Measuring Investor Outcomes

ABSTRACT Should measures of investment returns focus on representative or non‐representative investors? Should we consider only the magnitude of accumulated portfolio value relative to investment, or should we highlight the series of withdrawals to fund real activities that an investment can sustain? Should we ignore the importance of the time ordering of returns (distinct from the overall level), or should we consider “return sequence risk”? I argue herein that we should be thinking more about these issues, and in particular should be less focused on arithmetic means of short horizon returns, as employed in Sharpe ratios, alphas, and portfolio comparisons. I discuss some candidate methods, and highlight the need for expanded econometric tools to allow inference regarding alternative measures of returns.