TCS Daily


You Might Be Paying Too Much for Your Mutual Fund; We Figured Out Why

By Peter Wallison and Robert Litan - April 17, 2007 12:00 AM

Editor's note: One curious feature of today's financial markets involves the cost of mutual funds. While all the normal indicators of competition are present in the mutual fund industry, prices have not converged downward toward a common marginal cost. There is wide disparity of expense ratios across similar product offerings, including for products such as index funds.

Peter Wallison of the American Enterprise Institute and Robert Litan of the Kauffman Foundation and the Brookings Institution have spent the last year and half studying the mutual fund industry in detail. They believe they have solved the puzzle. The following essay is longer than a typical TCS essay and it is technical in places, but given the size and importance of the $10 trillion mutual fund industry, their findings merits study and further review.

A headline in the Wall Street Journal last year declared, "Independent Directors Strike Back." The article under the headline noted that the independent directors of the mutual funds managed by AIM Investments were demanding that the funds lower their advisory fees. The article attributed this singular behavior to a newly installed independent chair of the mutual funds. It evidently did not seem odd to the author of the article that the independent directors of a group of mutual funds would have to demand that the adviser compete with other advisers, yet by the lights of ordinary business practice it was quite strange indeed. For some reason, the adviser to the AIM funds did not want to lower its fees and expenses in order to attract more investors. Why would it be that in the mutual fund industry, among all the industries in the United States, the principal players—the investment advisers of mutual funds—refuse to compete on price with one another?

One of the academic papers most critical of mutual fund pricing begins its analysis with this statement: "This strangeness—tremendous popularity, proliferating consumer options, and less than robust price competition—arises in the realm of the most tightly regulated financial product sold in the country today. In the words of a former SEC chairman, 'no issuer of securities is subject to more detailed regulation than a mutual fund.'" But in the end, to address this "strangeness," the authors call for yet more regulation; it seems never to have occurred to them that the heavy regulation of the mutual fund industry and its peculiar price structure might be related: that regulation might be the cause of, rather than the remedy for, the industry's peculiar failure to live up to the promise of its competitive structure.

In our view the problem is not greedy advisers, co-opted boards, or inattentive or uninformed investors. The problem is regulation itself—the fact that the Investment Company Act of 1940 requires mutual funds to be structured as separate corporations, with boards of directors charged with responsibility for controlling the adviser's fees and the fund's expenses. This unique form—an artifact of the industry's origins in the 1920s—makes the mutual fund industry, in effect, a rate-regulated industry. And as has historically been the case with rate regulation, the result is to impair effective competition.

It is no coincidence that the unique organizational structure required of the mutual fund industry has produced a unique system of price setting, one that does not include vigorous price competition and the benefits that it can bring to consumers. Price convergence—usually toward the marginal cost of the most efficient producer—is present in most areas of the U.S. economy, but has not occurred in the mutual fund industry. There, fees and expenses—the price for the service of collectively managing a fund—vary widely and seemingly without relation to service quality. The 100 percent discrepancy in expense ratios between, for example, the Gartmore and the UBS index funds, when both offer similar services as measured by fund performance, is a case in point.

Simply put, price competition and price convergence have not occurred in the mutual fund industry because there is little incentive for investment advisers to reduce costs. Under section 36(b) of the ICA, the investment advisers of mutual funds are deemed to have a "fiduciary duty with respect to the receipt of compensation for services, or of payments of a material nature, paid by such registered investment company... to such investment adviser or any affiliated person of such investment adviser." The SEC, or any investor, is specifically permitted to enforce this requirement against the investment adviser and any other person who has a fiduciary duty with respect to the fund, which of course includes the board of directors. Thus every form of payment by the fund to the investment adviser—not just the advisory fee—must not result in compensation to the adviser or its affiliates that exceeds what the vague term "fiduciary duty" would allow, and the members of a board of directors that approves compensation that exceeds this permissible level are liable to suit from fund shareholders to the same extent as the investment adviser who receives the payment.

