The Wrong Kind Of Diversification

Asset class diversification is a good thing because it lowers the potential for a large loss. However, diversifying actively managed mutual funds within an asset class is a bad thing because it decreases the already low probability for outperforming a comparable portfolio of all index funds. If you can’t buy index funds, then pick one active fund per asset class and hope for the best.

Portfolios holding only actively managed funds have a low probability for outperforming a comparable all index fund portfolio. In A Case for Index Fund Portfolios, a newly released white paper I co-authored with Alex Benke of Betterment, we show that portfolios of actively managed funds have a low probability of outperforming similar portfolios holding only index funds.

The odds get worse when investors select two or more actively managed funds in each asset class. People do this because they believe they are diversifying their holdings. However, the data suggests the more active funds selected, the greater the odds favor an index fund portfolio.

Our study formed thousands of actively managed mutual fund portfolios using the CRSP Survivorship-Bias-Free Database. The database includes funds that had closed or merged over the years, thereby providing a true picture of investment opportunities available each year. In each test, we compared actively managed fund portfolio results to index fund portfolios that had the same asset allocation over the same time horizon.

In one scenario, we looked at portfolios holding more than one actively managed fund in each asset class to see the effect on portfolio performance. Five thousand separate portfolios were formed that had one, two, and three actively managed funds per asset class. These results were then measured against a portfolio of index funds.

Figure 1 illustrates the outcome of this scenario. It covers three asset classes (Broad US Stocks, International Stocks, and US Investment-Grade Bonds) over a 16-year period from 1997 to 2012. The active fund portfolios were then compared to an index fund portfolio using the same asset allocation. The index fund portfolio held 40% in the Vanguard Total Bond Market Index (VBMFX), 40% in the Vanguard Total Stock Market Index (VTSMX), and 20% in the Vanguard Total International Stock Index (VGTSX).

Figure 1: Probability an index fund portfolio would outperform an actively managed fund portfolio

Figure 1: Probability an index fund portfolio would outperform an actively managed fund portfolio

Source: A Case for Index Fund Portfolios (Portfolio Solutions® and Betterment)

The three-fund index fund portfolio had an 82.9% probability of outperforming the actively managed fund portfolios when only one fund was selected in each of the three asset classes. The odds increased in favor of index funds when two active funds were selected for each of the three asset classes (a total of six funds).  When three actively managed funds were selected for each of the three asset classes (a total of nine funds), the odds reached 91.0% in favor of index fund. The study has a margin of error of ±1.0%.

This finding has meaningful implications for all investors, particularly for participants in a company sponsored retirement plan that offers multiple mutual fund choices.  An all index fund portfolio has a high probability of outperforming active funds. However, if only actively managed funds are available, then it’s better to pick one fund per asset class and hope for the best.

Diversifying among actively managed funds within an asset class decreases the already low probability for beating a comparable portfolio of all index funds. If you’re able to buy index funds, then keep the number of actively managed funds down to a minimum in order to create the best odds for higher performance.

See disclosure regarding A Case for Index Fund Portfolios here.