Three Keys to Retirement Investing

Investors Aren't Pensions Back in the day, one couldn't attend an investment conference without hearing that asset allocation accounted for 90% of portfolio results. In 1986, Brinson, Hood, and Beebower published "Determinants of Portfolio Performance" (hereafter BHB), which found asset allocation to be much more important than security selection. Word spread like a prairie fire.

(However, the word was not quite correct. As later observers pointed out, the authors chose but one of several reasonable approaches to measuring the "determinants." Had they used other methods, they would found different answers. Within the institutional-investment community, the paper became a mildly infamous example of how research studies can be oversold.)

There were two problems with applying that lesson to individual retirement planning.

One, as the article was written for pension fund managers, it ignored the effect of contributions. Pension fund managers work with what is given to them. Individual investors do not. True, their contribution decisions are constrained, often severely, by the existence of other financial needs. But within reason, individuals have another tool for affecting their investment outcomes besides those that were considered by BHB.

Indeed, unless they hold very cheap funds, they have two tools, the other being costs. Which brings us to problem number 2. Effectively, BHB's study ignored expenses. (The bottom of this column explains how that happened.) You, of course, should not. As Jack Bogle wrote nearly 20 years ago, "Is performance determined by asset allocation or by cost? Both." To which we can add contribution rates, which lay outside the scope of Bogle's speech.

By the Numbers Let's take a look at those three items of asset allocation, costs, and contribution rates. Morningstar's Hal Ratner calculated how a balanced index would have performed over rolling 15-year periods starting from 1926 through 2001. We can use his program to see how the figures change as we adjust each of those items.

(The details: Ratner invested $1,000 at the start of each time period, with each percent of contribution rate representing $500, such that a 2% contribution rate is $1,000 per year and 10% would be $5,000. All monies are invested at the beginning of the year. The amounts shown are the average ending values for each of the 15-year calculations, expressed in real teams, and shown in thousands of dollars.)

A sample of the output appears below. It shows the average ending values for two asset allocations, 40% stock and 70% stock; two contribution rates, 4% per year and 7% per year; and two expense regimes, with "low" being a portfolio that is 100 basis points (that is, 1 percentage point) cheaper per year than "high."

The Big Picture Note: The relative importance of the three items will vary depending upon how the study is structured. For example, a higher initial balance, relative to contributions, will increase the importance of costs. But the very general pattern is clear. The contribution rate is paramount. Investors can't asset-allocate and low-cost their way to retirement success. They must accompany their good investment decisions with cold, hard cash.

After contributions, asset allocation and costs appear to be of roughly similar importance--very roughly. Once again, the initial finding could stand some further explanation. The gain from asset allocation has the significant drawback of coming with greater risk. Moving from 40% in stocks to 70% in stocks does boost expected wealth, but not without additional bumps along the way--and the occasional poor ending value. There is no catch to cutting costs, though.

The public discussion of investing suggests a different order. Although things are changing, historically fund companies, advisors, and the media have spent more words on asset allocation--and more yet on security selection--than on either costs and contribution rates. Such is the nature of the beast. More can be said about technical matters than on items as simple as saving more and spending less, particularly when the people doing the saying are technicians.

Also, costs and contribution rates can be awkward issues. Understandably, those who are paid to give investment advice can be reluctant to pound that table too loudly, unless (as with Vanguard) cost is among their competitive advantages. Contribution rates present the opposite problem. Because advice givers profit from additional assets, they must be careful when advocating higher savings rates lest they be seen as acting in self-interest.

Asset-allocation recommendations, in contrast, are mostly drama-free. True, some contend that fund companies overly favor stocks for their target-date funds, so that they can draw the higher management fees that accompany equity funds. That charge has not stuck, for good reason: It doesn't wash. Even if fund managers were so venal, the modest additional income that would be gained by such a strategy would be more than offset by the damage caused by a bad allocation. Ask Oppenheimer. Its target-date funds are no more, after being caught out with overall aggressive allocations (and security selections) in 2008.

Addendum The BHB article examined how a pension fund's actual performance compared with the theoretical performance of a composite benchmark, with the benchmark being created from the fund's average asset allocation over the 10-year period. Thus, if a fund had the admittedly unlikely (but pleasantly round) averages of 60% large-company U.S. stock, 30% investment-grade bonds, and 10% cash over that decade, then it would be measured by a 60/30/10 benchmark.

But--and here's the punch line--BHB's calculation captured sources of variance, not sources of return. To put the matter more plainly, BHB sought to explain why a fund might outleg the index in one quarter and then trail the next quarter. Costs were never the answer to that particular question, because costs are constant. The BHB study was structured to find only items that changed, not those that stayed the same.

That subtlety escaped a great many MBAs and CFAs for a great many years.

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.