The Valuation of Long-Dated Assets
The expected time- and risk-adjusted cumulative return on any asset equals one at all horizons. Nonetheless, I show that a typical asset's realized time- and risk-adjusted cumulative return tends to zero almost surely. As a corollary, the value of a typical long-dated asset is driven by extreme events: either by good news at the level of the individual asset or by bad news at the aggregate level. In the case of the aggregate market, the fact that its Sharpe ratio is higher than its volatility suggests that bad news is the relevant consideration in practice.
I am grateful to David Backus, Brandon Bates, Jonathan Berk, John Campbell, John Cochrane, Peter DeMarzo, Darrell Duffie, Lars Hansen, Michael Harrison, Ilan Kremer, Steve Ross, Jeremy Stein, Dimitri Vayanos, Martin Weitzman, and seminar participants at Berkeley, LSE, and the Stanford Institute for Theoretical Economics conference on "New Quantitative Models of Asset Markets" for their comments. The views expressed herein are those of the author and do not necessarily reflect the views of the National Bureau of Economic Research.
Ian Martin, 2012. "On the Valuation of Long-Dated Assets," Journal of Political Economy, University of Chicago Press, vol. 120(2), pages 346 - 358. citation courtesy of