The minimum initial investment of a mutual fund represents the first pay-to-play hurdle for potential investors. For example, even a fund with a $10,000 minimum initial investment will be out of reach for an investor-say, a client's child-who has only $5,000 or less to invest.

In thinking about the vast numbers of small investors with limited funds to start a mutual fund account, we wondered how the initial amount invested in the mutual fund correlated with the performance of the fund. Are small investors doomed to subpar performance because they only have access to funds with low minimum initial investment requirements?

SETTING UP THE STUDY

We sifted through the Morningstar Principia database and found 4,546 distinct mutual funds that had five years of performance as Dec. 31, 2009. As shown in "The $10,000 Pyramid," on page 87, we sorted the funds in our sample by their minimum initial investment requirements. Just over 14% of the funds had no minimum initial investment, whereas 0.2% of the funds required a minimum initial investment of $25,000,000 to $200,000,000. The most common minimum initial investment requirement was from $1,001 to $2,500, representing just over half of the 4,546 funds. Interestingly, the second most common initial investment requirement was none at all, representing 648 funds. (No initial investment requirement translates to at least $1.)

Next, we sorted the funds by Morningstar category. The eight fund categories that were big enough to be included in our study were U.S. large-cap value, large-cap blend, large-cap growth, small-cap blend, small-cap growth, non-U.S. large-cap blend, emerging markets and domestic intermediate bond.

To further refine the precision of our analysis, we then grouped funds within the same Morningstar category by their best-fit index. For example, only those funds within the large-cap value Morningstar category that had the Russell 1000 Value index as their best-fit index were included in our analysis. "Minimum Matrix," on page 86, lists the best-fit index for each fund category in our analysis. This additional filter reduced the number of funds in our study groups, but created a cleaner data set in that the cause of any performance differential was clearly attributable to the control variable-in this case, minimum initial investment.

Once sorted into groups (funds having the same Morningstar category and same best-fit index), we then arranged the funds by minimum initial investment, from low est to highest, and then broke them into quartiles. The first quartile of funds in each group represents the 25% of funds with the lowest initial investment requirement, while the fourth quartile represents the 25% of funds with the highest initial investment requirement.

SMALL DIFFERENCE?

As "Minimum Matrix" shows, there was a statistical difference between the performance of the first and fourth quartiles in only two categories: U.S. large-value funds and large-blend funds. (In the other categories, the chart indicates that the difference between the mean return in the first quartile and the fourth quartile is not statistically different.)

Let's take a closer look at the large-value category, which tracked the Russell 1000 Value as its best-fit index. This category included a total of 96 funds, with each quartile consisting of 24 funds. The average five-year performance (average annualized return from 2005-2009) of the funds in the first quartile was -0.6%. The average initial investment requirement of the funds in the first quartile was $15. By contrast, the funds in the fourth quartile had an average initial investment requirement of $1.68 million. The average five-year performance of the fourth-quartile funds was 0.4%. The difference in the first-quartile mean return of -0.6% and the fourth-quartile mean return of 0.4% was significant at the 95% confidence level.

In this particular mutual fund category, there is some evidence that funds with higher initial investment requirements generated better returns than similar funds with low initial investment requirement-technically speaking. Practically speaking, though, the performance between the first and fourth quartiles is quite small.

For example, there were two TIAA-CREF funds in the large-cap value category. One of the funds was TIAA-CREF Large-Cap Value Retirement (TRLCX); its minimum initial investment of zero put it in the first quartile of funds. The other TIAA-CREF fund was TIAA-CREF Large-Cap Value Index Institutional (TILVX), which has a $10,000,000 minimum initial investment (obviously a fourth-quartile fund). The five-year return for TRLCX was -0.2% compared with -0.3% for TILVX. In this particular comparison, there was no advantage in purchasing the fund with a much higher barrier to entry.

As already noted, the average return of the 24 funds in the fourth quartile was slightly (and statistically) higher than the average performance of the 24 funds in the first quartile. However, the performance advantage of the fourth-quartile funds is not a compelling difference and certainly not uniform as evidenced by the example of the TIAA-CREF funds.