The Separation of Funds and Managers: A Theory of Investment Fund Structure and Regulation
abstract.This Article offers a broad theory of what distinguishes investment funds from ordinary companies, with ramifications for how these funds are understood and regulated. The central claim is that investment funds (i.e., mutual funds, hedge funds, private equity funds, and their cousins) are distinguished not by the assets they hold, but by their unique organizational structures, which separate investment assets and management assets into different entities with different owners. In this structure, the investments belong to “funds,” while the management assets belong to “management companies.” This structure benefits investors in the funds in a rather paradoxical way: it restricts their rights to control their managers and to share in their managers’ profits and liabilities. The fund investors accept these restrictions because certain features common to most investment funds make these restrictions efficient. Those features include powerful investor exit rights that substitute for control rights and economies of scope and scale that encourage managers to operate multiple funds at the same time. This understanding illuminates a number of key areas of contracting and regulation and refutes the claims of skeptics who say that fund investors would be better off if they employed their managers directly.
author.Associate Professor of Law, Yale Law School. This paper benefited from workshops at the American Law and Economics Association Annual Meeting, Columbia Law School, UCLA Law School, the University of Illinois College of Law, the University of Virginia School of Law, and Yale Law School. It also benefited from comments by Douglas Baird, William Birdthistle, Anthony Casey, Quinn Curtis, Jonathan Greenstein, Henry Hansmann, Richard Hynes, Bruce Johnsen, Paul Mahoney, Peter Molk, Larry Ribstein, Roberta Romano, and Andrew Verstein. Stephen Sovinsky provided excellent research assistance.
Investment funds control a vast amount of wealth. By some estimates, the various types of funds—including mutual funds, hedge funds, private equity funds, venture capital funds, exchange-traded funds, and closed-end funds—collectively hold about $18 trillion in the United States. They hold substantially more assets than the commercial banking system and almost enough assets to equal the value of all domestic equity securities listed on the New York Stock Exchange and NASDAQ combined.1
Legal scholars have largely overlooked these enterprises, however. Although financial economists have extensively studied investment funds’ performance and operation,2 legal scholars have rarely examined their structure and regulation. Little is known about foundational issues such as why investment funds adopt their basic patterns of organization, why they are regulated differently from operating businesses, and how we can distinguish them from operating businesses.
In this Article, I offer a broad perspective by suggesting for the first time that the essence of investment funds and their regulation lies not just in the nature of their assets or investments, as is widely supposed, but also—and more importantly—in the nature of their organization. An investment fund is not simply an enterprise that holds “investments”—it is an enterprise that holds investments in a particular way. And the regulations that govern these funds are at least as concerned with the funds’ peculiar patterns of organization as with their peculiar kinds of assets.
Specifically, nearly every enterprise that we commonly think of as an investment fund adopts a pattern of organization that I call the “separation of funds and managers.” These enterprises place their portfolio securities, currency, and other investment assets and liabilities into one entity (a “fund”) with one set of owners, and their managers, workers, office space, and other operational assets and liabilities into a different entity (a “management company” or “adviser”) with a different set of owners. The Fidelity management company, for example, operates several dozen mutual funds under the Fidelity brand name. Each fund is a separate entity with separate owners, and so is Fidelity itself.3 In connection with this system, investment enterprises also tend to radically limit fund investors’ control. A typical hedge fund, for example, cannot fire and replace its management company or its employees—not even by unanimous vote of the fund’s board and equity holders.4
This pattern remains poorly understood. To my knowledge, it has never been identified as a common feature of the various types of investment funds. And in commentary about individual types of funds—particularly mutual funds—the pattern is often viewed as a kind of scam, depriving investors of the rights to control their managers and share in their managers’ profits.5
This Article thus identifies the separation of funds and managers as a central phenomenon and explains its functions. I argue, paradoxically, that this pattern benefits fund investors for precisely the reasons it is often said to harm them: it limits their control over managers and their exposure to managers’ profits and liabilities. These limits are particularly efficient in investment funds for three reasons. First, most fund investors have special rights of exit and are protected by unusually strong incentive compensation systems. Control thus resides more efficiently in the hands of management company investors than in the hands of fund investors. Second, fund investors have particularly strong desires for precision in the tailoring of risk. They desire exposure only to the risks of investment assets and not to the risks of management businesses. Third, fund managers can achieve economies of scope and scale by simultaneously managing multiple funds. For reasons I explain, this simultaneous management is only possible when fund investors’ control rights and exposures to managers’ earnings and liabilities are tightly restricted by the separation between funds and managers.
This multifaceted explanation may seem complicated, but the intuition behind it is simple: in terms of their rights and risks, fund investors look more like buyers of products or services than like investors in ordinary companies. In much the same way that product buyers can “exit” by refusing to buy more products, for example, fund investors can often exit by withdrawing their money and refusing to pay managers’ fees. And in much the same way that product manufacturers simultaneously produce multiple products, fund managers simultaneously operate multiple funds. The reasons why fund investors have no ownership or control over their management companies thus resemble the reasons why product buyers generally have no ownership or control over their manufacturers.6
This positive explanation for investment funds’ basic structure has extensive normative implications. Among other things, it illuminates the function of investment fund regulation. The key function of the Investment Company Act of 1940 (ICA),7 it turns out, is not merely to address the peculiar problems created by securities investing, as is widely believed, but also to address the peculiar problems created by the separation of funds and managers. This theory also suggests we should change the way we define investment funds. Regulation and academic research currently distinguish investment funds from operating businesses by looking at the nature of their assets.That is, an investment fund is said to be an enterprise that holds a large amount of securities. I suggest, however, that the essence of an investment fund lies more precisely in the nature of its organization. Focusing on organization solves an array of practical and conceptual problems, such as how to distinguish private equity funds from conglomerates.
The paper proceeds as follows: Part I reviews the literature and provides a basic description of investment funds. Part II describes the separation of funds and managers and its consequences. Part III explains this pattern by identifying the peculiar features of investment funds that make the pattern efficient. Part IV addresses objections to this explanation and applies the explanation to several special cases, including closed-end funds, asset securitization vehicles, and donative trusts. Part V suggests policy reforms.
The separation of funds and managers is nearly universal among the types of enterprise that we commonly think of as investment funds. An overview of the different types of funds may therefore be useful.8 By far the largest category of fund is the “mutual” fund. Mutual funds are commonly defined as pools of stocks, bonds, and other investment securities. Mutual funds issue shares in these pools to members of the public, who can buy the shares to invest for retirement or other purposes. Because mutual funds sell their shares widely to the general public, they must register with the SEC and comply with the ICA.
A key feature of mutual funds is that they allow their shareholders to “redeem” their shares. In other words, shareholders can return their shares to the funds in exchange for the cash value of their shares. This value equals the value of the portion of a fund’s net assets that corresponds to each share. Mutual funds typically allow their shareholders to redeem every day. “Exchange-traded” funds (ETFs), which have received a great deal of attention in recent years, are a special kind of mutual fund.9
Mutual funds are sometimes called “open-end” funds to distinguish them from “closed-end” funds. Closed-end funds are similar to mutual funds in many respects. They are pools of investment securities, they sell interests widely to the general public, and they must comply with the ICA. The primary difference is that closed-end funds do not allow shareholders to redeem. Rather than redeeming, closed-end fund shareholders dispose of their shares by selling them on stock exchanges, just as they might do with the shares of operating companies.
“Hedge” funds are pools of investment assets similar to mutual funds. But unlike mutual funds, hedge funds issue securities only to limited numbers of institutional investors and wealthy individuals. Hedge funds therefore do not have to register with the SEC or comply with the ICA.10 Like mutual funds, hedge funds allow their shareholders to redeem their shares. However, hedge funds typically allow redemptions only once per month or once per quarter, rather than daily as mutual funds do.11
“Private equity” funds are similar to hedge funds in the sense that they sell only to wealthy individuals and institutions and therefore do not register with the SEC or comply with the ICA.12 But private equity funds do not offer redemption rights. Instead, they exist for terms of years—usually around five to ten years—after which they wind up by distributing their assets or by selling their assets and distributing the proceeds.