In a series of cases testing this language, particularly Gartenberg v. Merrill Lynch Asset Management, Inc., the courts have given some content to the term "fiduciary duty." The Gartenberg court laid out a series of considerations that a fund's directors must take into account when judging whether the fees and expenses charged to a fund by its adviser meet the fiduciary duty test:

• the nature and quality of the adviser's services;

• the adviser's costs of providing these services;

• whether the adviser shares with the fund the economies of scale that result from growth of the fund;

• the volume of orders that the adviser must process; and

• the indirect benefits that the adviser receives as a result of operating the fund.

Note that the only objectively quantifiable elements in this list are the adviser's costs and the volume of orders. Because order processing does not involve a significant cost, and the other elements of the test are either hard to evaluate or discover, or of little significance when discovered, the directors will naturally focus on the adviser's costs. In practice, this means that fund boards of directors—when they are operating most diligently—pay considerable attention to the adviser's costs of providing a wide variety of services to the fund, with a view to determining whether these costs are within some undefined zone of reasonableness. These costs, supplemented by an allowance for profit, then become the basis for the adviser's compensation. Oddly, the Gartenberg court rejected comparisons with the fees and expenses of advisers to other funds as the standard to which the directors should look, thus specifically precluding any director responsibility for promoting competition among funds or advisory groups.

If this focus on costs in price or rate setting sounds familiar, it is because one encounters the same focus in the rate-setting process for public utilities. One major difference is that public utilities usually have large installed fixed capital investments, whereas investment advisers tend to have large variable personnel costs. Otherwise the rate-setting process for both is similar (although, as we will shortly explain, with different types of regulators: government agencies in the case of utilities, boards of directors in the case of mutual funds) and results in similar outcomes.

Government (or for that matter any) regulation of rates or prices is subject to several well-known deficiencies, many of which are present in the mutual fund industry. As one of the country's leading regulatory economists, Alfred Kahn, wrote in his treatise, The Economics of Regulation, "[R]egulation as such contains no built-in mechanism for assuring efficiency. To the extent that it effectively restrains public utility companies from fully exploiting their potential monopoly power, it tends to take away any supernormal returns they might earn as a result of improvements in efficiency, thereby diminishing their incentive to try."

To be sure, mutual funds are not regulated utilities and do not have monopoly power, so the quotation seems inapposite. However, it is not the funds themselves that are rate-regulated under the ICA, but the investment adviser's charges to the fund. Because there is only one adviser to the fund, the adviser is in fact in the same position with respect to the fund as a monopoly public utility is with respect to the consumers in its market. The board of directors, then, functions in the same way that the public utility rate commission functions: it regulates or limits the profits of the adviser to be sure that consumers—the fund and, ultimately, its shareholders—are fairly treated.

Looked at from this perspective, the analogy to public utility rate regulation actually is quite close. Let us rephrase Kahn's statement so that it applies more clearly to the mutual fund case: in controlling the degree of profit that the investment adviser earns from its services to the fund, the board weakens the incentives of the investment adviser to reduce its costs so as to earn higher profits. This appears to be what has been happening all along in the mutual fund business. The regulation of investment advisers' compensation under section 36(b) has reduced their incentive to cut costs and operate more efficiently.

There is, however, one other important respect in which the investment adviser is different from the monopoly public utility: because it has a monopoly on production in its geographic area, the public utility presumably does not have to compete with other producers in supplying, say, electricity. In contrast, competition clearly places some limit on what an adviser might charge as an advisory fee, since at some point a fee that is too high would drive investors to other funds, lowering the adviser's profit. So the question arises: would not advisers still seek to keep their fees low in order to retain investors, or—more to the point—would they not lower their fees in order to attract investors away from competing funds? If so, don't advisers still operate in a competitive market, with incentives for lower costs, innovation, and efficiency, even though they must obtain approval of their fees from a board of directors?

If we consider how boards actually determine what is reasonable for an investment adviser to earn, we find that the answer to this question is no. First, section 36(b) does not require that the board attempt to ensure that the adviser's fees in any sense replicate what a competitive market would produce. On the contrary, section 36(b) seems to assume that the adviser will charge the fund more than a competitive price, and the Gartenberg court specifically precluded a fund's directors from considering the fees and expenses of competing advisers. Underlying section 36(b), as well as the Gartenberg ruling, is the notion, noted earlier, that advisers have interests that conflict with those of the fund's shareholders, and that the adviser is not operating in a competitive market. As the court said in Gartenberg, "Competition between money market funds for shareholder business does not support an inference that competition must therefore also exist between adviser-managers for fund business. The former may be vigorous even though the latter is virtually non-existent. Each is governed by different forces. Reliance on prevailing industry advisory fees will not satisfy."

