The Importance of Asset Allocation 53
the portfolio's value. If a manager is particularly bullish about stocks, she will increase the allocation of stocks toward the 80 percent upper end of the equity range and decrease bonds toward the 20 percent lower end of the bond range. Should she be more optimistic about bonds, that manager may shift the allocation closer to 40 percent of the funds invested in bonds with the remainder in equities.
A review of historical data and empirical studies provides strong support for the contention that the asset allocation decision is a critical component of the portfolio management process. In general, four decisions are made when constructing an investment strategy:
V What asset classes should be considered for investment?
V What normal or policy weights should be assigned to each eligible asset class?
V What are the allowable allocation ranges based on policy weights?
V What specific securities should be purchased for the portfolio?
The asset allocation decision comprises the first two points. How important is the asset allocation decision to an investor? In a word, very. Several studies have examined the effect of the normal policy weights on investment performance, using data from both pension funds and mutual funds, from periods of time extending from the early 1970s to the late 1990s.8 The studies all found similar results: About 90 percent of a fund's returns over time can be explained by its target asset allocation policy. Exhibit 2.7 shows the relationship between returns on the target or policy portfolio allocation and actual returns on a sample mutual fund.
Rather than looking at just one fund and how the target asset allocation determines its returns, some studies have looked at how much the asset allocation policy affects returns on a variety of funds with different target weights. For example, Ibbotson and Kaplan (see Footnote 8) found that, across a sample of funds, about 40 percent of the difference in fund returns is explained by differences in asset allocation policy. And what does asset allocation tell us about the level of a particular fund's returns? The studies by Brinson and colleagues and Ibbotson and Kaplan (Footnote 8) answered that question as well. They divided the policy return (what the fund return would have been had it been invested in indexes at the policy weights) by the actual fund return (which includes the effects of varying from the policy weights and security selection). Thus, a fund that was passively invested at the target weights would have a ratio value of 1.0, or 100 percent. A fund managed by someone with skill in market timing (for moving in and out of asset classes) and security selection would have a ratio less than 1.0 (or less than 100 percent); the manager's skill would result in a policy return less than the actual fund return. The studies showed the opposite: The policy return/actual return ratio averaged over 1.0, showing that asset allocation explains slightly more than 100 percent of the level of a fund's returns. Because of market efficiency, fund managers practicing market timing and security selection, on average, have difficulty surpassing passively invested index returns, after taking into account the expenses and fees of investing.
Thus, asset allocation is a very important decision. Across all funds, the asset allocation decision explains an average of 40 percent of the variation in fund returns. For a single fund, asset allocation explains 90 percent of the fund's variation in returns over time and slightly more than 100 percent of the average fund's level of return.
Good investment managers may add some value to portfolio performance, but the major source of investment return—and risk—over time is the asset allocation decision. Investors who
8Findings discussed in this section are based on Roger G. Ibbotson and Paul D. Kaplan, "Does Asset Allocation Policy Explain 40, 90, or 100 Percent of Performance?" Financial Analysts Journal 56, no. 1 (January-February 2000): 26-33; Gary P. Brinson, Brian D. Singer, and Gilbert L. Beebower, "Determinants of Portfolio Performance II: An Update," Financial Analysts Journal 47, no. 3 (May-June 1991): 40-48; Gary P. Brinson, L. Randolph Hood, and Gilbert L. Bee-bower, "Determinants of Portfolio Performance," Financial Analysts Journal 42, no. 4 (July-August 1986): 39-48.
Was this article helpful?