Diversification Bias and the Law of One Price: An Experiment on Index Mutual Funds

ABSTRACT

Individual investors select high-fee index mutual funds despite the fact that the future payouts are nearly identical. The authors offer an explanation for this violation of the law of one price based on investor desire to diversify. While diversification in some settings may be beneficial, in the case of assets with identical payouts, fee minimization is the only rational strategy. The evidence confirms that investors diversify by selecting multiple higher fee funds rather than minimizing fees when investing in index mutual funds.

KEYWORDS:

• Mutual funds
• S&P index funds
• Behavioral finance
• Diversification bias

Notes

1. Actual payouts may differ slightly due to tracking error.

2. Diversification bias differs from the diversification heuristic in that a bias implies systematic deviations from optimal decisions, while the diversification heuristic remains compatible with traditional economic ideas of rationality in that they can be utility maximizing.

3. A fact sheet can be found at http://us.spindices.com/indices/equity/sp-500.

4. Marketing has been put forth as another potential explanation for the violation of the law of one price in index funds (Khorana and Servaes [2012]).

5. Participants consisted of master of accountancy, master of finance, and master of business administration students (the master of business administration students were in their capstone course). Despite the fact that our task was very low in integrative complexity, we included finance and accounting students because we wanted to provide a strong test of our hypotheses (Elliott et al. [2007]). The use of these students likely biases against our results and accordingly may explain the lack of an effect for education.

6. None of the demographic variables differ significantly between experimental conditions (ps > 0.10), indicating adequate randomization of participants.

7. We constructed a fee by allocating a hypothetical $10,000 in accordance with participants' allocations. The resulting dollar amounts were then multiplied by the fee to get the estimated total fee. For example, in the equal fee condition (fee equal 2.6%), an allocation of 25% to each of the four funds would produce a total fee of $260.

8. Post experimental questions suggest that the educational materials effectively conveyed the desired information. Those with different fees and historical returns also better understood the implications of those fees and returns when they were provided with Securities and Exchange Commission educational literature than when literature was not provided (F = 3.25, p = 0.08). Participants who were provided with the SEC educational literature were also better able to recognize that historical returns do not predict future returns than those without the literature (F = 12.0, p < 0.01).