People sometimes refer to “private equity” funds by other names that more specifically describe the funds’ investment strategies. “Venture capital” funds, for example, tend to invest primarily in companies that are relatively new and risky.“Buyout” funds tend to buy large and controlling stakes in a small number of established operating companies. Other types of private equity funds invest in real estate, distressed debt, and other asset classes.
“The separation of funds and managers,” is, in a sense, widely known. In fact, it is almost too widely known. It runs so deep in the DNA of the investment management industry that its significance has tended to be felt implicitly rather than identified expressly. It is the forest that has gotten lost among the trees.
To my knowledge, no one has yet identified the separation of funds and managers as a unifying feature common to all types of investment funds. The pattern has tended to attract attention only in the context of individual types of funds—primarily mutual funds.
Among observers of mutual funds, discussion of the separation of funds and managers has produced two related themes. The first is that despite the separation of mutual funds and their managers, mutual funds tend to be dominated by their managers.13 This theme appeared recently, for example, in amicus briefs and commentary regarding the Supreme Court’s opinion in Janus Capital Group, Inc. v. First Derivative Traders,14 an important case about the structure of mutual funds, which I address below.15
A second theme is that this managerial dominance is deeply problematic. The best-known proponent of this view is Jack Bogle, founder of Vanguard, one of the nation’s largest mutual fund management companies. Bogle has written numerous books and articles arguing that the separation between mutual funds and their managers harms fund investors.16 In the late 1990s, Bogle went so far as to prophesy that separately owned mutual fund management companies would cease to exist in the ensuing twenty-five years.17 His basic argument is that separation places funds and managers in conflict by depriving fund investors of the rights to control their managers and share in their managers’ profits.18
Bogle’s skepticism is widely shared19 and has been for a long time. Suspicion of the separation of funds and managers was particularly widespread in the 1960s, when the SEC criticized this pattern and called for reforms to address it.20 The reforms included the “excessive fee liability” that the Supreme Court recently addressed in Jones v. Harris Associates.21
Despite all of this criticism, however, no one has yet made a sustained attempt to explain just why the separation of funds and managers occurs. This Article thus reaches beyond the existing literature in two ways. First, it explains the separation of funds and managers in terms of efficiency, rather than regulation or exploitation. Second, it identifies the separation of funds and managers as the central and defining feature of investment funds.
Every type of investment fund adopts a pattern of organization that I am calling the “separation of funds and managers.” I choose this phrase partly to evoke the infamous “separation of ownership and control” in ordinary companies,22 while making it clear that investment fund organization involves something different from and more extreme than the simple delegation of decision-making authority that occurs in ordinary companies. Let us now turn to describing this pattern and its consequences.
Investment funds begin life through the efforts of management companies. Management companies, which are also sometimes known as “general partners” or “advisers,”23 establish funds for the purpose of attracting investors and charging them fees. The ranks of management companies include small businesses with only a few employees, as well as large financial conglomerates, such as Fidelity, Goldman Sachs, BlackRock, and Bank of America. Financial conglomerates often operate investment fund management units alongside other lines of business, such as investment banking, commercial banking, and brokerage.
A management company establishes a fund by taking several steps, often simultaneously. First, the management company organizes the fund as a legal entity distinct from the management company. Next, the management company causes the new entity to sign a contract with the management company.24 Under the terms of the contract, the management company agrees to supply all of the operational and administrative services the fund requires.25 The fund thus has no employees or other operational assets and may even be prohibited from having them.26 The agreement further gives the management company sole authority to direct the fund’s operations and investment strategy. In private equity and hedge funds, agreements also often limit the funds’ abilities to remove or replace management companies and their employees.
Once the agreements and entities are in place, the management company solicits investments from outsiders. Note that the management company and its employees might invest some of their own money alongside the outsiders.27 However, so long as the ownership of funds and their management companies do not perfectly overlap, funds and management companies must deal with one another at arm’s length and may be said to have separate ownership.
Management companies often repeat this procedure to establish and operate many different funds simultaneously or serially over time. The largest management companies operate thousands of funds. Management companies often market their many funds under a single brand name, such as Fidelity, BlackRock, Cerberus, or D.E. Shaw, and funds that share a common management company are often said to be part of the same “family” or “complex.”
Once the management company has established the fund and its contracts, the fund will own what I am calling “investment” assets and liabilities, and the management company will own what I am calling “management” assets and liabilities. Investment assets typically (though not always) consist of things like securities, currency, real estate, and assorted financial instruments, and investment liabilities consist of things such as outstanding loans and unperformed derivative obligations. Management assets include the written and unwritten contracts that entitle management companies to the services of workers (such as portfolio managers, accountants, and secretaries), as well as physical assets such as office space and computers and intellectual property such as corporate logos. Management liabilities include obligations under office leases, employment contracts, and other commitments. In management companies that operate inside of larger financial conglomerates, such as Goldman Sachs, the management assets may be far-flung, highly complex, and only loosely related to the management of investment funds. The risks associated with these vast assets and liabilities will turn out to be especially important in explaining the separation of funds and managers in financial conglomerates.
This pattern of organization produces two functional features of great importance: first, the use of separate entities to hold investment assets and management assets, and second, the construction of separate ownership for those entities. These two features in turn produce three important limits on fund investors’ rights and risks with respect to management companies. These limits are important, because the essence of my theory is that the separation of funds and managers has arisen for the purpose of achieving them.
The first of these limits is a product of the use of separate entities to hold management assets and investment assets. The use of separate entities limits the exposure of fund assets to management companies’ liabilities and creditors. In other words, the use of separate entities achieves “asset partitioning.”28 By virtue of the fact that funds and management companies are separate entities, any creditors who contract with a management company have no claims against the assets of the funds. This reality became dramatically evident during the financial crisis of 2008, when several management companies, including Lehman Brothers and Washington Mutual, collapsed and went into bankruptcy. When these companies entered bankruptcy, the assets of their funds remained free from the managers’ creditors, except to the extent that the management companies owned shares in the funds.29
The second limit is a product of the construction of separate ownership for funds and management companies. Separate ownership limits fund investors’ right to claim management companies’ residual earnings or “profits.” Because funds and management companies have separate owners, it is the management companies’ equity holders, and not the funds’ equity holders, who possess the rights to claim the management companies’ residual earnings. As a consequence of these limits, fund investors experience neither the benefits nor the burdens of management companies’ profitability.
The third limit is also a product of separate ownership. Separate ownership prevents fund investors from exercising residual control over management companies.30 Because funds and management companies are separately owned, it is the management companies’ equity holders, and not the funds’ equity holders, who exercise residual control over management companies and their assets.
We can construct a complete list of alternatives to the separation of funds and managers by imagining what would happen if we did away with one or both of the two key functional features that produce the limits on fund investors’ rights and risks just described—i.e., the use of separate entities and the construction of separate ownership for those entities. None of the alternatives that omit one or both of these two features would be sufficient to achieve the three limits on fund investors’ rights and risks just noted (i.e., the limits on liabilities, residual earnings, and control).
One alternative might be to do away with both separate entities and separate ownership. Investment enterprises might accomplish this by placing both operational assets and investment assets into a single entity with a single set of owners. In other words, investment enterprises could “internalize” their management by employing their workers and owning their operational assets directly. This would cause investment enterprises to resemble prototypical operating companies.
This alternative would achieve none of the three limits on fund investors’ rights and risks discussed above. The funds’ assets would be exposed to the management creditors because the funds’ assets would reside inside of the same entities that contracted with the management creditors. And fund investors would exercise both residual control and residual earnings rights over management assets, because fund investors would own the management assets.
