Getting What You Pay For

Critics complain that hedge funds are too expensive for the returns they have lately been producing. Setting performance aside (which is as varied as the strategies pursued), on first blush, the ante does seem somewhat dear: Most hedge funds charge a 1 percent management fee, and they also keep 20 percent, and sometimes more, of any profits — this is the so-called “carry.”

Hedge fund pros haughtily retort that they are worth it since many earn consistent returns no matter what the market does. In fact, hedgies argue, it's really the other way around: It's the actively managed mutual fund managers who charge excessively for their services, because, very often, much of their return is driven by the direction of the stock market at large.

Who is right? Ross Miller, a risk consultant and professor in the finance department at the State University of New York at Albany, has concluded that hedge funds are actually cheaper than has commonly been supposed and that mutual funds are more expensive than most have believed. In a paper he published last summer, Miller created something he calls the “active expense ratio”; it's a way to unbundle a manager's return explained by the market from the return that he earned from his own skill. Miller then compares what that market return would cost in a comparable index fund and deducts it; the rest of the management fee is then considered what you are paying for active management. In a somewhat complex math formula, Miller figures out how much you are paying for the manager's skill. “Mutual funds appear to provide investment services for relatively low fees because they bundle passive and active funds management together in a way that understate the true cost of active management,” Miller writes in his study.

The math is complicated (it helps if you can identify Greek symbols), and not everyone agrees that it makes sense to assume the existence of beta in an active manager is bad or if it is worth paying (something) for. But Miller's premise, which is starting to get noticed by other analysts, is: It makes no sense to pay a manager (much) for returns that are explained by the market (beta). Rather, an investor should focus on the excess return (alpha) won by the manager's own stock-picking skills.

And that is best done by, say, getting market returns using low-cost index funds and using alternative strategies to get alpha, Miller says. (More mutual funds are pursuing low-beta, low r-squared strategies, too; see page 61.) On average, hedge funds generate excess return while producing a low beta. An unpublished study by the Frank Russell Investment Group says that the average hedge fund return derives about half of its return from beta and the other half from alpha. To be sure, hedge fund styles differ wildly, and so do their betas. For example, the beta of the CSFB/Tremont Hedge Fund Index — a benchmark drawn from a universe of 1,000 portfolios of all stripes — is 0.26; but its fixed-income arbitrage subsector, comprised of two-dozen narrowly focused funds, carries a beta of just 0.01.

Stock mutual funds, on the other hand, are often primarily beta generators — which, in a bull market, are not bad. As an example, take the large-cap blend mutual fund category; its beta is 95 percent, says Morningstar. Obviously, that's a lot of market exposure.

For certain investors, closely tracking the market is not a bad move. But the question is: How much should you pay for it? Vanguard and other index managers charge basis points. While actively managed domestic large-cap funds, on average, levy a 1.21 percent annual expense charge to cover overhead and manager's fees, says Morningstar. Again, these expenses might at first appear modest compared to the average hedge fund one-and-twenty fee structure.

The Real Cost

Cost-conscious advisors in search of cheap beta and consistent alpha have traditionally addressed the fee issue with a little back-of-the-envelope math. Using the r-squared coefficient data, they dissect portfolios into active and passive components. The passive portion represents the part that conceptually could be replaced by a low-cost index fund, while the active portion represents the uncorrelated alpha contribution. Let's say you have a large-cap fund called Big Bucks (BGBUX) and its r-squared correlation is 95 and it has an expense ratio of 1.21. This means that 95 percent of BGBUX's variance is explained by contemporaneous movements in the S&P 500 benchmark. If an S&P 500 index fund can be held for an expense ratio of 10 basis points (0.10 percent), BGBUX shareholders appear to be overpaying for S&P beta exposure.

You can calculate what shareholders are being charged for active management within the portfolio by backing out the “passive” (market-driven return) management costs. If the index portfolio management is worth 10 basis points, the rest of BGBUX's expenses — 111 basis points — must be tied to active management. Knowing that, you can spread the active cost over the portion of the portfolio's assets that are, in fact, actively managed. The r-squared statistic implies that's just 5 percent, so the effective cost of the portfolio's actively managed segment appears to be a whopping 22.2 percent — a lot higher than the fund's advertised expense ratio.

