This 1997 grant produced a paper, "Life Cycle Investing, Holding Periods, and Risk, " which analyzed the interaction between investment time horizon and investment risks and returns of different classes of equity and fixed-income securities. The asset classes used in the study were: large capitalization stocks; small capitalization stocks; long-term government bonds; intermediate-term government bonds; corporate bonds; and treasury bills.
The authors calculated means, medians, and standard deviations for each asset class over holding periods such as 1, 5, 10, 20, and 30 years. For each holding period, they randomly drew months from the dataset, sampling with replacement, until they drew the number of months corresponding to the period of interest. The statistics provide several insights. First, there were large differences in wealth accumulation between asset classes over long holding periods. Second, the expected reward from investing in riskier versus less-risky bonds did not appear to exist for most risk preferences. There was little or no advantage to investing in long-term government bonds as opposed to intermediate-term government bonds. Ending wealth was nearly identical over most holding periods while the risk was consistently lower. Third, relying solely on standard deviations as the measure of wealth variability, at least in the case of equities, may be misleading. This was because over longer holding periods mean wealth accumulations greatly exceed median accumulations, suggesting that additional risk was mostly on the up side.
The researchers showed that the asymmetry of risk and reward increases as the holding period lengthens. Over a 20 year holding period there was a small probability that an equity investment would slightly underperform a bond investment. For example, at the 5th percentile a $1 (large capitalization) stock investment generated an accumulation 94 cents less than a comparable long-term corporate bond investment. However, at the 95th percentile the stock investment was $31.19 greater than the bond investment. Hence the risks were not at all symmetric, with relatively insignificant downside exposure and substantial upside potential.
The paper is forthcoming in the Journal of Portfolio Management.