A second option would be to hold investment assets and management assets in separate entities, but combine the ownership of those entities. For example, an investment enterprise could place investment assets and management assets into distinct but wholly owned subsidiaries of the same parent company.
This alternative might limit the exposure of fund assets to management creditors, because management creditors could be forced to contract only with the management entity. But this alternative would nevertheless give fund investors residual earnings and residual control rights over the management assets, because the management and fund entities would reside under common ownership.
A final alternative would construct separate ownership, but not separate entities. Investment enterprises might, for example, hold both investment assets and management assets in a single entity and then issue two classes of “tracking stock.” One class might offer residual earnings and control only with respect to the investment assets, and the other class might offer residual earnings and control only with respect to the management assets.
This approach would fail to insulate investment assets from management creditors, because, as Edward Iacobucci and George Triantis have observed, creditors’ claims are always entity-wide.31 That is, management creditors could always potentially seize any of the investment assets that were held by the same entity that held management assets, regardless of whether the entity had issued tracking stock. Further, conflict-of-interest rules and other restrictions prevent the attachment of residual earnings and control rights to subsets of an entity’s assets.32 These problems together have made tracking stock unworkable for ordinary companies, and would make tracking stock similarly unworkable in investment enterprises.
Of course, it may be possible to achieve asset partitioning inside of a single legal entity more effectively by using a specialized business form, such as a Delaware statutory business trust. These forms allow a single entity to divide assets among two different series of beneficial interests and to give each series its own distinct creditors and its own distinct legal personhood.33 The availability of these forms does not diminish the need for separate entities with separate owners, however, because these forms just create functionally separate entities inside of nominally unified ones. Although to my knowledge no fund families actually use these forms to hold the assets of both funds and managers, if they did they would be achieving the essential features of the separation of funds and managers—i.e., separate creditors and separate owners.
The explanation for the separation of funds and managers may now be stated simply: this pattern has arisen for the purpose of achieving the previously noted limits on fund investors’ control rights over management companies and on fund investors’ exposures to management companies’ liabilities and residual earnings. These limits are desirable because they maximize the aggregate value of investment enterprises and so increase the total wealth available to everyone who contracts with them. This pattern permits managers to obtain greater profits and hence enables investors to obtain lower fees and better returns with fewer risks.
This explanation may also be stated a bit more technically. The separation of funds and managers represents a peculiar approach to two optimization problems. These problems arise in every type of enterprise and together shape the modern theory of the firm. The first optimization problem involves the degree of asset partitioning. Enterprises can partition their assets to a greater or lesser degree by choosing to divide their assets among a greater or lesser number of distinct legal entities. As Iacobucci and Triantis and others have shown, asset partitioning involves both costs and benefits, and enterprises generally choose the degree of partitioning that they find optimal in their particular situations.34
The second optimization problem involves the allocation of ownership rights. As Henry Hansmann and others have shown, enterprises can allocate ownership rights to any of their various classes of patrons, such as their investors, workers, customers, and so on.35 Enterprises generally assign ownership rights among these patrons so as to minimize the aggregate costs to all of these patrons and to maximize the aggregate benefits. To put this point a bit crudely, enterprises typically give ownership rights to the patrons who value those rights most highly and who are least likely to use them to the detriment of others.
The separation of funds and managers may thus be viewed as a particular way of resolving these two problems. The separation of funds and managers resolves the first problem (asset partitioning) by partitioning investment assets and management assets into separate entities. The separation of funds and managers solves the second problem (ownership) by giving ownership of management companies to management company investors, rather than to fund investors. This peculiar resolution of these two problems creates the limits on fund investors’ rights and risks that are so important.
To be clear, investment funds probably settled upon this efficient pattern of organization through natural selection rather than deliberate choice. A few of the earliest mutual funds experimented with combining funds and managers in the 1920s, and a small handful of them continued to do so up through the mid-1960s.36 Eventually, however, competitive pressures apparently forced almost every mutual fund and other type of investment fund to adopt the separation of funds and managers.
The goal in this Part will therefore be to identify the peculiar features of investment funds that shifted the balance of costs and benefits in the two optimization problems noted above in favor of enterprises that limited fund investors’ rights and risks through the separation of funds and managers. I identify three such features: (1) exit rights and performance incentives that substitute for control; (2) an unusually strong need for precision in the tailoring of risk; and (3) the economies of scope and scale that arise when managers simultaneously or serially operate multiple funds.
First, most fund investors have unusually strong exit rights and most fund managers have unusually strong performance incentives. Both of these features take the place of control, and as a consequence control over management companies resides more efficiently in the hands of management company investors than in the hands of fund investors. Fund investors thus benefit from the limits on control, because the placement of control in the hands of its most efficient users allows investment funds to offer lower fees and better returns. Let us begin by focusing on exit rights before turning to performance incentives.
The mechanical details of exit are complicated, and so a brief initial summary may be useful. Except for a few special types of funds, which I discuss in the next Part, investors in every type of investment fund have the right periodically to withdraw their money or otherwise remove it from managers’ control. In mutual funds and hedge funds, investors withdraw by redeeming their shares and demanding the value of the corresponding assets in cash. In private equity funds, investors receive periodic liquidations. Every five to ten years, private equity funds liquidate all of their assets and distribute the proceeds in cash.
These rights of exit place investment funds in stark contrast to ordinary companies. In ordinary companies, investors can sell their shares, but they cannot remove the assets that underlie the shares from the companies’ possession. An ordinary company retains ownership of its assets no matter who owns its shares.
As a result of exit and withdrawal rights in investment funds, fund investors value control rights less than ordinary company investors do. Fund investors do not value control, because if they are unhappy, they can simply remove their money and take it elsewhere.
Of course, the strength of exit rights varies significantly among different types of funds. Mutual funds offer strong exit rights in the form of daily redemptions while private equity funds offer much weaker exit rights in the form of relatively infrequent liquidations. This variation in the strength of exit rights actually bolsters the claim about the significance of exit rights, however. As we shall see, as fund investors’ exit rights become stronger, control rights and contractual protections become weaker, and as exit rights become weaker, control rights and contractual protections become stronger. The separation of funds and managers is thus most problematic in closed-end funds and private equity funds, where exit is relatively limited and is least problematic in mutual funds and hedge funds, where exit is relatively free.
The effect of withdrawal and liquidation rights can be expressed in several different ways using the various conceptual languages that appear in the economic literature on the theory of the firm. Using the language of Albert Hirschmann, we can say that in investment funds, “exit” substitutes for “voice.”37 Using the language of Oliver Williamson, we can say that fund investors do not demand control, because they do not make “specific investments.”38 Specific investments are those that make it costly for investors to move their money or resources to other investments. Investors who make specific investments demand control, because they would otherwise be vulnerable to hold-up and exploitation. Investments in a fund are less “specific” than investments in ordinary companies, because in a fund, investors can move their money with relative ease by redeeming or refusing to reinvest after liquidation. Investors in ordinary companies make highly specific investments, because their money is locked in and cannot be removed unless directors decide to pay dividends.39
We can also express this point using the language of Henry Hansmann by returning to the product analogy noted above.40 Exit rights cause fund investors to resemble product buyers. Just as product buyers can sever their relationships with suppliers by refusing to buy the products any longer, so too can fund investors sever their relationships with management companies by removing their assets and refusing to pay the managers’ fees any longer. If I do not like Pepsi, I will simply switch to Coke. Similarly, if I do not like Fidelity, I will simply switch to Vanguard. The right to vote thus carries almost as little value for mutual fund investors as it would for product buyers.41
Let us now turn to the mechanical details. I first consider how exit rights diminish the value of control in mutual funds. I then examine hedge funds and private equity funds. I address closed-end funds and related vehicles as special cases in Part IV below.