This peculiar statement was essential to the court's analysis, and of a piece with the thinking that led to the adoption of section 36(b) in 1970. If the market for advisory services were in fact competitive, there would be no need for a board of directors to control the charges of the adviser, since the adviser would have to price its fees on a competitive basis. Section 36(b) seemed necessary to Congress because Congress was led to believe that prices were not set competitively in the mutual fund industry.

In any event, to make its judgments under section 36(b), the board receives data from the adviser that document the adviser's costs in providing various services to the fund. In addition to the advisory fee itself, these costs often include 12b-1 (distribution) fees, subadvisory services, and fees for other activities such as account maintenance, accounting and auditing, legal and administrative services, transfer agency and custodial services, and keeping books and records. After satisfying itself that the costs borne by the adviser for providing these services are within a reasonable range, and comparable to the costs of services that the adviser might purchase elsewhere, the board allows the adviser to charge the fund—as a reasonable profit—somewhat more than the adviser's costs.

This process creates an economic environment for the adviser's management that is entirely different from a fully competitive environment. In the latter, other advisers would be offering lower costs to attract investors, and—assuming that investors respond—this would require that higher-cost advisers lower their rates to meet the competition. They would be forced to find less expensive inputs in order to maintain their profitability, or perhaps even to survive. This is how competition increases efficiency and promotes innovation. The presence of a board of directors with power to determine what would be a reasonable profit, and to add that profit to the adviser's costs in setting fees, changes the calculus for advisers.

It may well be that an adviser could attract more investors by lowering its fees and costs, but if the adviser does that, and thereby succeeds in increasing its profit, the board may take that increased profit away in the next fee negotiation. Indeed, the common requirement that advisers establish breakpoints in their fee schedules—in order to recapture for the fund some of the benefits of an adviser's economies of scale—is a perfect illustration of Kahn's point that rate regulation reduces the regulated industry's incentive to operate more efficiently and thereby lower its costs. Having taken the revenue risk associated with reducing its fees and expenses, the adviser discovers that it has to give back some or all of its resulting profit in the form of a breakpoint or two. This is a clear disincentive to reducing fees and expenses, making it rational for an adviser instead to maintain its expense ratio at a level that will not cause a reduction in profits at the next fee negotiation, but that is not so high as to chase away investors. When all investment advisers follow this course, the result is the mutual fund market we see today: all the normal indicators of competition are present, yet prices have not converged downward toward a common marginal cost.

Note that in this system neither the adviser nor the board is in any way at fault. Rather, highly motivated advisers and highly diligent boards are both caught in the process established by section 36(b) of the ICA, as explicated by Gartenberg, in which boards are required to regulate the adviser's profits just as utility commissions are required to regulate utility rates—and cannot in the process consider the prices of competitors. Once this regulatory process was established, it was foreordained that fund pricing would never converge to a single competitive price for similar services. That, in turn, means that the costs of investing in the average mutual fund will inevitably be higher than would prevail in a competitive market.

The wide disparity in mutual fund expense ratios makes clear that boards of directors are not effective in controlling costs. This is entirely understandable and certainly not the result of any lack of independence. The only cost information that an adviser furnishes to a board of directors is information on average or fixed costs, not the marginal cost of providing the advisory service. In a competitive market, most economists believe, the marginal cost of the most efficient producer—that is, the cost to that producer of furnishing the last profitable unit of the product or service—sets, and should set, the market price. All other competitors, if they wish to stay in business, must either differentiate themselves from the most efficient producer or match that price.

