**Shortfall risk goes down over long time horizons**. If there’s an equity risk premium then the probably of a shortfall relative to the risk-free rate declines over time. This shortfall risk definition (or odds of success) is what many Monte Carlo simulations seem to use.

**…but loss severity risk goes up**. Quantified as the cost of insuring against a shortfall or more explicitly the price of a put option that guarantees at least the amount you would have got if you had invested in a treasury bond with a comparable maturity. For the S&P500 for example, the ‘insurance cost’ for a 25 year time horizon is about 6 times more expensive than the cost of a 1 year time horizon (as I understand). So loss severity risk goes up over time even though shortfall risk goes down. So time diversification does not reduce risk as much as is often assumed.

**Implications?**

- rely more on asset class diversification to reduce risk, not time.

- direct holdings of treasury inflation protected securities held to maturity as fixed income holdings, rather than bond mutual funds can reduce ‘loss severity risk’ by ensuring a ‘guaranteed’ real return for at least part of a portfolio.

Just my understanding - any other interpretations?

Robert

.

(Zvi Bodie highlights this point in “The Next Generation of Life-Cycle Investment Products “ for those interested here is the link to a webcast of this presentation http://www.cfawebcasts.org/cpe/what.cfm?test_id=698)

.