Mutual funds offer the freest exit of any major type of investment fund. In mutual funds, each investor can redeem her shares unilaterally every day.42 For a detailed description of how exit rights reduce the value of control rights in mutual funds, I refer readers to a recent article that I co-wrote with Quinn Curtis. In that article, we developed a conceptual model of the effect of exit rights on control rights in mutual funds.43 We defined how investors choose between exit and voice and answered objections in detail. In this Article, I will describe only the basic intuition, drawing on that prior paper.
The intuition begins with the observation that as a consequence of exit rights, mutual fund share prices always equal exactly the net asset value (NAV) of the issuing funds.44 The NAV is the value of the pro rata portion of a fund’s assets—net of liabilities—that corresponds to each share. NAV’s key feature is that it does not reflect expectations about future fees or future portfolio changes. The value of fees yet to be paid and portfolio changes yet to be made does not show up in NAV, because NAV is based simply on the value of the securities in a fund’s portfolio on any given day. Put differently, the NAV is liquidation value, not expected value. To be sure, NAV reflects fees paid and portfolio decisions made in the past, but not in the future.45 And of course, NAV might reflect expectations about the future performance of the companies in a fund’s portfolio, but not expectations about future fees and portfolio changes in the fund itself.
The intuition about how exit undermines voting is thus very simple: investors in mutual funds can exit before bad news affects their share prices; investors in ordinary companies cannot. If, for example, a mutual fund adviser announces today that its fees will increase starting tomorrow, investors can redeem today at a price that does not reflect the fee increase. The fees will not be charged until tomorrow, and so they simply do not enter into the calculation of NAV today. Rather than voting to lower fees, therefore, mutual fund investors will simply redeem and move to other funds. By contrast, if an ordinary company announces today that its operating expenses will increase starting tomorrow, the price of the company’s shares will decline immediately in expectation of the increase. Investors can sell today, but only at a low price reflecting the costs expected tomorrow. Unlike investors in mutual funds, therefore, investors in ordinary companies might sometimes use voting to prevent the cost increase and raise the share price.
This simple intuition can also be stated somewhat more formally. An investor who is unhappy with some aspect of a fund—for example, the fees46—has three options: (1) she can exit by redeeming; (2) she can do nothing; or (3) she can try to improve the situation by voting. Exit makes the third option (voting) much less appealing in mutual funds than in ordinary companies. To see why, we need only compare the costs and benefits of voting to the costs and benefits of the other two options (exiting and doing nothing).
Consider first the relative benefits of exit over voting. Exit rights are simply not available in ordinary companies, and so if exit offers even slightly greater benefits or slightly lower costs than voting in mutual funds, it will tend to make voting less attractive in mutual funds than in ordinary companies. In practice, it turns out that the net benefits of exit are almost always greater than the net benefits of voting. The greatest possible benefit that voting can produce is to reduce an adviser’s fees to its marginal costs.47 It turns out, however, that this is precisely the same benefit that exit can achieve. Marginal cost is the price that basic microeconomic theory predicts will prevail in a perfectly competitive market. And it is the level that most observers agree does in fact prevail in at least some portion of the mutual fund market (though perhaps not in all of the mutual fund market).48 Hence, redeeming and switching to a fund that operates in the competitive portion of the market can almost always produce at least as much benefit as voting. And redeeming and switching will tend to produce even more benefit than voting whenever voting is not perfectly effective. That is, exit will produce more benefit than voting whenever voting fails to reduce fees all the way to marginal cost, which is probably most of the time.
The balance of costs also favors exit over voting. Voting in a mutual fund is costly for the same reasons that voting in an ordinary company is costly: investors face a collective action problem. The burdens of voting and activism are concentrated among the activists, while the benefits are dispersed among all shareholders. Exit, by contrast, does not create a collective action problem, because exit does not involve collective action at all. Mutual fund investors can redeem unilaterally.
Of course, exit involves costs of its own, but many of these costs also appear in voting.49 Exit requires time and sophistication, for example, but voting requires far more. If redemption is hard, running a proxy contest is vastly harder. And although exit requires the realization and payment of taxes, the significance of this problem is limited.50
Now consider option (2) from above: the possibility that an investor might simply do nothing. Unlike the option to exit, the option to do nothing is not unique to mutual funds—investors in ordinary companies can also choose to do nothing. Nevertheless, doing nothing turns out to be more appealing in mutual funds as a result of exit rights. Exit rights diminish the benefits of voting—and hence increase the appeal of doing nothing—because exit prevents shareholders from realizing the full benefits of voting. When shareholders vote, they produce improvements whose benefits may be realized both in the present and the future. The NAV pricing produced by redemption rights, however, does not reflect expectations about the future. The NAV today is unaffected, for example, by a vote that reduces fees in the future. Hence, although activists may enjoy the benefits of their voting during the time that they remain invested in a fund, they cannot sell at prices that reflect the benefits of their voting that will be realized after they withdraw. This is not the case in ordinary companies. In ordinary companies, activists can sell at prices that reflect the full expected value of their shares—including the benefits of activism to be realized in the future—because these benefits are built into share prices. With the benefits of voting in mutual funds diminished in this way, the costs of voting loom even larger, making option (2)—doing nothing at all—even more appealing.
The bottom line, therefore, is that investors will rarely, if ever, choose to vote in a mutual fund. They will choose instead either to redeem or to do nothing.
This argument holds under any reasonable view of mutual fund market competition, because almost every commentator agrees that at least some funds in almost every investing style offer competitive fees and returns.51 Thus, even if a large portion of the market is not competitive, investors will nevertheless still prefer exit over voting, because they can still expect to locate at least one competitive fund in almost any investing style.
This argument also holds under any reasonable view of mutual fund investors’ size or sophistication. Although large and sophisticated investors often become active in ordinary corporations, they do not become active in mutual funds, because even large and sophisticated investors stand to gain more from exit than from activism. And although small and unsophisticated investors will sometimes fail to exit mutual funds because they lack the time or knowledge to do so, they will also fail to vote for the very same reasons.
Real-world practices confirm that shareholders do not value control rights in mutual funds. As noted above, shareholders in these funds give up residual control over management companies by separating funds and managers. And these shareholders also go even further, actually giving up control over their own funds by completely neglecting voting rights and the other institutions of control in those funds.
In a well-intentioned effort to protect investors, the ICA requires mutual funds to give their shareholders a minimum set of control rights. Funds must have boards of directors who are nominally elected by shareholders and must also have the right to terminate and replace their management companies.52 In actual practice, however, investors almost never use these mandated control rights. Unlike in ordinary public companies, in mutual funds, shareholder activism is not simply uncommon—it is unheard of. I know of no director election or other important matter that has ever been contested by the impetus of shareholders in the ninety-year history of the open-end mutual fund industry.53
Moreover, funds invariably minimize the ICA’s mandated control rights to the greatest extent allowed by law. The ICA requires mutual funds to give shareholders the right to vote for directors, but funds minimize the effect of these rights by allowing board members to serve indefinitely without reelection and to appoint many of their own replacements.54
Hedge funds confirm this understanding of exit and control. Like mutual funds, hedge funds offer exit in the form of redemption rights at NAV. The primary wrinkle is that hedge funds do not allow redemption every day; they allow it instead only once every month or quarter. The effect of redemption on the value of control rights nevertheless remains extremely powerful in hedge funds. These funds not only restrict fund investors’ control over management companies by separating funds and management, they also restrict fund investors’ control over their own funds by writing express limits into the funds’ operating agreements.
Hedge funds accomplish this through an array of contractual devices. Most obviously, many hedge funds organize as limited partnerships.55 This gives managers exclusive control, because the investors in these funds are limited partners while the managers are general partners. Under most states’ limited partnership statutes, limited partners cannot exercise control without exposing themselves to unlimited personal liability for the fund’s debts.56 Thus, in many hedge funds, investors are not even the formal residual controllers of their funds—managers are.
