Sometimes, More Is Indeed More
Accumulators win by keeping things simple. In the mutual fund industry, more has been less. The best funds have been those that have been the most straightforward--stock, bond, and balanced funds that have transparent investment schemes that are readily understood by their investors. Such funds pretty much go where their markets take them (in today's parlance, their returns are mostly "beta" rather than "alpha"), and they deliver few surprises. They are run at a low cost.
In contrast, high-concept funds have not served investors well. From "government plus" and option-income in the 1980s through various exotic bonds in the 1990s to market-neutral and 130/30 strategies a decade ago, complex funds have not shown staying power. They arrive with a roar, being heavily marketed and wrapped in expensive packaging, they deliver mediocre-to-poor returns, and they depart a few years later with barely a whimper.
Neither do most accumulators require intricate planning strategies. Invest early, invest often, trade infrequently, keep a close watch on costs and taxes, and let time do its thing. That's pretty much it for the rank-and-file accumulator.
Investing during retirement, however, is another matter--a situation wherein more might indeed be more. Some things don't change before and after retirement. Low costs and funds that use transparent investment schemes remain attractive. However, the combination of a shrinking time horizon and the need to make withdrawals complicates the retiree's planning strategies. Buy-and-forget no longer suffices. The retiree's situation is more dynamic than that of the accumulator, demanding more attention and greater flexibility in response to market movements.
Varying Amounts of Variable Annuities
Such, implicitly, is the argument of Wharton's Olivia Mitchell (who was kind enough to send me the paper) and three faculty members of Goethe University Frankfurt, in "Variable Payout Annuities and Dynamic Portfolio Choice in Retirement." In the article, the authors add two levels of complexity to the retiree's situation. First, they advocate using variable annuity funds rather than plain-vanilla mutual funds. Second, they treat the annuitization decision as being gradual, with the investor purchasing additional variable annuity funds during retirement.
As the authors acknowledge, many previous papers have studied the effects of adding fixed-payout annuities to a retiree's portfolio. Whether fixed or variable, goes the argument, annuities are attractive to retirees because they offer a survival credit. That is, because annuity owners who die at an earlier age effectively pass on a portion of their assets to those who die later, annuities distribute at a higher rate than retirees can achieve by investing on their own. (With enough money, one could emulate a standard mutual fund by investing in the same securities that it holds, in the same proportions. One cannot emulate an annuity in a similar fashion because it contains features that can be obtained only through pooling.)
Nor are the authors the first to extend the argument from fixed-payout annuities, which have an unchanging income stream that does not benefit from market gains, to variable annuities, which can grow over time in both value and income. More articles have been written about fixed annuities, as they are the older of the two investments, but variable annuities have also come in for attention in recent years.
However, the paper gives the most thorough treatment of the dynamic purchase of annuities, that is, the strategy of easing into annuities during the retirement period. Previous articles had treated the subject as an on/off switch or had greatly simplified the dynamic analysis. At the start of retirement, the retiree had the option of purchasing a certain amount of variable annuities, but, after that, the path was pretty much fixed. That clearly is the easier strategy to model than the one of ongoing purchases--but, the authors argue, it is not the better strategy.
The authors find that using variable annuities rather than conventional mutual funds (or directly held stocks and bonds), and doing so dynamically, is better for a variety of retiree investor types. They model low, moderate, and high wealth investors with low, moderate, and high levels of risk tolerance and find variable annuities to be the superior option in each of the nine cases. They also add a moderate-size bequest motive (that is, inheritance). Once again, variable annuities perform better.
That doesn't mean that I would rush out to buy annuities, were I retired. For one, I haven't gone through the paper's math (meaning, in this paper's case, that I haven't asked one of Morningstar's Ph.D.s to check the formulas). In addition, there are considerations that are outside the scope of the paper, such as the potentially higher costs of holding annuities rather than mutual funds and the psychological difficulty of giving up an asset to be annuitized. However, it's an intriguing thesis. I'll be giving it future thought.
Also, the paper serves as a useful example of how the investment framework changes after retirement. The number of moving parts increases, as does the need for investor flexibility. Investing during retirement is a different problem than is investing for accumulation. Previous lessons cannot automatically be ported.
John Rekenthaler has been researching the fund industry since 1988. He is now a columnist for Morningstar.com and a member of Morningstar's investment research department. John is quick to point out that while Morningstar typically agrees with the views of the Rekenthaler Report, his views are his own.