Back to course

Performance Statistics That Matter

Tear sheets print a dozen statistics; a handful carry information. This lesson is the short list, and what each number is unreadable without.

The short list

Sharpe, with its error bars. The workhorse, read as an estimate: its uncertainty shrinks with track length and widens with skew and kurtosis, as the probabilistic Sharpe lesson quantified. A useful habit is mentally attaching ±2/Tyears\pm 2/\sqrt{T_{\text{years}}} as a rough annualised confidence band: three years at Sharpe 1.0 is consistent with a true value near zero.

Maximum drawdown, against expectation. Drawdown is often read as a personality trait; it is mostly arithmetic. A strategy with annualised volatility σ\sigma running for years should be expected to spend long stretches underwater, and the right question is whether the observed drawdown is consistent with the strategy's own volatility and Sharpe, not whether it is small. A suspiciously shallow maximum drawdown over a long span is evidence of smoothed marks or a short-volatility profile waiting to pay it all back at once.

The rest of this lesson is for subscribers

Unlock every lesson in Research Validation and Backtesting, and every other premium course.

Subscribe to continue

Test your knowledge

Questions are only available to subscribers.
Continue
NextTransaction Costs and the Paper Sharpe

Keep reading Research Validation and Backtesting

21 lessons in this course, and every other premium course, on one subscription.

Subscribe now
  • Every lesson in every course, with the worked examples and interactive simulators
  • Graded questions on every lesson, with explanations for the wrong answers as well as the right one
  • The trainers, timed assessments and brainteaser library that go with them