Working Paper: NBER ID: w16219
Authors: Ian Martin
Abstract: 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.
Keywords: No keywords provided
JEL Codes: D53; G1; Q54
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
Cause | Effect |
---|---|
expected time and risk-adjusted cumulative return on any asset equals one at all horizons (G12) | realized time and risk-adjusted cumulative return tends to zero (G17) |
extreme events (good news at asset level or bad news at aggregate level) (G14) | value of long-dated assets (G19) |
Sharpe ratio being higher than volatility (G17) | bad news is significant driver of asset valuations (G14) |
expected risk-adjusted return tends to zero (G40) | rare extreme events cause temporary explosions in asset prices (E32) |