But the fund board cannot be expected to determine the marginal cost of the adviser and thus the price that the adviser would charge in a fully competitive market. Alfred Kahn again explains why:

"Manifestly, the operating expenses and capital outlays of public utility companies are by far the most important component of their rate levels.... Therefore, in terms of their quantitative importance, it would be reasonable to expect regulatory commissions to give these costs the major part of their attention. But in fact they have not done so; they have given their principal attention instead to the limitation of profits.

"The reasons for this perverse distribution of effort illustrate once again the inherent limitations of regulation as an institution of effective social control of industry. Effective regulation of operating expenses and capital outlays would require a detailed, day-by-day, transaction-by-transaction, and decision-by-decision review of every aspect of the company's operation. Commissions could do so only if they were prepared completely to duplicate the role of management itself."

Kahn describes exactly what happens when a board of directors attempts to negotiate what should be the adviser's compensation for management services, and thus the adviser's price to investors in the fund. The board, as required by section 36(b), attempts to limit the adviser's profits, not to find the price for the adviser's services that would replicate a competitive price.

The investment adviser does not furnish marginal cost information to the board for two reasons. First, as noted above, the adviser has an incentive to furnish only average costs, because its profit is determined by reference to these costs, which typically are higher than marginal costs. Second, and perhaps more important, the adviser does not know what its marginal costs actually are, since these can only be determined in a competitive market. In a famous essay, "Competition as a Discovery Procedure," Friedrich Hayek pointed out that only through actual competition can one determine what a product or service might actually cost in a competitive environment. This is because the most important facts—such as the price at which the last consumer will buy and the marginal cost of the last input—are not known and cannot be known in advance of actual competition. The costs of inputs keep changing, as does the price the last consumer is willing to pay for a given amount of a product or service. None of these things can be forecast or discovered in advance, and thus they cannot be communicated to boards of directors, even if investment advisers had an incentive to do so.

One other point is also important to note: at least some of the costs reported by advisers to boards must be arbitrarily determined. This is not because of any wish on the part of advisers to mislead fund boards, but simply because there is no right or rational way for any organization engaged in a joint enterprise—such as managing two or more mutual funds—to allocate certain costs between them. To take a simple example, how would an investment adviser that initially advises only one fund and later adds another allocate its investment research costs between the two? The new fund may at first take more time of the adviser's research staff than the old one, so does that mean that these costs should be allocated principally to the new fund? Perhaps, but the effect of this will be to substantially reduce the adviser's costs on the old fund, and perhaps require the board of the old fund to reduce the adviser's profit on that fund to a "reasonable" level under section 36(b). Once both funds are operating and start-up costs are no longer a factor, how should the adviser allocate its costs? If we assume that the new fund is half the size of the old, but the adviser's staff spends the same amount of time on each, should the adviser's personnel costs for the old fund be cut in half, or should the adviser allocate personnel costs so that each fund bears them in proportion to its size? Alternatively, the adviser could argue plausibly that the new fund was started principally to take advantage of the economies of scope associated with having two funds rather than one; for example, once the adviser is doing NAV calculations for one fund, it will not cost much more to do the same for two. In this view only the adviser's incremental costs should be allocated to the new fund.

None of these methods—and one can think of others—is obviously wrong, but they lead to entirely different results when the board comes to consider the adviser's compensation fund by fund. The obvious problem here is that the board cannot, under the language of section 36(b), review and pass upon the adviser's overall profits for both funds jointly, even though they are in fact managed jointly. Section 36(b) is based on the fiction that the board and the adviser have fiduciary responsibilities to the shareholders of each fund separately, and thus the board must review the adviser's compensation for each fund apart from the others.

Under these circumstances, advisers are inevitably led to allocate costs so as to produce "reasonable" profits where they want these profits to appear, and to produce lower profits, or losses, where they want these to appear. Most independent directors probably know this but permit this arrangement because compliance with the literal language of section 36(b) is impossible. As one expert in cost accounting noted in a book on the subject, "All methods of allocating joint costs are arbitrary. . . . Allocating joint costs poses issues for accountants because the needs of allocated joint costs for financial reporting differ from those for decision making. Cost accounting systems serve multiple purposes, including product pricing, product emphasis, cost control, and reporting to internal and external constituents. The same set of computations rarely satisfies every purpose" (emphasis in original). The inherent arbitrariness of the cost allocation process, an essential element of the responsibilities laid on advisers and boards in negotiating fees and expenses, suggests that a new and different system for determining the fees and expenses of advisers is necessary—one that does away with the need for section 36(b) altogether.