Moreover, even when funds do not organize as limited partnerships and instead adopt more permissive organizational forms, such as the LLC, the funds’ organizing documents invariably contain provisions that expressly prohibit fund investors from ever directing the management companies on matters of business strategy.57 Additionally, many hedge funds organized as LLCs do not have boards of directors or similar bodies of investor representatives.58 To the extent that funds do have directors, it is typically because quirks of law in offshore jurisdictions require it.59
Operating agreements also often explicitly cut off fund investors’ authority to hire and fire the management companies and their employees. A fund typically has no authority at all to fire its individual managers, and it cannot fire the management company except in response to egregious fraud or misconduct. This makes the removal of a manager an extreme and rare remedy. The prevailing ethos is that if a fund investor wishes to change how her money is managed, she should just redeem.
In private equity funds, exit rights are much more limited than in open-end mutual funds and hedge funds. Private equity investors cannot redeem. Instead they must wait for the funds to liquidate, typically every five to ten years. At the end of these periods, the funds sell all of their assets, distribute the proceeds, and wind up their affairs.60
These limits make exit a much less complete substitute for control in private equity funds than in mutual funds and hedge funds. Nevertheless, the value of control is still lower in private equity funds than in ordinary companies as a result of exit. Private equity relationships are much briefer than equity investor relationships in ordinary companies. By the time private equity investors become dissatisfied, a fund’s life may be largely over. Shareholders thus place less value on the right to make changes through the exercise of control than they would if their commitments to the funds lasted forever, as in operating companies.
Additionally, because each fund periodically liquidates, management companies must periodically raise new funds if they wish to stay in business. The desire to raise new funds creates a kind of product-market competition dynamic, because management companies can only attract new investors if the management companies can demonstrate success in prior funds.61 Private equity management companies are thus subject to frequent reputational pressures and discipline from capital markets in a way that managers of ordinary companies are not. And this discipline may indeed occur quite frequently. Most private equity funds invest their money early on in their lifecycles, and then wait for a few years to let the investments grow before liquidating. Most private equity agreements therefore allow management companies to begin raising new funds as soon as all of the money in prior funds has been invested.62 Thus, even though funds typically last for five to ten years, and can last even longer, managers may attempt to raise new funds as often as every two to five years.
Nevertheless, as a result of the restrictions they place on exit rights, private equity funds give investors greater contractual protections and slightly greater control rights than hedge funds do. Private equity funds thus nicely illustrate the powerful link between exit rights on the one hand, and control rights and contractual protections on the other.
As in hedge funds, in private equity funds the authority to make strategic decisions belongs almost exclusively to management companies, rather than to the funds, and the funds cannot fire management companies or management company employees except in extreme circumstances.63 Unlike hedge funds, however, private equity funds protect investors with an array of contractual limits. Private equity fund operating agreements typically specify many more details about investment strategy than hedge fund operating agreements do, and private equity agreements often have provisions that allow fund equity holders to dissolve a fund if specified management company employees die, leave, or otherwise fail to devote sufficient time to management of the fund.64 Moreover, as I explain below, private equity funds place much stronger restrictions on managers’ ability to take on conflicts of interest by operating multiple funds than hedge funds do.65
Additionally, private equity funds give investors some voting rights, albeit very limited ones. Private equity funds often have shareholder-representative bodies that go by names like “advisory boards.”66 Unlike boards of directors in ordinary companies, advisory boards in private equity funds do not have day-to-day authority to manage the funds’ affairs and cannot hire and fire employees.67 But these advisory boards often do have the right to veto certain conflict-of-interest actions and some other types of transactions.68 They also sometimes have authority to call votes of the equity holders on certain fundamental matters such as dissolution of the fund or removal of its managers. These modest control rights place private equity funds in an intermediate space between hedge funds and ordinary companies, which is consistent with the intermediate level of exit rights they offer.
Many readers will object that exit rights in investment funds are not always perfectly free. In private equity funds, for example, investors must wait for liquidation, which typically occurs every five to ten years, and can sometimes take longer. In hedge funds, investors face initial lock-up requirements, which force them to keep their money committed for a fixed amount of time when they first invest, and also occasionally face “side-pocket” arrangements that allow managers to restrict redemptions with respect to portions of the funds’ portfolios. In mutual funds, investors often lack the time or financial sophistication to monitor their funds and redeem as necessary. Investors in all types of funds must realize and pay taxes when they exit.
I have addressed some of these objections in detail elsewhere and have explained why they may not be as serious as they initially seem.69 For present purposes, however, it is enough to stress that my point here is not to say that exit rights are perfect. Rather, my point is to build a framework in which we can understand the effect of variations in the strength of exit. The general pattern observed above is that as exit rights become weaker, control rights and other contractual protections become stronger, and as exit rights become stronger, control rights and other protections become weaker.
This framework is useful for explaining many previously inexplicable phenomena. It tells us, for example, why private equity investors demand so many more contractual protections than hedge fund investors do. This understanding will also explain below why closed-end funds are so unpopular. The idea, basically, is that closed-end funds offer no exit rights to compensate for their restrictions on control.
It is also important to keep these restrictions on exit rights in perspective. Even if exit is costly or uncertain, it can still diminish the value of control. So long as exit remains even a possibility—though perhaps a costly or difficult one—voting will be less valuable than it would be if exit were not a possibility at all (as in ordinary companies). If exit diminishes the value of control to fund investors even just partially, then fund investors will be more likely to give up control to management company investors than if exit were not a possibility at all.
Although exit is the most important of fund investors’ substitutes for control, it is the not the only one. In private equity and hedge funds, investors are further protected by extremely powerful performance incentives. Managers’ fees are tied to funds’ performance,70 often giving managers around twenty percent of a fund’s positive returns.71 Managers are also often required to invest their own money in the funds. These incentives take the place of control by encouraging managers to act in investors’ interests even in the absence of control.
The natural variation in the strength of these incentives across different types of funds also helps to confirm the observations above about the role of exit. Performance incentives tend to be strongest and to receive the greatest emphasis in the types of funds that offer the weakest exit rights. Mutual funds—which have the strongest exit rights—rarely charge large performance fees.72 This is consistent with the theory that exit and performance fees are substitutes for one another, as well as for control.
With this understanding of control rights in place, let us now consider a second reason for the separation of funds and managers: fund investors place an unusually high value on precision in the tailoring of risk. They wish only to be exposed to the risks of investment assets, and not to the risks of the management businesses that administer those assets. Returning to the product analogy, we might say that one of the products that fund management companies sell is a willingness to bear operational risks in return for a fee, so that fund investors can enjoy exposure only to the risks of investment assets.
Consider, for example, S&P 500 index mutual funds. Roughly speaking, these funds hold shares in the five hundred companies that comprise the S&P 500 index. These funds offer no discretionary management, and so their only purpose is to make exposure to these five hundred stocks easy and convenient. This is only possible because of the separation of funds and managers.
If an S&P 500 index fund did not separate its investment assets from its management assets, the fund would have to employ its managers, secretaries, and other workers directly. It would also have to own its operational assets, such as computers and equipment, and sign operational contracts, such as leases for office space and employee health insurance. As a consequence, the fund’s value would reflect not just the fluctuations in the value of the S&P 500, but also fluctuations in the markets for office space, secretarial labor, health insurance, and so on. In the end, the risks associated with these fluctuations would amount to a kind of new stock holding for the fund, creating, in effect, an “S&P 501” fund.
The separation of funds and managers solves these problems by giving operational expenses to the management company and making the management company’s shareholders the residual risk-bearers for these expenses. A typical S&P 500 index fund, for example, signs a contract with its management company under which the fund receives its operational services in return for a contractually specified fee, which generally will not fluctuate with the management company’s actual operational expenses.
