Illusionist

One of the banes of modern investment manage- ment is the existence of closet indexers–money managers who are ostensibly active managers but who in fact mimic an index. The last thing that either you or your clients want to do is pay active management fees for an index product. Are you able to avoid closet indexers and pick winning managers, or have you thrown in the towel and simply index client portfolios?

As the debate between active management and passive management continues to rage, total assets managed by closet indexers within the universe of open-end U.S. equity funds have grown to almost 30%, according to a recent study by two Yale University professors, Martijn Cremers and Antti Petajisto, called How Active Is Your Fund Manager? A New Measure That Predicts Performance. Avoiding closet indexers has become more challenging than ever.

In this article, we will examine the cost of investing in closet indexers; hypothesize why this phenomenon has arisen; explain why it is difficult to avoid closet indexers when using the style grid to allocate portfolios; and show how the strategy we’ve developed at AthenaInvest and ICON Advisers–Strategy Based Investing–can be used to find winning active managers.

The Cost of Closet Indexers

Cremers and Petajisto (C&P) provide information bearing directly on this question in a January 2007 working paper based on a comprehensive study of open-end U.S. equity mutual funds for the period 1980 through 2003. They measure how active a manager is by comparing a fund’s holdings to the best-fit index holdings and then calculating the fund’s “active share,” which ranges from 0% (index fund) to 100% (fully active). Active share is defined by C&P as a measure of how aggressively a manager attempts to beat the market, not how actively the portfolio is being traded.

The chart below shows the performance of each active share quintile from the C&P study. The top two quintiles, comprising the most active managers, have average alphas (net of fees and the market benchmark return) of 139 basis points and 39 basis points, respectively, while the lower three quintile managers all produced negative average alphas. These lower three quintiles comprise closet indexers, which, according to the C&P study, underperform the top active managers by nearly 300 basis points. Thus any methodology that drives advisors towards closet indexers, which fall in the middle between active managers and true indexers in the chart at right, by our definition underperformed during the period from 1980 through 2003.

The C&P study is just one of several recent studies that indicate market-cap/value-growth box investing which we will argue leads directly to closet indexing and produces poor investment performance. The results of our own studies have been reported in the Journal of Investment Consulting (Winter 2005-06) and in a series of articles in Investment Advisor (September 2005, January 2006, and February 2006). In those articles, we make the case that constraining investment managers to boxes, defined by market-cap and value-growth characteristics, costs about 300 basis points per year in underperformance. Further, we argue that style boxes are not asset classes and that diversifying funds among those boxes provides minimal, if any, risk reduction beyond what can be obtained by randomly selecting funds.

Our studies, along with others, are summarized in the table, “Performance Advantage of Drifters vs. Box Huggers” below. Each of these studies focuses on the question of whether it is better for a manager to stay in a box (i.e., be a hugger) or drift around the U.S. equity universe (i.e., be a drifter). These studies use different methodologies, and collectively cover a 24-year time period, and come to the same conclusion: historically, it is better to be a drifter than a hugger. The annual risk-adjusted drifter-minus-hugger return differential ranges from a little under 200 basis points to more than 400 basis points.

While past performance may not be indicative of future results, by reading these studies, advisors can see for themselves that historically, drifting has helped performance. The first study cited in the table can be replicated by any investment professional with access to Zephyr by simply sorting fund returns by the Zephyr Style Drift measure and noting that those that drift the most generate the highest returns.

Why the Growth in Closet Indexing?

Over the last 20 years, the style grid has become more prominent as a diversification tool in portfolio construction. In this approach, an advisor seeks out small-cap value and large-cap growth managers, for example, in order to fill out the “box” allocations for a client portfolio. Managers are expected to stay within a box or be “style pure,” an aspect of portfolio management that has been seen as very important. Consequently, boxes played a central role in categorizing, evaluating, and constructing equity portfolios.

How did the style grid get started? We recently met with several investment industry veterans who said they believe it started in 1984. At the time, they were consulting with a finance professor about how their company might build its indexes. The professor went to the board in the conference room and began laying out a simple four-box system which was based on the evolving academic research dealing with market-cap and P/E ratios. Thus was born the style grid, and the company began to build indexes based on this concept. Others in the industry began using these indexes to assess managers, and shortly thereafter, the idea of “style purity” took root, much to the chagrin of these industry pioneers. The rest, as they say, is history as the industry has bought into the style grid concept lock, stock, and barrel.

How does “style purity” breed closet indexers? Using boxes to allocate portfolios and correspondingly setting the expectation that managers be “style pure” is the same as asking active managers to provide products that look like the box indexes. Commonly used measurements such as low style drift, high R-squared, and low tracking error all measure the degree to which an active manager matches a box index. As these measurements have become the standard over time, there has been a rise in closet indexing as active managers are pigeon-holed into the boxes by advisors, consultants, and the distribution system in general. To be fair, this growth is also the consequence of fund companies responding to the new box-driven demand. We believe this stampede towards the style grid has resulted in the commensurate rise of closet indexers.

Results from the C&P study (reported in “The Rise of Indexing” above), confirm the emergence and growth of closet indexers. C&P found that in the early 1980s, essentially all open-end U.S. equity funds were active (defined as having an active share of 60% or greater) and there were no closet indexers nor true indexers to speak of. When the industry began to box managers in the mid-’80s, both closet and true indexers began to grow as a percent of industry AUM. By 2003, the latest data in the C&P study, true indexers had grown to 13% of AUM while closet indexers had grown to nearly 30% or roughly $1.5 trillion.