Recently, we focused on the Fidelity mutual funds that are actively managed and their over-diversification. As we wrote then, funds grow as big as they do and then have difficulty beating the market. Those funds track their respective index, yet charge an active management fee. Charlie Ellis writes an interest piece in the Financial Analysts Journal last Summer that dissects those fees.
Seen for what they really are, fees for active management are high - much higher than even the critics have recognized.
When stated as a percentage of assets, average fees do look low - a little over 1% of assets for individuals and a little less than one-half of 1% for institutional investors. But the investors already own those assets, so investment management fees should really be based on what investors are getting in the returns that managers produce. Calculated correctly, as a percentage of returns, fees no longer look low. Do the math. If returns average, say, 8% a year, then those same fees are not 1% or one-half of 1%. They are much higher - typically over 12% for individuals and 6% for institutions…
[I]nvestors should consider fees charged by active managers not as a percentage of total returns but as incremental fees versus risk-adjusted incremental returns above the market index.
Thus (correctly) stated, management fees for active management are remarkably high. Incremental fees are somewhere between 50% of incremental returns and, because a majority of active managers fall short of their chosen benchmarks, infinity. And when market returns are low, as in recent years, management fees eat up even more of an investor's return.
We can look at funds in one of three ways- first as a full alpha generating products; second as a split between an active component and a passive one; third as an incremental return over that produced by an index. In our Wednesday post, we concluded that it is too difficult for a fund holding hundreds of securities to be a fully alpha-generating vehicle. In regards to splitting the components, we wrote on January 13th Out of the Closet about closet indexing of funds. We highlighted a ratio called Active Share that measures the percentage of a mutual fund that can be attributed to independence from its respective benchmark index. For instance, the mighty Fidelity Magellan Fund has an active share of approximately 70% and a relatively low 55 basis point expense ratio. Magellan is benchmarked against the S&P 500, which one can replicate for 10 basis points. If 30% of the fund is essentially an index, then investors are paying 75 basis points for the balance that it actively managed. Still not bad. Fidelity's Growth Company Fund (MUTF:FDGRX) has an Active Share of closer to 55 with a 90 basis point fee. The FDGRX is benchmarked to the Russell 3000. Fidelity's MidCap Index funds charge 22 basis points. As such, investors are paying 1.5% for their active component. When looking at its excess returns, it may be a better value than that of Magellan.
Finally, but most importantly, we can look at performance attribution versus the index - a feature that Mr. Ellis highlighted effectively. Forget how much of the managers selection is different than his/her benchmark, but look at what are you getting for your incremental fee. The answer is not much - or said differently, you are paying a lot for modest amounts of return differential (and in some cases, underperformance). Once again we highlight the same Fidelity Funds, looking this time at fees, R-squared, Beta and Information Ratio. The Info Ratio measures a fund's active return (fund's average monthly return minus the benchmark's average monthly return) in relation to the volatility of its active returns. RR² (R-Squared) is a historical measurement, calculated in this case over 36 months, which indicates how closely a fund's fluctuations correlate with the fluctuations of its appropriate benchmark index. An RR² of 1.00 indicates perfect correlation, meaning all the fund's fluctuations were explained by fluctuations in the benchmark index, while an RR² of 0.00 indicates no correlation. For example, a fund with an R-squared of 0.80 indicates that 80% of the fund's past fluctuations were explained by changes in the benchmark index. Generally, the higher the R², the more meaningful the beta figure.
Each of the Fund's have R² greater than 90% with the exception of the somewhat aptly-named Independence Fund at 88%. The median fund had a negative Information Ratio. As for the Magellan Fund, that 75 basis points of "excess" fee doesn't look as attractive anymore. The fund trades in-line with the S&P but consistently has underperformed. The Fund has 10% of its assets in non-U.S. securities which is one contributing factor in its even modest differentiation, but looking at the following portfolio characteristics, its not surprising that it's returns are close to benchmark.
This isn't necessarily a case advocating an index approach. Rather, an investor needs to understand what they are paying for, and how they can potentially improve upon that strategy on their own. Creating a thoughtful, time-tested strategy with a reasonable mix between concentration and diversification can allow you to generate alpha without diluting yourself with fees. Try our Strategy Builder at www.BloodhoundSystem.com to produce your own.