The separation of funds and managers further tailors fund investors’ risks by insulating funds from the management company’s creditors. If a management company becomes insolvent, its creditors have no claims against the funds. This suggests that separation may have important risk-tailoring functions even when the residual risks of operating expenses are small in comparison to investment risks. Although fluctuations in the cost of secretarial wages might not have mattered to investors in Lehman Brothers’ hedge funds, the risks created by Lehman’s bankruptcy surely did. Because advisory businesses are often part of much larger businesses, the insulation that the separation of funds and managers provides from managers’ creditors is crucially important.
This last observation leads us to contemplate a final motivation for the separation of funds and managers: it facilitates the simultaneous or serial management of multiple funds. As noted above, it is quite common for management companies to operate multiple funds simultaneously or serially over time. And many management companies also operate other lines of business, such as investment banking, commercial banking, brokerage, and so on. The motivation, presumably, is that by operating multiple funds and other lines of business, management companies can achieve economies of scope and scale.73 These economies ought to benefit investors as well as managers, because managers who achieve these economies can use the savings to compete for investors by promising lower fees and better returns.
Returning to the product analogy, we might say that the separation of funds and managers enables fund management companies to operate multiple funds in much the same way that product manufacturers operate multiple product lines.
The separation of funds and managers facilitates the operation of multiple funds and other lines of business in three ways: first, it prevents liabilities and residual earnings risks from spilling between one fund or business line and another; second, it allows management companies to continue as going concerns even when their individual funds have liquidated; and third, it enables the efficient resolution of conflicts of interest among multiple funds by giving management companies the authority and incentives to resolve these conflicts efficiently.
First, the separation of funds and managers limits the spillage of risks across funds and between funds and other business lines. It does so by preventing creditors of one fund or business line from seizing other funds’ assets and by insulating each fund’s investors from fluctuations in the profitability of the other funds and business lines.
In essence, the simultaneous management of multiple funds simply strengthens the argument about risk-tailoring above: as a management company operates a greater number of funds and other business lines, the potential risks to investors become larger and less correlated with the investors’ desired investment risks. The need for distinctions between the risks of any one fund and the risks of the management business becomes more urgent.
Management risks can be substantial. When several financial conglomerates, such as Lehman Brothers and Washington Mutual, collapsed during the financial crisis of 2008, only the separation of funds and managers prevented creditors of these conglomerates’ far-flung businesses—including their mortgage trading businesses—from seizing assets in the funds these conglomerates managed. Less dramatically, the risks associated with residual earnings—as distinct from liabilities—can also be important. Goldman Sachs’s profits, for example, depend primarily on the success of far-flung and diverse business lines that are almost entirely uncorrelated with the success of any particular Goldman mutual fund. Investors in Goldman’s mutual funds therefore naturally demand insulation from these risks.
The separation of funds and managers also facilitates the operation of multiple funds by allowing management companies to continue as going concerns even after their individual funds have liquidated. This is especially important for private equity funds, which liquidate every few years.
If a fund’s assets and its managers’ assets were combined into a single entity, the managers would have a very hard time operating a new fund after they had liquidated their first fund. The managers could not dissolve the entity that held the fund’s assets—and thus terminate its liabilities—without also dissolving the entity that held the management company’s assets. If the managers dissolved such an entity, they would lose the contracts and assets attached to that entity, such as office leases, employment agreements, and corporate logos. The cost of reassembling all of these contracts and assets could potentially become very high, particularly for large repeat players such as KKR, Bain Capital, and Goldman Sachs.
The separation of funds and managers also enables the operation of multiple funds by addressing the conflicts of interest that inevitably arise among the various funds. Because managers owe fiduciary duties to each of their funds and because resources are scarce, the allocation of virtually any resource to one fund places the manager in a conflict of interest with its other funds. These conflicts arise in innumerable aspects of management companies’ operations, such as the allocation of investment opportunities; the allocation of workers’ time; the allocation of administrative resources; the order in which trades are executed; the way shares are voted in portfolio companies when different funds hold conflicting interests in the companies;74 and the speed and order in which redemptions are processed, as well as myriad other matters.
These conflicts resemble the conflicts of interest that have helped to make tracking stock unworkable in ordinary companies. Tracking stock fails not simply because it cannot partition assets, but also because it requires managers and directors to be loyal to different groups of stockholders.75
Financial economists and legal scholars have thus found the conflicts that arise from the simultaneous management of multiple funds in investment management companies extremely alarming. Indeed, we are witnessing a boom in academic articles that catalog new sources of inter-fund conflicts for fund managers, particularly with respect to mutual funds.76 The SEC has occasionally offered ad hoc solutions to some of these problems.77
This ad hoc approach misses the forest for the trees. Existing research seems to treat conflicts among funds as though they were rare and problematic, much as conflicts are among shareholders in ordinary companies. In fact, however, conflicts among funds in investment management companies are completely pervasive. They are far more common than in ordinary companies and indeed far more common than anyone seems prepared to acknowledge. Rather than trying to catalog and address each of these conflicts, therefore, we should instead ask a deeper question about why fund investors so freely permit these conflicts in the first place.
The answer to this deeper question may be found in three factors associated with the separation of funds and managers: (1) the exit rights discussed above; (2) the limits that the separation of funds and managers places on fund investors’ control over management companies and their conflicted resources; and (3) the limits that the separation of funds and managers places on fund investors’ rights to claim managers’ residual earnings from the allocation of conflicted resources.
First, exit rights are important because they ensure, at least in ordinary situations, that no fund will be substantially worse off as a result of conflicts of interest among a manager’s various funds. In other words, exit rights increase the odds that conflict resolution schemes will be Pareto optimal. If shareholders of a fund anticipate that a conflict resolution scheme will make them worse off than they would be under some alternative scheme, then they can exit and invest in a different fund that offers such a scheme, or else they can invest in a fund whose managers operate only a single fund and face no inter-fund conflicts at all.
To be sure, exit rights do not provide investors much direct help with respect to past conflicts; they only provide direct help with respect to future conflicts. Once a fund has lost some resource, redemption or liquidation rights will not bring it back. But exit rights are nevertheless useful in avoiding future conflicts, because exit rights reduce the difficulty of foreseeing future conflicts and the costs of failing to foresee them. In ordinary companies, it is not possible for shareholders to foresee and price conflicts over the life of their investments, because their assets are locked in with no clear point of exit in the future.78 In investment funds, by contrast, equity holders only have to foresee and price conflicts up through the next exit date. Mutual fund shareholders, for example, only have to foresee conflicts up through the end of the business day, because mutual funds offer redemption daily.
Indeed, exit may be helpful even when investors do not actually foresee and exit in response to future unfavorable conflict resolutions, because exit enhances the power of reputational penalties. Redemption rights and periodic liquidations force fund managers constantly to seek new investors. Managers must therefore constantly consider how conflict resolutions will affect their ability to attract new investors.
The second important feature that enables efficient conflict resolutions is the set of limits on fund investors’ control. These restrictions prevent fund investors from fighting with each other over the resolution of conflicts. If investors in a management company’s various funds could all simultaneously exercise residual control over the management company’s resources, then each fund would try to use that control to allocate the management company’s resources to itself. Conflicts would always be resolved in favor of the funds that had the most voting power, rather than the funds that valued the resources most highly.79
Lastly, limits on fund investors’ claims over management companies’ residual earnings are also important. They align control rights over conflicted assets with incentives to exercise those control rights efficiently. The people who control management companies and their conflicted assets are, of course, the shareholders of the management companies. Because these shareholders can also keep these management companies’ profits, these shareholders will generally choose conflict resolution schemes that maximize the aggregate value of the management companies. When fund investors have robust exit rights and full information, management companies will choose the conflict resolution schemes that are most efficient across the funds and the management companies, because these schemes will attract the most investors and entice them to pay the highest fees. Managers thus act as central planners across all of the funds they operate, allocating resources to their most efficient uses, or at least adopting conflict resolution schemes that create the greatest aggregate value across all of the funds and the management company.
