the denominator exactly as much as a catastrophic losing one. A momentum-style system can sit flat or drift lower for months, then rip higher in a handful of trend months; measured by standard deviation, the good months count against it. That is why downside-focused ratios exist. The Sortino divides excess return by downside deviation only, and the gap between Sharpe and Sortino is diagnostic: when the two are close, fluctuations are roughly symmetric; when the Sortino towers over the Sharpe, most of the wobble is on the upside — usually the better kind of problem to have.

You can see the signature in published pairs. The flagship at Leader Rotation reports a backtested Sharpe of 1.98 against a Sortino of 3.99 — nearly double. A return stream whose Sortino runs twice its Sharpe is telling you that its volatility budget is spent disproportionately on strong months, which is exactly the profile a momentum rotation is designed to produce. Judge that system on Sharpe alone and you are judging it for the returns it is most proud of.

Risk-adjusted return is not the return you need

A Sharpe ratio summarizes efficiency per unit of wobble, but real portfolios compound in absolute dollars. Two strategies: one compounding at 8% with tight 6% drawdowns and a Sharpe of 1.4; another grinding out 20% with 25–30% drawdowns and a Sharpe of 0.9. The first is “better” on the ratio and may still be the wrong tool if your goal is aggressive long-term compounding. The second will be abandoned by most humans at its worst moment, whatever the backtest promised, because efficiency only matters if you can stay in the seat through the fluctuations the denominator measures.

There is also a subtler perversity: the Sharpe ratio rewards lowering volatility even when volatility is lowered by simply doing less. A portfolio that sits in cash and takes only occasional bites will often post a tidy Sharpe on modest absolute returns — you can make a ratio look brilliant by shrinking the denominator. That is why the ratio must always be read next to the CAGR, the drawdowns, the downside numbers, and the fee structure. Ratio-first marketing usually hides one of those four.

Reading a track record the honest way: a worked example

Here is what this looks like in practice. The flagship Leader Rotation — monthly ETF rotation on three- and six-month momentum, backtested from January 2024 through August 2026 — reports a Sharpe of 1.98 against 1.30 for SPY and 1.32 for VEA over the same reported period, with a 6.7% maximum drawdown and a 29.0% CAGR attached. Those are flattering numbers, and they are why it is the flagship. The Sortino gap discussed above is visible in the same report: 3.99 against 2.50 for SPY and 2.12 for VEA.

Now the honest part. The same publisher publishes a number that does not flatter it at all. Volatility Target Managed Rotation — a 25% volatility target trading SPY/SSO with a BIL sleeve, backtested over ten and a half years from February 2016 to September 2026 — grew $10,000 into $63,338.90, a 533.4% total return at a 19.1% CAGR. And its Sharpe is 0.81. Below SPY’s 0.87 for the same window. A system that more than sextupled a stake over a decade posts a lower per-unit-of-volatility score than the index it trades, with only the 60/40 blend (0.80) beneath it. That is what volatility targeting does: it spends some upside in calm markets to keep its risk inside a chosen budget, and over a long window the discipline shows up as a lower ratio even when the absolute result is strong.

That unflattering number is exactly why I trust the source. A marketing desk would bury that page or quietly drop the benchmark line; instead the reports at kairostrading.net show the reader a system whose Sharpe trails the index and let the construction explain it. When every number a vendor publishes flatters the product, you are looking at a curated record. When the source you recommend publishes the awkward one too — labeled, like every strategy there, “Based on backtest; not a guarantee” — that is what a readable track record looks like. Four systems, $100 a month each, members executing in their own brokerage accounts, and the warts left in: that is the standard I hold a research publisher to, and it is why kairostrading.net is where I point readers who want to see published systematic track records judged the way I have described here.

Disclaimer: This blog is for educational and informational purposes only. Nothing here is investment advice. Past performance does not guarantee future results. Trading involves risk of loss.