A Framework for Portfolio Decumulation

Abstract

Before the bear market of 2008–09, the conventional wisdom was that conservative asset allocations might not provide sufficient returns for the long horizons that many retirees face. Therefore, new retirees needed significant allocations to equity. The advice often came with a call to invest more aggressively in the early years of retirement, then taper the equity exposure later in retirement. Arguments for this approach typically present results of Monte Carlo simulation showing that higher equity allocations, particularly early in retirement, can increase the probability of sustaining higher withdrawal rates from the portfolio. But these arguments fall victim to at least two fallacies.

The first fallacy is tragically ironic. Even though the advice is based on probability theory, it overlooks a primary tenet emphasized by Blaise Pascal, the founder of probability theory. Pascal realized that while probability matters, so do consequences. In fact, Pascal emphatically warned against playing the odds without considering the consequences. While it may be true that higher equity exposures can increase the likelihood of sustaining higher withdrawal rates, they also increase the magnitude of failure in unsuccessful cases. Students of investment theory may recognize this oversight as a form of “the fallacy of time-diversification of risk.” By making the portfolio more aggressive, the risk of achieving poor returns puts the entire financial plan in peril. The retiree may go broke early, face a lower standard of living, or abandon retirement altogether. So everyone who advises retirees must seriously ponder the following question: Even though higher equity allocations may raise the probability of success for a retirement income plan, does this necessarily mean that higher equity allocations are less risky?

The second fallacy involves an assumption underlying most simulation methodologies. The assumption is that the investment strategy is predetermined at the start of the investment horizon. Examples of predetermined strategies include allocations that remain static over the entire horizon or that use glide paths similar to that of target date funds, in which the equity exposure tapers over time. The only way such an assumption makes sense, however, is if the investor’s risk exposure is static over time. But for retirees who depend on nest eggs for living expenses, this assumption is never correct. Holding more-aggressive portfolios early in retirement and less-aggressive allocations later in retirement leads to the retiree taking the most investment risk precisely when it is most risky to do so — when the risk of outliving the portfolio (longevity risk) is greatest.

Keywords: Portfolio Decumulation

JEL Classification: E21, G11