The above is an excerpt from "Competitive Equity: A Better Way to Organize Mutual Funds."


6 Comments

mutual funds
These funds are for dummies who are too lazy or too dumb to check out companies themselves. Imagine paying all those fees for guys to sit in cushy offices, flirt with the secretaries, and UNDERPERFORM the market. How stupid is that. Anybody can just buy the actual companies themselves anytime. You can even check out any fund to see what their top positions are, and then get the same ones to mimic them. You can also do the same with Berkshire hathaway. Financial planners is one of the biggest scams around. but it's so lucrative, that if I were younger, I would get into the business myself. I like making money from fools.

Timing
Although you can find out what companies a fund holds at a particular time, the funds do not report these holdings often. For a fund that trades actively, the information may be out of date when reported and can be very badly inaccurate just before the next update.
Also, not all people are intelligent or wise; it's easy to lose money on stocks.
That said, your basic points are good; most funds and planners are inferior to random choices or index averages. My own recommendation for someone not willing to do a lot of work is the QQQQ.

I wonder if the same concept...
...can explain CEO salaries? Like mutual fund advisers their salaries too are set by a board of directors.

You are making this too difficult...I figured out why...
There is no reason to expect competitive forces to fully engage when competition is severely regulated or not allowed. By definition.

Of course, "the failure of pricing to approach the marginal cost of the most efficient producer...present in most areas of the U.S. economy" should not be expected with a regulated industry. However, the authors are very wrong to characterize normal pricing policy (in most industries) in America as "cost plus" with the most efficient player setting the price structure.

This paradigm might only be valid with the market for a commodity product where the most efficient producer dominates the market and uses his low costs as a predatory weapon to eliminate competition. This was the case with Swift & Company 100 years ago. And they got into a little trouble for it, didn't they?

Otherwise, the (actual) most efficient player should be able to generate extraordinary margins by selling into a larger market where his own share is minor. This is almost never the largest player. But if it was the big guy then what would be his motivation to drop prices? He already has enough market strength to hold his prices up! He is certainly not going to cut his margins to the bone across his entire line in order to protect a couple of share points. Unless he is as foolish as Dell.

In a purely competitive market with commodity products and many small producers (such as farmers) some of those entities risk producing at a loss if the market moves after the seeds are already in the ground.

Note: I wish to point out that the authors' comparison of price competition in another regulated industry (power utilities) makes no mention of The Public Utility Regulatory Policy Act (PURPA) passed in 1978. This law specifically provides for price competition through innovation, especially regarding renewable energy.

The authors are highly regarded lawyers and dedicated public servants. Peter Wallison (Harvard Law) and Robert Litan (Yale Law). Of course, Dr. Litan took his bachelors at Wharton (that's good) and his PhD in Economics at Yale. But neither seem be licensed as CPAs nor have they worked as professional accountants. It is not remarkable that garden variety, very boring Cost Accounting and Pricing routines might have been overlooked, here.

As we consider the Investment Company Act of 1940 Gartenberg v. Merrill Lynch (1981) is not nearly as interesting on its own today as certain more recent cases such as Chamberlain v. Aberdeen (2005) dealing with Section 36(a) and Wicks v. Putnam (2005) regarding Section 36(b).

"Section 36(b) explicitly confers a private right of action by a shareholder against an investment adviser for a beach of the duty not to charge excessive fees...The factors to consider in judging the disproportionality of a fee, as laid out in Gartenberg, are: (1) the nature and quality of services provided to fund shareholders; (2) the profitability of the fund to the adviser-manager; (3) fall-out benefits; (4) economies of scale; (5) comparative fee structures; and (6) the independence and conscientiousness of the trustees.

http://www.ballardspahr.com/press/article.asp?ID=1203

According to the authors Gartenberg stated "Reliance on prevailing industry advisory fees will not satisfy." However, this does not seem to say that comparative fees are completely off the table. Gartenberg simply states that competitive pricing is not sufficient.

