Measuring Investment Risk Based on Tail Thickness

Abstract

In recent years, both institutional andindividual investors have come to rely heavilyupon techniques for analyzing (defining andmeasuring) risk. In this respect, the issue thatcontinues to require the attention of academicresearchers and practitioners alike is how toconcisely define investment risk and, moreimportantly, how to best measure it. Selecting anappropriate risk definition involves trade-offsamong ease of measurement, forecast ability, andintuition of individual investors. The purpose ofthis paper is to present an alternative index formeasuring unconditional (or total) risk. Theproposed measure reflects behavior in general, andthickness in particular, of the lower tail of thedistribution of returns. We therefore argue itprovides a more useful and reasonable indexbecause, unlike measures frequently used, itsestimation depends upon the most relevant datafrom the sample distribution. We describe riskanalysis based on lower tail behavior and identifyits advantages over existing methods. Finally,using data of weekly returns to the CREF StockFund, we provide an empirical example toillustrate the technique.