The promise of active investing is compelling: the opportunity to earn higher risk-adjusted returns! And paying a fee to an active manager--who is doing something unique--can make sense. And as we know, the only way to beat a benchmark is by being different from the benchmark, as we discuss here. It follows that if a fund has a large fraction of holdings that are different from its benchmark, it becomes more likely that the fund will be able to outperform (or underperform) the benchmark, gross of fees. Intuitively, in order to outperform, one must be different (Pat O'Shaughnessy also has a good piece here). Howard Marks says it best in his piece "Dare to be Great:"
This just in: you can't take the same actions as everyone else and expect to outperform.But what if 50% of the portfolio of a so-called "active" fund actually holds positions that are identical to those found in an index fund tracking the S&P 500? What about if these passive positions made up 75% of the "active" fund's portfolio? Well, that would not really be active or different at all, right? This would be much closer to pure passive, index-based investing, which can be done at low cost. In cases like these, we probably should not be willing to pay more expensive, active fees, since we are not receiving "active" stock selection services. Sadly, however, many do pay high fees for mostly passive exposures. Welcome to the world of "closet indexing," where funds hold many of the same securities as their passive benchmarks, but investors still pay full freight for what they think is "active" management. In a recent paper, "Do Mutual Fund Investors Get What They Pay For? The Legal Consequences of Closet Index Funds," by Cremers and Curtis, the authors investigate the evidence for closet indexing and how closet indexing affects investors. Before getting into the details of the paper, it's useful to discuss the concept of Active Share, which can help us understand the closet indexing phenomenon.
What is Active Share?Active Share measures how different a fund's holdings are from the fund's benchmark. How does this work in practice? We provide an example below, from the paper, to demonstrate the measurement mechanics. Consider a hypothetical active fund (column 1) that invests in 4 stocks, while the fund's benchmark invests in 5 stocks, in different weights, as per the chart below: Note that there are two dimensions of difference between the fund holdings and the benchmark holdings:
- The overlap of stocks in the fund and the benchmark: Of the fund's 4 stock portfolio, 3 are also in the benchmark (AAA, DDD, and FFF), and 1 is not in the benchmark (EEE), while the benchmark contains 2 stocks that the fund does not hold (BBB and CCC).
- The weights of stocks in both the fund and the benchmark: We can see that the weights of stocks held in the fund that are also in the benchmark sometimes differ; for instance the fund holds 25% less of FFF than the benchmark does, but twice as much DDD.
Why is Active Share Important for Investors?Active Share is undoubtedly a useful concept for investors to understand (visual active share will be even better). In our example above, the fund (which we'll call Fund A) is effectively investing 65% of its investors' money in the same positions as its benchmark index. The Active component is only 35%. Let's say another fund, Fund B, is more active, with an Active Share of 70%. What are the prospects for outperformance by each of these two funds? In short, they are quite different. Fund A would need to generate double the return on the active component of its portfolio to match the returns of Fund B on its managed assets. Obviously, this is very hard for Fund A to achieve. And in the real world, it turns out that closet indexers often simply fail to outperform. Let's try and make this more concrete: Consider a smart-beta value ETF that charges 0.50% and holds 250 value stocks and "smarter-beta" ETF that charges 1% and holds 50 value stocks. For simplicity, let’s assume the smart beta and the active ETFs follow the same strategy and that value strategies have similar systematic risk to the broad market. Which value ETF should we buy? The naïve answer is “buy the cheapest,” but that answer is incorrect. The smart-beta ETF and the smarter-beta active ETF have very different return profiles and fees. As the authors highlight, investors need to look at both fees and expected returns to make the right call. For example, if the 50-stock value portfolio generates a 1.5% average expected return above the market (0.50% after fees), and a 250-stock value portfolio generates a 0.5% average expected return above the market (0.00% after fees), the smarter-beta ETF is a much better value despite being twice the cost. (See Wes's WSJ article).
The Poster Child for Closet Indexing: The Magellan FundThe Fidelity Magellan Fund became famous based on Peter Lynch's strong management in the 80s. What is less well known is that even as the fund became the largest in the U.S., with $103 billion under management in 2001, its Active Share plunged during this time. (sidenote: we've also heard from a credible source that the fund was launched as a spaghetti against the wall fund). The Chart below plots on the left axis the Active Share under a series of portfolio managers who ran the fund: Note that during much of the time that Robert Stansky managed the fund, Magellan's Active Share fell to the 30%-40% range. How did Magellan do during this time? Perhaps unsurprisingly, performance over the period was very close to the benchmark -- minus 1% per year for management fees, of course -- as is further discussed in this article. A quote from the paper says it all:
During the Stansky years...regardless of the direction the market went, Magellan investors were likely to do poorly relative to the benchmark, because they were - in effect - buying the benchmark at high cost. For example, while the expense ratio was only 0.75%, the Active Fee or how much investors of the Fidelity Magellan fund were actually paying for active management at the end of 2001 amounted to 1.95% per year.
What does the evidence say regarding closet indexers?How prevalent are closet indexing funds? The authors use data from 2014 from the Thompson-Reuters mutual fund holdings database, and examine funds that invest at least 80% of their assets in U.S. equities, and calculate Active share versus the self-declared benchmark. The sample includes 1,225 U.S. mutual funds managing $2.7 trillion. The authors use 60% Active Share as a cutoff for closet indexing; that is, if a fund has Active Share < 60% it is considered a closet indexer. Below are the summary statistics for Active Share for 2014: Two things that are striking about these results are the following:
- Closet indexing is more common than you might expect overall
- Closet indexing is primarily a Large Cap Fund phenomenon.
How Does Closet Indexing Impact Fund Performance?By now, it should be clear that closet indexing is a real phenomenon, and has at least the potential to reduce returns to mutual fund investors. But what actually happens to returns in the real world, in the aggregate? The authors focus on the large cap funds identified above, and track the performance of 25 different mutual fund portfolios by sorting the funds along two dimensions:
- Expense Ratio: construct five quintile portfolios (1 through 5 on the vertical axis) based on the expense ratio of the funds. The average annual expense ratios are 1) 0.58%, 2) 0.96%, 3) 1.16%, 4) 1.38%, and 5) 1.80%.
- Active Share: construct five quintile portfolios (1 through 5 on the horizontal axis) based on Active Share
Active Fee: Adjusting Active Expense Ratios to Account for Passive ExposuresSo how much are you really paying for the active component of a strategy? It is a straightforward matter to back out the fee per unit of active management. The authors present "Active Fee," which uses the following formula (where "Index Fund Fee" is the expense ratio associated the benchmark): The formula involves three muscle movements: 1) active share, 2) the expense ratio, and 3) the index expense ratio. We explain each below:
- Expense Ratio: The expense ratio associated with the active fund.
- Index Expense Ratio: The expense ratio associated with the costs of investing in the active fund’s index benchmark.
- Active Share: The “active share” element of active fee is a bit more complicated, mildly controversial, but fairly intuitive. Active share simply measures how different a fund’s holdings are from its benchmark and helps an investor determine what percentage of a fund is unique, or active, and what percentage of a fund is not differentiated, or passive.
- Diluted Fund ActiveFee = 1.08%
- Concentrated Fund ActiveFee = 1.08%
Do Mutual Fund Investors Get What They Pay For? The Legal Consequences of Closet Index Funds
- Cremers and Curtis
- A version of the paper can be found here.
- Want a summary of academic papers with alpha? Check out our Academic Research Recap Category.