On the Risk of Stocks in the Long Run: A Comment

Abstract

In a recent article in the Financial Analysts Journal, Zvi Bodie [1995] uses a clever insurance paradigm as the justification for assessing stocks’ risk as a function of the investment horizon. He concludes that stocks’ risk increases monotonically with the investment horizon. This is counter-intuitive for most people. Bodie then provides several important conclusions. One is that, all else equal, young investors should have lower equity allocations than old investors. Another is that large equity allocation proportions are not a good idea for guarantors of annuities, such as the PBGC. Finally, Bodie notes that his finding is inconsistent with the idea that equities are a good inflation hedge. Bodie’s investment conclusions appear to hinge on his proof that stocks’ risk increases with the investment horizon. But Bodie’s methodology also can be used to prove that any asset class’s risk increases with the investment horizon. Thus, if Bodie’s methodology is accepted, then whatever are the implications for stocks also apply to virtually all assets. This paradoxical state of affairs justifies a re-examination of Bodie’s methodology and its implications for his conclusions. We show that diametrically opposite conclusions about stocks’ long-term risk follow from Bodie’s methodology. Bodie’s con-clusion that stock’s risk increases monotonically with the investment horizon is incorrect. Conventional wisdom about stock’s long-term risk has not been refuted.