Real-world conflict resolution schemes in private equity and hedge funds are consistent with this simple conceptual model. In particular, the differing approaches of private equity and hedge funds demonstrate the central role of exit.80 In hedge funds, exit is relatively easy, and so managers have almost unlimited discretion to allocate resources among funds and to resolve conflicts as they arise. Under the terms of hedge fund operating agreements, managers are expressly freed from any restrictions on their ability to appropriate opportunities for themselves and their other funds and are not obliged to devote any particular resources to any particular fund or any amount of time beyond what they deem necessary in their sole discretion.81
By contrast, in private equity funds, exit is much more restricted, and therefore so is management companies’ freedom to take on and resolve conflicts among multiple funds.82 Private equity managers generally cannot operate multiple funds with similar investment objectives simultaneously (although they may do so serially), and they are required to devote specific resources, including substantially all of designated employees’ time, to each fund.83 Managers are also extremely tightly regulated in their ability to appropriate opportunities for themselves. This pattern thus suggests that fund management companies maximize the value of their funds and management businesses by choosing conflict resolution schemes that are efficiently tailored to fund investors’ levels of vulnerability.
For a stylized illustration of how conflict resolution works inside of fund management companies, let us return once again to the product analogy. Much like fund management companies, product manufacturers simultaneously operate many different product lines. And much like fund management companies, product manufacturers resolve these conflicts among the customers of these product lines to create the greatest possible aggregate value for those customers.
Consider PepsiCo. It manufactures dozens of different snack products, ranging from Pepsi-Cola and Mountain Dew to Doritos and Fritos to Quaker Oatmeal. Conflicts of interest among the buyers of these different products are ubiquitous. For example, if PepsiCo gets a limited supply of particularly cheap or particularly high-quality corn, PepsiCo must decide which of its various product lines will receive the corn. The boost in quality or drop in price that would come from the corn could potentially benefit the buyers of many different product lines. PepsiCo is thus similar to a fund management company, except that in PepsiCo, conflicts relate to corn, rather than to investment opportunities and administrative resources.
Because customers’ exit rights are highly robust in snack product markets—customers can just buy from other manufacturers—PepsiCo operates under a conflict resolution scheme very similar to that of hedge funds. PepsiCo has few if any contractual commitments to its customers that limit its discretion to resolve conflicts among the various product lines as they arise. PepsiCo’s managers thus exercise broad discretion to resolve conflicts in a manner that will attract the most customers at the highest prices and thus generate the greatest profits for PepsiCo’s shareholders.
For example, if PepsiCo resolves conflicts over the high-quality/low-price corn in favor of, say, Doritos, rather than Pepsi-Cola, the demand for cola may be lower than it would have been if the corn had been used to make cola. Nevertheless, PepsiCo will use the corn for Doritos if it expects that Doritos buyers will value the corn highly enough that the increase in Doritos demand will offset the forgone increase in cola demand.
People who buy cola will not be substantially worse off as a result of PepsiCo’s decision to favor Doritos, however. If the cola becomes sufficiently bad or sufficiently expensive, then the cola buyers will simply switch to Coke.
In fact, the conflicts PepsiCo faces may actually become a source of efficiency and competitive advantage. PepsiCo may be able to offer better quality and lower prices on all of its products than competitors who manufacture only a single product are able to do. PepsiCo might be able to do this because it can achieve various economies of scope and scale through the simultaneous operation of many product lines. For example, it can spread fixed costs among its various product lines and can buy in large quantities. Additionally, it might be able to move resources to their most efficient uses more cheaply than the commodities markets can.
Indeed, this last insight suggests what may be the most satisfying aspect of the analogy between investment funds and product manufacturers: my account of conflict resolution in fund management companies is really just a version of the internal capital markets model of the firm.84 The internal capital markets model suggests that one of the advantages of firm organization is that capital can sometimes move more freely within firms than between firms. The internal capital markets model may be applied to fund management companies if we simply think of administrative resources and investment opportunities as capital to be allocated and if we think of funds as projects or product lines calling for capital.
One objection to this model is that because fund managers sometimes invest in some of their own funds, the system of conflict resolution I have just described might become corrupted. One might worry that if a fund manager invests in one of its funds and not in the others, the manager will unduly favor the fund in which the manager has an interest. This is not a problem, however. The temptation to favor funds in which a manager invests is conceptually no different from the temptation to engage in any number of other forms of self-dealing. The forces that restrain managers from favoring the funds in which they invest are thus no different from the forces that restrain other forms of self-dealing: managers can generally make more money by using resources to attract fund investors than by appropriating the resources for themselves. And investors can simply invest elsewhere.
Again, the product analogy illustrates. If PepsiCo simply sold its corn for cash and distributed the proceeds to its shareholders, this would, technically speaking, constitute an outrageous conflict of interest. But of course, no one would find this kind of action at all troubling, and indeed no one would even conceptualize it as a conflict of interest. The reason is that if PepsiCo’s outside customers were unhappy, they could simply buy Coke.
One might also argue that in many investment funds—particularly private equity funds—exit is not as easy as it is in snack product markets. This is true. But the response here mirrors the response given above to concerns about the strength of exit85: the point here is not to say that exit is perfect. Rather, the point is to build a conceptual model to understand exit’s consequences and to identify its central role in making conflicts of interest work. The model explains, for example, why private equity funds regulate inter-fund conflicts so much more strongly than hedge funds: it is because of private equity funds’ limits on exit.
To be clear, in building this explanation for the separation of funds and managers, I am assuming, as most of the law and economics literature does, that entrepreneurs begin with an idea for an enterprise and then choose the patterns of ownership and asset partitioning that will structure it most efficiently. I am arguing, in other words, that the factors I have identified cause the separation of funds and managers, and not the other way around.
Note also that one might locate causation somewhat differently than I have by reaching even deeper into the causal chain to attribute the motivation for the separation of funds and managers to the nature of funds’ assets, rather than the nature of their investors’ rights. One might argue that the factors I have identified—exit rights and performance compensation, tastes for precise risk-tailoring, and economies of scope and scale—are themselves possible and desirable only because of the liquid character of investment funds’ assets. One might argue that hedge funds can only offer exit rights, for example, because their assets can be sold easily enough to generate cash for exiting investors.
This view is certainly correct, and it explains why the separation of funds and managers tends to be so closely associated with securities investing. I address this association in more detail below.86 I have chosen to focus on a slightly higher set of links in the causal chain, however, because that approach seems both more precise and more accurate. The lower-level fact that some funds hold liquid assets, for example, does not help to explain the separation of funds and managers until we understand the intermediate fact that liquid assets enable exit rights.
I now examine a few special cases of the separation of funds and managers and some of the key objections to the framework I have proposed for explaining it.
Closed-end funds present perhaps the most obvious special case of the separation of funds and managers. Like other investment funds, closed-end funds separate funds and managers. But unlike other investment funds, closed-end funds do not offer any special exit rights. They do not offer redemption and do not periodically liquidate. On first impression, therefore, closed-end funds might seem to prove that exit is not an important motivator for the separation of funds and managers.
In fact, however, closed-end funds show precisely the opposite. Investors apparently find closed-end funds’ limited exit rights so troubling that very few investors are willing to pay full price for closed-end shares anymore. Closed-end funds now comprise only about one-and-a-half percent of the assets managed by the types of funds discussed in this paper.87 And those funds that do exist tend to trade at discounts to their NAVs, suggesting that investors do not value them as highly as open-end funds, whose shares always sell for exactly NAV (plus the costs of any initial fees).