The idea that the Board would be unaware of and unmotivated by prevailing advisory fees in the industry is both naive and specious. Of course, such people are industry professionals who are completely aware of normal business practice and the risks under Sections 36(a) and 36(b) surrounding apparent egregious misbehavior regarding their fiduciary responsibilities. They dare not ignore the competition.

One manner of calculating prices in any industry without pure competition (and commodity products) is known as "what the market will bear". This probably applies here.

Further, a more highly regarded adviser should be expected to earn a premium. So he probably gets paid better. He earnings would be factored into total costs.

The efficiencies the authors speak of as being not in the interest of the individual adviser are also not much of his concern. The firm itself does stand to gain, thereby, and the larger entity invests in productivity tools, etc.

Nevertheless, the fundamental problem here seems to be with cost accounting. With any scheme where prices (fees) are very close to costs it is critical that such costs should be understood. Otherwise the business risks operating at a loss. This is especially true when the product is a service and such costs are fixed or semi-variable.

Simply put, a Standard Cost is arrived at by adding up all the planned costs (plus a fudge factor) and dividing that total cost by the "projected" unit volume for the planning period. This calculation creates a "burden rate" to be applied per transaction unit or to some other measure of volume.

Immediately, actual volumes (and actual costs) vary from the numbers in the plan. With modern computers burden rates might be adjusted to actual results in real time. More typically Standard Costs are revised annually, twice a year at Plan Update time, quarterly or monthly (lagged). Positive and negative variances are accumulated and charged off according to GAAP.

The authors are agonizing over such cost accounting difficulties as if this is something new. These problems have been solved in the accounting profession for a very long time.

One more thing. The authors quoted Friedrich Hayek "...only through actual competition can one determine what a product or service might actually cost in a competitive environment. This is because the most important facts—such as the price at which the last consumer will buy and the marginal cost of the last input—are not known and cannot be known in advance of actual competition. The costs of inputs keep changing, as does the price the last consumer is willing to pay for a given amount of a product or service. None of these things can be forecast or discovered in advance..."

This unfortunate cherry picking from the writings of a economist from the University of Chicago whose best work was during the 1930's and whose theories were shown to be flawed by Keynes (whose own work is also now out of date regarding modern financial capitalism) might be a fine way for a lawyer to support an argument. But it leads to fundamental error.

The price at which the last consumer will purchase is very close to zero. Similarly, the marginal cost for one more transaction at the fund adviser's desk is also very nearly zero. This is all well known, far in advance and it does not mean a thing.

your school of economics?
So you don't like Hayek, or the chicago school, and admit that Keynes too is not valid. Then what sort of economics do you actually like? If you like competition and free markets, then wouldn't you also like Austrian economics? They worked out all this stuff years ago too, in the 30s, and it's still the best around.

Economics is a philosophy...
Just like most philosophical paradigms economics attempts to present a "theory of everything" that will allow economists to understand and to predict economic events.

This is like sociology, Dietmar. I don't mind if the scholars try to understand the phenomena if they would genuinely look into the underlying behavior of the mechanisms.

What bothers me is these lame explanations for events that have already gone past (as if the economists have it nailed) and this pathetic guessing game regarding what the markets will do going forward.

There are too many variables for their calculations. They cannot develop perfect data and they cannot solve the algorithms.

Therefore, just as we do in the business world as we mature into decision makers, an economist should be fully immersed in the processes of financial capitalism (and general management) and he should thereby develop a pointed understanding of what seems to be going on and what should be coming next. Economic models are fun. But they are more vocabulary than formulas.

What I see with most economists are brilliant academics who have never worked. They reason from assumptions that are fundamentally ridiculous. Any heavy CFO type sees these flaws immediately. They are the same junior mistakes we must correct in our new hires. Economists will vastly improve their understanding of market behavior if they work in finance and accounting enough to understand that such matters are often not intuitive. And to refine the stupidities out of their assumptions.

I did not answer your question about my school. I went to Wharton. I studied accounting and finance. Economics kept countradicting itself. I also don't believe a thing the sociologists have to say. Before the MBA I was a (Masters level) behavioral biologist.

TCS Daily Archives