The envelope approach — naively subtracting the r-squared from 100 — isn't always a true indication of how much a portfolio is actively managed. Miller has a formula to help calculate the active weight (both the paper and the math are available at econwpa.wustl.edu/eps/fin/papers/0506/0506010.pdf). Miller's calculation is necessary because using only the r-squared can grossly overestimate the cost of active management. In fact, Miller says in his study that the active expense ratio may be actually a half or a quarter of that indicated by the quick-and-dirty method explained above.

Active Weight and Cost

To reconcile the true active management component, Miller adjusts r-squared to calculate the portion of the portfolio that is actively managed. (Again, please visit Miller's paper for math details.) From the Miller calculation, 18 percent of BGBUX's portfolio, our hypothetical fund, is actively managed. Put another way, 82 percent of the portfolio is explained by market moves. Is that a closet index fund? Some say no, but many market observers would say yes. One thing is plain: BGBUX shareholders are paying plenty for the four-fifths of the fund's returns that are explained by the S&P 500.

You can derive the weighted active expense ratio by concentrating the portfolio's 111 basis point fee premium on its actively managed component (the math isn't simple; see Miller's article). By Miller's math, investors are paying BGBUX's managers 6.25 percent, or more than five times the advertised expense ratio for their alpha-seeking skills (but much lower than the 22.5 percent management fee its r-squared implied). Suddenly, those hedge fund management fees don't seem so high after all.

It gets worse for the hypothetical fund, BGBUX: Say the alpha is -2.5 percent, then Miller's calculus determines that the fund's weighted active alpha is -13.4 percent. Dividing the active alpha by the active expense ratio then yields the benefit-to-cost ratio for the fund as a rather shabby -2.14. A good-performing fund should crank out a positive ratio of 1.00 or higher.

The Critics

Not everyone agrees that the active expense ratio is the best way to examine a portfolio. Ron Surz, president of PPCA, a company that analyzes manager performance, says of Miller's theory, “It's an interesting perspective and it ought to cause some controversy.” But Surz is not sure whether it's the best way to analyze a fund. Because it's somewhat complicated, Surz compares it to buying a steak dinner for $32 in a restaurant. The eight-ounce steak comes with carrots and potatoes, but those vegetables can be had for, say, $2 at a grocery. If you back out the vegetable costs, you're paying $30 for the steak, or $60 per pound. “That's the rationale Miller is painting,” Surz says. “But is it a reasonable perspective? I don't know. People don't go to restaurants and think like that. They look at it as a steak meal.” They are willing to pay the additional fee for the service, the restaurant's ambiance and for not having to cook or do the dishes, Surz says.

Like Surz, Leola Ross, a senior research analyst at the Russell Investment Group, has problems with Miller's assumptions. “One must not assume that the only goal of long-only active management is to produce alpha,” Ross says. “This assumption may be challenged on several fronts. First, beta is a great source of return and long-only active managers are often strictly mandated to deliver a beta of one with some alpha kicker. This is an expensive process and needs to be compensated. Second, long-only active managers are often charged with creating this beta with an alpha kicker while maintaining a sensitivity to tax considerations, or any host of investment restrictions.”

However you come down on the arguments, it is interesting to isolate a manager's stock-picking skills and then price it using Miller's method. And surprisingly, excluding the effect of incentive fees, that cost — at least for the largest domestic equity mutual funds — seems to be on par with that charged by hedge funds.

Value Fund Fees Viewed Via Miller

Compared to the Russell 1000 Value Index (and an exchange-traded fund that tracks the benchmark for 20 basis points), value managers fare poorly. Note that the weighted average active expense ratio at 1.89 percent is better than three times the rather slim costs advertised for these portfolios.

How Growth Funds Look

Growth managers produced much better results when benchmarked against the Russell 1000 Growth Index. Growth exposure can be obtained from an ETF for 20 basis points. The weighted average active expense ratio of this actively managed group is 2.25 percent.