The Closet Indexer That Could

The U
In the words of Morningstar's Don Phillips, today's fund investors buy the U. On the left end of the U are the cheapest funds with the least originality: index funds. At the right end are the costliest funds with the most originality: hedge funds for the institutions and idiosyncratic and/or go-anywhere funds for retail buyers. (Reverse the directions if you like; this is not a political matter.) Currently, these two ends get the money, while the middle of the U languishes.

It's easy to understand why this trend exists. The two ends are pure ingredients. The left end is beta without alpha--solely exposure to a given market, untainted by manager intervention, sold at a very low cost. The right end is the opposite: alpha without beta. It delivers solely a manager's contributions, untethered from market movements, and is sold at a high cost. (This isn't strictly correct, as funds do not free themselves entirely from the markets, but conceptually that is the notion.) Mix and match the two ends for the desired blend of alpha/beta, and for the desired cost. Avoid the muddle in the middle.

By this approach, particular contempt is reserved for "closet indexers"--funds that claim to be actively managed, and which therefore charge the fees of active management, but which in practice look a whole lot like a market index. Closet indexers can be detected either by examining their portfolios or, more commonly, their performance streams. The latter consists of an analysis of the fund's monthly returns, compared with those of an index. If the two series of returns are utterly unrelated, then the R-squared statistic will be 0. If they move completely in unison, then the R-squared will be 100.* Typically, actively managed funds that score in the upper 90s are called closet indexers.

The Puzzle
This makes  Nuveen Real Estate Securities (FREAX) among the industry's greatest sinners. Over the past 10 years, the fund has an R-squared of 99 (rounded down, not up!) against both the Dow Jones Real Estate Index and  Vanguard REIT Index  (VGSIX). Given that the A shares version of Nuveen's fund have annual expenses of 1.28%, when compared with a mere 0.24% for Vanguard's fund and none for the Dow Jones Index, it's pretty clear that Nuveen's fund has no chance. Active managers rarely beat indexes at all over a full decade, goes the story (which is false, but that's a different issue). And when they do, it's by being different, not by offering more of the same. Active funds won't outgain an index by copying the index.

The 10-year numbers, through June 30, 2013, are as follows:

  - source: Morningstar Analysts

Hmmm.

There's more work to be done, though. After all, there's no magic about beating an index during a bull market. Just do what the index does, but more. If a fund assumes greater exposure to a market than the index does, either through leverage or by owning the most volatile securities within the index, then the fund will beat the index when the market rises. Voila! The fund will appear to be a closet index fund via its R-squared measure (which is unaffected by its higher volatility), but actually it's something quite different--it's an index fund on steroids. That's not a better fund; it's just a fund that lies at a different point on the risk-return spectrum.

We can test if that happened. In addition to correlation measures, such as R-squared, we can calculate the 10-year beta of Nuveen's fund against Vanguard's fund, and separately, against the Dow Jones index. If Nuveen's beta is significantly higher, then we know that the fund did nothing special. It merely assumed more risk and achieved proportionately more reward.

  - source: Morningstar Analysts

Hmmm again. The direction is wrong. Nuveen's fund beat both indexes during the upswing despite having modestly less market exposure. Additional market-related risk is not the answer to the Nuveen-fund puzzle.

There is, however, the possibility of additional non-market risk. Perhaps Nuveen's fund has a significant amount of idiosyncratic risk that doesn't show up in the beta calculations (which measure only the portion of the fund's risk that is directly related to the overall real estate market). If that were so, the fund's risk did pay off by delivering higher returns--but the lunch did not come for free. That pattern would make sense.

We can capture the non-market risk by examining the 10-year standard deviations (calculated off monthly returns, expressed in annualized fashion) of the three funds. Once again, Nuveen's figure should be higher:

  - source: Morningstar Analysts

Once again, hmmm.

There is, of course, some reason--or reasons--why Nuveen's fund has been superior over the past decade. The first clue I'd seek would be the fund's relative annual returns, to see if the fund's excellence owes to a single very strong period, or to multiple good periods. Then I would attempt to link those periods of strength to portfolio differences between the fund and the indexes.

However, that analysis is past the scope of this column. Today's argument is that the general rules that are used to identify closet indexers--and in the most extreme form, to brand funds as being unworthy of existence--are just that: general rules. These guidelines are useful in aggregate, but not necessarily correct for a given case.

Also, R-squared is a tricky statistic. It's helpful as a relative figure for comparing one fund with another. That Fund A's R-squared is 85 and Fund B's 93, when compared with the same index, is good information. The calculation is harder to use as an absolute measure to define closet index funds, though. Where should the line be drawn? To answer the rhetorical question, it would seem nowhere. If 99 is too low, as it seems to be with Nuveen Real Estate, then there is no reliable number. It does not exist.

John Rekenthaler has been researching the fund industry since 1988. He is now a columnist for Morningstar.com and a member of Morningstar's investment research department. John is quick to point out that while Morningstar typically agrees with the views of the Rekenthaler Report, his views are his own.