Explain how reported performance can be gamed, and why standard performance metrics become unreliable as a result.
The idea
Reported performance can be manipulated, and the standard measures then mislead. Window dressing means trading just before the reporting date so that the portfolio holds recent winners. Valuing illiquid assets with stale or smoothed prices understates reported volatility and correlation with the market, which inflates Sharpe and Treynor figures. Selling out-of-the-money options earns a steady premium with rare large losses, so a short track record can show a high Sharpe ratio before the crash arrives. The standard deviation misses the skewed risk. Survivorship bias and backfilling mean databases hold only the funds that survived or chose to report, which overstates average returns. The Sharpe ratio, alpha and beta assume roughly normal, honestly priced returns, so they can be raised without adding value. Treat unusually smooth, high-Sharpe records with suspicion.
A practice question
Window dressing by a fund manager refers to:
- valuing illiquid holdings with modelled prices to reduce reported volatility.Why not: This is return smoothing through stale or modelled valuations, not window dressing.
- excluding liquidated funds from a published performance database.Why not: This is survivorship bias in performance databases, not window dressing.
- selling options to earn a stable premium and boost Sharpe ratios.Why not: This is option-selling, a risk-hiding strategy, not window dressing.
- trading just before reporting dates so that the disclosed holdings or returns look better. - correct
Window dressing is cosmetic trading around report dates, for example selling losers and buying recent winners to improve how the portfolio appears.
A harder one
A database holds 100 funds at the start of a year. The 80 funds that survived averaged 9% and the 20 that were liquidated averaged -6%. By how many percentage points does the survivors-only average overstate the average for all 100 funds?
- 15.0Why not: The gap between the two group averages, not the effect on the overall mean.
- 3.0 - correct
- 1.2Why not: Weights only the liquidated group's return (20% x 6%).
- 1.8Why not: Weights the survivors' return by the liquidated share (20% x 9%).
All-fund average = 0.8 x 9% + 0.2 x (-6%) = 6.0%. The survivors-only average of 9% overstates by 3.0 percentage points = 0.2 x (9 - (-6)).
The hardest tier
An illiquid fund's true monthly returns are independent with a standard deviation of 6%. Because of stale valuations, each reported return is 0.6 times the current true return plus 0.4 times the previous true return. If the mean is unchanged, the reported Sharpe ratio is overstated by a factor of closest to:
- 1.67Why not: Uses 1/0.6, the weight on the current return, as the volatility scale.
- 1.92Why not: Uses the variance ratio 1/0.52 instead of the standard deviation ratio.
- 1.39 - correct
- 1.00Why not: Assumes weights summing to 1 leave the volatility unchanged.
Reported variance = (0.6² + 0.4²) x 36 = 0.52 x 36, so reported standard deviation = 6% x sqrt(0.52) = 4.33%. Sharpe is overstated by 6/4.33 = 1.39.
In the app: 6 practice questions on this outcome, spaced review that brings it back before you forget it.
Start free →More in Portfolio Performance Evaluation
- Distinguish a portfolio's time-weighted return from its dollar-weighted return and explain when each is the right one to use.
- Characterise risk-adjusted measures including Sharpe's measure, Treynor's measure, Jensen's measure (Jensen's alpha) and the information ratio, and pinpoint when each fits best.
- Outline what Modigliani-squared and Treynor's measure are used for when comparing two portfolios, and how each is shown on a graph.
- Use the standard error and t-statistic to decide whether a performance measure is statistically significant.
- Characterise what style analysis is.
- Account for why it is hard to gauge the performance of actively managed portfolios.
- Outline regression and call option model techniques for gauging a fund manager's market timing skill, and work out the return that a manager earns from timing.
- Describe how performance attribution is done and apply it to split results into asset allocation, sector and security selection effects and their combined total.