C.W.K.
Stream
Lesson 05 of 08 · published

Sharpe ratio — return per unit of shake

~25 min · sharpe

Level 0Numeracy Apprentice
0 XP0/68 lessons0/14 achievements
0/100 XP to next level100 XP to go0% complete

Return per unit of shake

From Track 3 lesson 6: r_f is the anchor. From Track 3 lesson 2: σ measures volatility. Combine them and you get the most-used "risk-adjusted return" measure in finance:

Sharpe ratio = portfolio's excess return over the risk-free rate, divided by the portfolio's σ. It tells you "how much extra return did this portfolio generate per unit of risk it took?" Higher = better.

Named after William Sharpe (1990 Nobel laureate, alongside Markowitz and Merton Miller). It's the most-cited single statistic when comparing investment performance.

Why "excess return" matters

You don't compare portfolios on raw return — you compare on excess return over what you'd have earned doing nothing risky (parking money in T-bills). A 10% return when r_f is 4% gives you 6% of "extra" — that's what risk earned you. The 4% you'd have earned anyway.

So Sharpe ratio asks: "for the σ you took, did you earn enough excess return to be worth it?" The answer depends on what counts as "enough" — usually compared to other portfolios.

Sharpe ratio interpretations

Rough benchmarks (annualized):

  • Sharpe < 0.5: not great. The risk wasn't well-rewarded.
  • Sharpe 0.5-1.0: acceptable for many real-world portfolios.
  • Sharpe 1.0-2.0: good, often excellent.
  • Sharpe > 2.0: exceptional. Sometimes too good to be true (overfit, lucky, or fraudulent).
  • Sharpe > 3.0 sustained: very rare; treat with skepticism.

Numbers depend on time period and methodology. Bull markets boost Sharpe ratios; bear markets crush them. Strategies that quietly accumulate small gains can show high Sharpe ratios that disappear in tail events (Madoff's fund showed unbelievably high Sharpe — turned out to be fraud).

Why Sharpe is the natural performance measure

From lesson 9-1: the Sharpe ratio is the slope of the Capital Allocation Line. Higher Sharpe portfolio = steeper CAL = better risk-return tradeoff. So Sharpe directly answers "which portfolio is closer to the efficient frontier?"

It's also scale-invariant. A portfolio that returns 6% with σ = 6% has the same Sharpe as one returning 12% with σ = 12% (assuming r_f = 0). Both deliver the same return per unit of σ. Mixing the second with r_f at 50/50 gives the same risk-return as the first.

Limitations

Sharpe assumes σ is the right measure of risk. For some assets and strategies, σ doesn't capture the full picture:

  • Skewed returns. Strategies with frequent small gains and rare large losses can have high Sharpe but real tail risk (e.g., selling deep OTM options).
  • Non-normal distributions. Real returns aren't normal; Sharpe under-penalizes fat-tail strategies.
  • Time-varying risk. Sharpe assumes constant σ; real volatility shifts.

Variants exist — Sortino ratio (only penalizes downside σ), Calmar ratio (uses max drawdown), etc. — but Sharpe is the universal default.

The takeaway

Sharpe ratio = (R − r_f) / σ. Excess return per unit of σ. The slope of the CAL. Higher Sharpe = closer to the efficient frontier. The most-used risk-adjusted performance measure. Limitations: σ doesn't capture all risk, especially for tail-risk strategies. Despite imperfections, it's the universal benchmark for comparing portfolios.

Exercise

  1. Portfolio A returns 12% with σ = 18%. Portfolio B returns 8% with σ = 8%. r_f = 3%. Sharpe ratios?
  2. Which portfolio is closer to the efficient frontier (per Sharpe)?
  3. Why does selling deep OTM puts on a stable index give a misleadingly high Sharpe ratio?
  4. What does it mean if a portfolio has Sharpe 0 — is it earning a return?

Progress

Progress is local-only — sign in to sync across devices.
Spotted a bug or have feedback on this page?Report an Issue

Comments 0

🔔 Reply notifications (sign in)
Sign inPlease sign in to comment.

No comments yet — be the first.