[DROP BELOW -- TOO MUCH DETAIL?] - BUT REVIEW ANYWAY.

PERHAPS USE AS COMMENT PAGE FROM INVESTMENT POLICY AND IRSK? --- Second, the way we approach this is different from some of the other studies you may have read. The simple fact of the matter is that there is no clearly superior way to approach post-retirement investing issues. Let's briefly look at some other approaches. One starts with your risk tolerance, and uses it to calculate your "optimal" asset allocation. Based on your relative preference for income versus leaving a bequest, it then divides your assets between annuities (both fixed rate and variable) and an investment portfolio that holds a range of asset classes. The higher your preference for current income, the greater your allocation to annuities. Our problem with this approach is that in real life, things don't usually work this way. In our experience, people start with their current savings and an uncertain estimate of their future needs (e.g., will I or won't I need long-term care?) and their expected life. Using these they derive the asset allocation that will maximize the probability of achieving their income and savings goals. If this probability isn't high enough to make them comfortable, they then evaluate a range of other options (e.g., work part-time, reduce bequest, reduce income, etc.). In addition, in our experience, people's preference for annuities also depends on the current state of their health, as well as their preference for income relative to savings and bequests. Someone who doesn't expect to live very long tends to have a much lower preference for annuitization than this approach assumes.

Another approach to post-retirement investing assumes a fixed time period (say, 30 years), and asset allocation (say, 50% to domestic equities, and 50% to domestic bonds), and then estimates the maximum percent of the portfolio's value that can be withdrawn each year and still produce a 95% chance that something still will be left at the end of the period. To save you the trouble of reading them, we'll tell you the answer usually given by studies that use this approach: about 4.0% to 4.5% (see, for example, "Retirement Savings: Choosing a Withdrawal Rate That is Sustainable" by Cooley, Hubbard, and Walz, or "Making Retirement Income Last a Lifetime" by Ameriks, Veres, and Warshawsky). From our perspective, this approach has a number of shortcomings. First, its maximum time period (usually thirty years) is probably too short, given retirees' healthier lifestyles and increased life expectancy. Second, it fails to include precautionary savings and bequest goals. Third, with only a couple of exceptions, studies using this approach fail to include the impact of partially annuitizing one's savings. Fourth, they use too few asset classes, and fail to consider the potential benefit from greater diversification. Fifth, their calculations are based only on historical asset class returns, rather than a more informative mix of historical and forward-looking estimates. Sixth, they fail to address another critical issue: whether to implement their proposed strategies using index funds, actively managed funds, or direct security holdings. And seventh, they usually ignore potentially important tax issues.

In contrast, we take the approach used by many university endowment funds. We start with the goals you want to achieve, the funds you have available, and the risk limits you would like to set. Then, given our expectation for future asset class returns, we determine the asset allocation that best meets your goals. If this isn't possible, we help you to understand the impact of changes to your income and savings goals, your annuitization decisions, and your risk limits, and to make informed trade-offs between them.