Spot Closet Indexers
In the investing world, an actively managed fund that clings tightly to its benchmark index is sometimes referred to as a closet index fund. Closet indexers are often scorned because their operating fees are on par with those of other actively managed funds, which means that their expenses are generally much higher than those of index funds. If a closet indexer's returns are similar but its fees are much higher, why not just invest with a real index fund?
For investors, it has historically been tough to spot closet indexers. But that may soon change, thanks to two Yale professors who have come up with a way to measure the portion of a fund that is actively managed. At late June's Morningstar Investment Conference, in Chicago, Antti Petajisto of the Yale School of Management explained the concept of "active share" and discussed its implications for fund investors. Morningstar plans to begin incorporating the measure into its fund data later this year.
The calculation is not particularly complicated. To derive a fund's active-management quotient, Petajisto and Professor Martijn Cremers determine the portion of portfolio holdings reported in SEC filings that differ from holdings in its benchmark index. For example, if a fund holds 500 stocks and only 100 of those differ from its benchmark index, the fund's active share is 20%. The professors also examine the extent to which a holding that is in a particular index deviates from the security's weighting in the benchmark. So, for example, if ExxonMobil represents 3.4% of Standard & Poor’s 500-stock index and a fund has 0.5% of its assets in the oil giant, the fund has an "active short position" of 2.9% in the oil giant. If a fund has 6.3% of assets in ExxonMobil, it has an "active long position" in the stock of 2.9%. The sum of all active short positions plus all active long positions divided by two is the fund's "active share."
Petajisto and Cremers define a closet index fund as one in which the active share is less than 60%. In a study that examined data from 266 domestic stock funds between 1980 and the end of 2003, the professors found that the average fund using the S&P 500 as a benchmark had an active share of 66%.
To illustrate the concept, Petajisto discussed the evolution of Fidelity Magellan (symbol FMAGX) from a fund with high active share to a closet index fund. (To be fair, the study used data from the period before Harry Lange took the fund's reigns in October 2005 and overhauled its portfolio.) In the 1980s, under manager Peter Lynch, Magellan's active share versus the S&P 500 index was primarily in the 70% to 90% range. It stayed above 60% under subsequent managers Morris Smith and Jeffrey Vinik. But when Robert Stansky took Magellan's helm in June 1996, its active share began to slip. In 2000, Magellan's active share was a bit more than 30%.
The study suggests a relationship between active share and fund performance. Petajisto and Cremers found that funds with the highest active share beat their benchmark index by an average of 1.39 percentage points per year, while those with the lowest active shares produced returns that lagged their benchmark by an average of 1.41 percentage points. They also found that small funds -- those with assets of less than $500 million -- had the highest active shares, and funds with more than $2 billion in assets had the lowest active shares.