From the October 2006 issue of Wealth Manager Web • Subscribe!

October 1, 2006

Pricing Matters

The life of a middling manager isn't getting any easier, especially if he charges a premium. An ever-growing assortment of software and analytical metrics enhance transparency in the money game, helping investors shine a light on the mediocre funds that aren't shy about levying high fees. The latest addition to the clarity-boosting arsenal is the active expense ratio (AER), a quantitative gauge devised by Ross Miller, a professor of finance at the State University of New York at Albany and president of Miller Risk Advisors.

AER's goal is simple enough: Measuring the portion of expenses directly tied to the active investment management. Think of AER as a refined version of the gross expense ratio that's reported in every fund prospectus and otherwise dispensed by data vendors such as Morningstar. Refining is vital because the gross expense ratio is often misleading when it comes to measuring the true price of active management, Miller says.

The reason is that most actively managed mutual funds harbor some degree of influence from the market. Beta, in other words, is a familiar, and sometimes dominant companion in most actively managed mutual funds. But gross expense ratio puts a price on the entire portfolio, and so by that measure it's unclear how much a fund charges for alpha delivered, if any.

No longer. In a paper Miller penned last year, he has come up with what he says is a practical solution for calculating the cost of alpha. Titled "Active Expense Ratios and Active Alphas" and available at MillerRisk.com, it outlines "a method for allocating fund expenses between active and passive management and constructs a simple formula for finding the cost of active management."

In a recent interview with Wealth Manager, Miller summarized the central issue for fund pricing this way: "How much does the alpha cost? It's not obvious from looking only at the gross expense ratio."

Crunching the data for Miller's AER is enlightening, but not necessarily encouraging. For example, at the end of 2004, the paper notes, the mean AER for large-cap equity mutual funds tracked by Morningstar was 7 percent--roughly six times higher than their published expense ratio of 1.15 percent.

The good news is that by outing the real costs of active management, the pressure is growing on funds to shave expenses. Exactly how far any newfound thrift extends remains to be seen, although Miller's AER has received a fair amount of press over the past year. The publicity, he suggests, has convinced some fund companies to think twice about what they're charging.

Because Miller's paper is forthcoming in the Journal of Investment Management, a new round of publicity for AER may be in the offing. To learn more about his formula, we recently caught up with Miller.

What does your active expense ratio measure?

It allocates the costs of money management between the passive and active components of a portfolio. You can think of an actively managed mutual fund as having two parts: One portion is comparable to an index fund; the other is an active part. When you buy a mutual fund, you're buying a package deal--a package of beta and alpha. The active expense ratio measures the costs of the assets that are being actively managed.

How does it work?

Let's say you had a money manager, and all he did was put your money in a domestic stock market index fund, and for that privilege he charged 100 basis points. You're paying 100 basis points for something you can get for 10 basis points on a retail index fund like Fidelity Spartan. So, our manager is charging something like a 90-basis-point overcharge.

That's an extreme case. A subtler example is an active fund that closely tracks an index--not perfectly, but closely. Historically, funds that closely track an index have low management fees relative to other types of funds. Nonetheless, the question is, "How do you figure out how much of your money is going for active management and how much to passive management?" That's a critical issue because money managers often try to reduce tracking error relative to an index. There's nothing wrong with that, but investors need an accounting of the costs because managers aren't always doing a lot of active management. A classic case is the PIMCO Stocks Plus Fund, which is basically an S&P 500 index fund with an overlay of Bill Gross's bond picks. You're not buying a lot of active management in the fund overall and you're paying a fair bit for Bill Gross.

None of this, by the way, is an issue of judging the intentions of the portfolio managers running funds. Rather, it's an issue of accounting for what's actually happened. The managers could have the best of intentions, but if you find that over a period of time that you could do what the manager's doing a lot less expensively, money's going to flow out of his fund.

How is your active expense ratio calculated?

The formula has three key variables: The R-squared of the fund you're looking at against the relevant benchmark; the active fund's expense ratio; and the expense ratio of the investable index fund that's an appropriate alternative for capturing the beta portion of the active fund.

Let's say you have a large-cap domestic equity fund that charges 100 basis points for expenses and is said to be actively managed, but in fact is a shadow index fund with an R-squared of 99 percent against the S&P 500. If you ran the math for dividing beta and alpha out, based on my paper, it turns out that roughly 91 percent of the assets are passively managed and the remaining 9 percent, actively managed. As a result, the active expense ratio is a steep 9 percent.

How do you come up with 9 percent?

In my paper, there's a formula for converting the R-squared, which tells you how much of the variance in the active fund is explained by the benchmark. I convert the R-squared into active and passive shares of the portfolio. You can think of it as separating alpha and beta. That's the tricky intermediate step that converts the 99 percent R-squared into 91 percent beta and 9 percent alpha in the example I just gave you.

The next step is determining how to replicate what the fund's doing. Based on the numbers generated by my paper's formula, 91 percent of the money would go into an appropriate index fund, which you can get for 18 basis points in the Vanguard 500 Fund, for instance. If you put all of your money in this index fund, the total charge would be 18 basis points. But in this example of trying to replicate the active fund, you're putting 91 percent of the money into the index fund. That works out to a charge of around 16 basis points. In other words, 91 percent of 18 basis points comes to a charge of 16 basis points.

What this means is that for the active fund that charges 100 basis points, you can reproduce the passive side for 16 basis points. That leaves 84 basis points, which is the price of the 9 percent of the fund that's actively managed. That translates into an active expense ratio of more than 9 percent. So, in this example you're paying quite a bit for the active management--much more than implied by the gross expense ratio of one percent.

You might also think of the calculation in absolute-dollar terms, which makes the formula easier to understand. Using the same fund example, a $10,000 investment in the active fund incurs total fees of $100--that is, a 1 percent expense ratio. Of that $100, you're paying $16 for the roughly $9,100 that's being passively managed. And you're paying $84 for the $900 that's being actively managed. It's that $84 divided by the $900 that's being actively managed that gives you the roughly 9 percent active expense ratio.

Is your active expense ratio the first of its kind?

Yes, and that's why it's been so popular. I've been at conferences where I've had notable economists come up to me and say, "I've thought about something like that." But there are a lot of wrong ways to do it. In the practitioner literature, for instance, people have thought about using the R-squared number, but that approach gives results that don't make any sense, which is why I modified R-squared in my paper.

What's the range of active expense ratios for mutual funds?

From 1 percent to over 20 percent. The best-case scenario is found with the active funds from Vanguard, like the Wellington and Windsor funds which tend to have active expense ratios in the 1 to 2 percent range. In the next range up are the portfolios from American Funds, for instance, which tend to have a 3 to 4 percent active expense ratio. Above that are the low-fee shadow indexers, which are in the 5 to 8 percent range. For funds with the highest active expense ratios, the typical characteristics are portfolios with an R-squared of 98 or 99 percent and a gross expense ratio in excess of 2 percent.

What's your take on fees as related to active money management these days?

I think people are getting wise to the fee issue, and so it's putting a lot of downward pressure on fees. There are still plenty of active managers, and there probably always will be. But many investors now see the portfolio process as one of capturing beta and sector exposure through passive investments, and getting alpha through targeted active investments.

Does your active expense ratio work with hedge funds?

It's more problematic if you're trying to compare hedge funds [because there's not always an obvious benchmark or because leverage employed can be relatively high.] On the other hand, for a pure hedge fund--one that's uncorrelated with the markets--its overall fee is the active expense ratio. That's because with a hedge fund that's pure active management, what you pay is only for active management.

What are the caveats to using your measure?

Within the current mutual fund world, the active expense ratio does just fine. But like any other measurement--alpha, Sharpe ratio, whatever--the simple form of my active expense ratio is something that a clever manager can game because a clever manager can game any performance measure.

Looking backward, however, the active expense ratio is a great tool because no one knew it existed. But the world is changing, and funds are being punished for being shadow indexers. If you look at the funds that are working on becoming more active, they're not necessarily more active in absolute terms; they're just more of a hybrid. As a result, Morningstar is going to have to start abandoning its single benchmark approach. Instead of asking what it costs for a single benchmark, the question's becoming: What does it cost to reproduce the package of benchmarks? For a fund that engages in active asset allocation, my method fails; in fact, pretty much all of portfolio analysis method fails for that hybrid approach.

Is that because there's no obvious benchmark?

Right. I was sensitive to that from day one, because back when I worked in asset management, I was head of research for a family of products that involved sector rotation and country rotation. There was no fixed benchmark. If you performed style analysis on the portfolio, the results changed every month. But the number of funds like that currently is still tiny.

Is it fair to say that the value of your active expense ratio is dependent on having a reliable investable benchmark?

Right. You need to have something that's a passive alternative that consists of one or more investable indices.

If mutual funds managers are targeting more than one benchmark, the challenge of coming up with a good investable index becomes tougher.

Yes, but the same is true for any performance analysis.

James Picerno (jpicerno@highlinemedia.com) is senior writer at Wealth Manager.

Reprints Discuss this story
This is where the comments go.