Perhaps one reason (though perhaps not the only reason88)why closed-end funds trade at these discounts is that the absence of exit makes the separation of funds and managers in these funds seriously problematic. These funds use the separation of funds and managers to limit investors’ control, but they do not compensate by offering investors exit rights.
The separation of funds and managers limits fund investors’ control in closed-end funds in two key ways. First, unlike boards in ordinary companies, boards in closed-end funds cannot back up their control by threatening to hire and fire individual senior executives. Closed-end fund boards can only hire and fire entire management companies.89 This is costly and difficult, since firing an entire management company entails resetting all of a fund’s administrative operations and business strategy. Additionally, even when a board decides that firing an entire management company is worth these costs, the board must overcome large legal and practical obstacles. Under the ICA, firing a management company and replacing it with a new one requires a majority shareholder vote.90 Boards in ordinary companies, by contrast, can hire and fire CEOs without ever asking the shareholders. The net result of these difficulties in closed-end funds is that although these funds fire their managers more often than open-end funds do, they fire their managers far less often than ordinary companies do.91
Second, because shareholders in closed-end funds do not exercise residual control over management and operational assets, managers can retain rights over those assets that they typically could not retain in ordinary companies. Most important, closed-end fund managers retain the right to devote management resources to many different funds simultaneously, and thus to assume myriad conflicts of interest that would not be permitted to the managers of ordinary companies.
To be clear, closed-end funds are not necessarily bad investments when they trade at discounts to NAV. At the time of their initial public offerings, closed-end funds must sell at slight premiums to NAV. Sophisticated investors generally avoid buying shares in these offerings, and this is why closed-end funds are so rare.92 But once closed-end funds have successfully passed the IPO gauntlet, they typically begin trading at discounts to NAV. Once this occurs, buying shares in a closed-end fund can be quite sensible. Even funds with serious agency conflicts can be good investments if the price is low enough. Further, the closed-end form offers benefits that at least partially offset its shortcomings. Most important, because closed-end funds do not have to pay redemptions, they do not have to worry as much about liquidity. They can invest more of their cash and they can place it in higher-return, lower-liquidity assets.
Asset securitization vehicles present another special case. These vehicles are not commonly regarded as investment funds, but they nevertheless bear a striking resemblance to investment funds. Most asset securitization vehicles hold pools of receivables on financial obligations such as mortgages, auto loans, and credit card debt.93 Asset securitization vehicles buy these receivables from the banks and other financial institutions that originate them. These vehicles may be said to separate funds and managers because they have no employees and receive all of their administration from “servicers” with distinct legal existences and distinct owners. Asset securitization vehicles’ organizing documents also give investors almost no control over the vehicles’ administration.
The motivations for the separation of funds and managers in asset securitization vehicles parallel those in investment funds. Much as investment funds use the separation of funds and managers to insulate fund investors from managers’ liabilities and residual earnings, asset securitization vehicles use this pattern to insulate the securitized assets from originators’ and servicers’ liabilities and residual earnings. This has been well documented in the literature on asset securitization.94
Less well documented are the reasons why securitized asset investors accept the limits on control that grow out of this pattern of organization. Whereas investors in investment funds accept limits on control because they hold exit rights, investors in asset securitization vehicles may accept these limits for at least two different types of reasons. First, interests in asset securitization vehicles are almost all structured as debt, rather than equity. Investors in these vehicles thus accept limits on control because they receive fixed entitlements to take the place of control. Second, asset securitization vehicles hold only relatively fixed pools of assets with very predictable cash flows. Servicers are thus ideally supposed to act only as functionaries, and are not expected to make discretionary decisions that might necessitate the exercise of oversight or control by investors.
Another special case is the donative trust. In the prototypical modern donative trust, a person who owns assets (a “settlor”) hires a financial institution (a “trustee”) to manage the assets on behalf of a third person (a “beneficiary”) to whom the settlor wants to give the assets as a gift. Modern donative trusts may be said to separate funds and managers, because the trusts themselves prototypically have no employees or operational assets, and because trust beneficiaries generally do not own the trust institutions that manage their assets.95
Legal scholars have long been aware that trust law insulates trusts’ assets from trustees’ creditors, and that this is one reason for use of the trust form.96 Of course, use of the trust form for this purpose closely parallels investment funds’ use of the separation of funds and managers to insulate funds’ assets from management companies’ creditors.
Additionally, trusts are often employed to limit beneficiaries’ control, much as the separation of funds and managers is used to limit fund investors’ control. But in trusts the motivations for restricting control are different. In investment funds and asset securitization vehicles, restrictions on control appear because investors have alternatives to control, such as exit. In donative trusts, by contrast, restrictions on control often appear as a primary goal, imposed for their own sake. Settlors often restrict control to protectbeneficiaries from themselves or others.97A father might give assets in trust for the benefit of his son, for example, and instruct the trustee to pay out the assets only if the son fulfills some condition intended for his benefit, such as finishing college or marrying a Jewish girl.98
The creation of distinct ownership for trust assets and trust institutions helps to make these restrictions on control possible. The reason beneficiaries cannot control trust assets is that beneficiaries do not own the trust institutions that manage them.
The unique motivations for restricting control in donative trusts can create uniquely fascinating and difficult problems that do not appear in other enterprises that separate funds and managers.99 Because control restrictions appear for their own sake, rather as a response to the presence of alternatives to control, trust beneficiaries often find themselves in the precarious position of having neither control nor substitutes for control. Trust beneficiaries must therefore rely much more heavily on fiduciary law and contractual protections than investors in investment funds do.100
Let us now consider a few objections to the explanation for the separation of funds and managers offered in Part III.
One possible objection is that the separation of funds and managers is actually driven by some quirk of tax law. I know of no sustained attempt to explain the separation of funds and managers by reference to tax law, but surely some readers will intuit that the separation of funds and managers is really just an attempt to avoid entity-level taxation, for example, or to take advantage of the “carried-interest” device that allows private fund managers to pay capital gains rates on their fee income (and which was famously associated with Mitt Romney’s time at Bain Capital).101 Tax-based explanations are insufficient, however. Although taxes might explain a few details of funds’ organization,102 they cannot account for the basic fact of the separation of funds and managers.
The main reason is that the separation of funds and managers has appeared across too wide an array of times and tax regimes for it to be a product of tax law alone. When the separation of funds and managers first appeared in mutual funds and closed-end funds in the 1920s, these funds were taxed at the entity-level as corporations.103 The pattern of separation then persisted even after Congress dramatically altered these funds’ taxation by adopting a kind of quasi-pass-through taxation system in the 1930s and 1940s that allowed ICA-regulated funds to avoid most entity-level taxation by distributing their income every year.104 The pattern now also persists in private equity and hedge funds that are taxed as true partnerships under an entirely different system from modern mutual funds and closed-end funds.
In particular, the “carried interest” device fails as an explanation, because the separation of funds and managers predated the carried interest device by several decades. And although the carried interest device is popular among private equity and hedge funds, it remains completely unavailable to the managers of mutual funds and closed-end funds.105
It might also seem that the separation of funds and managers results from some quirk of the ICA or the Investment Advisers Act (IAA). This explanation fails, however, for the same basic reason that the tax-based explanation fails: the separation of funds and managers has appeared across too wide an array of regulatory systems for regulation to explain it. When mutual funds and closed-end funds first began separating funds and managers in the 1920s, they faced essentially no special regulation at all. And modern private equity and hedge funds also separate funds and managers even though many of them were subject to no special regulation under either the ICA or IAA until after Dodd-Frank.106 The separation of funds and managers is also ubiquitous among funds in Europe, despite differences in regulation between Europe and the United States.107 Furthermore, although American fund regulation presupposes the separation of funds and managers (as I explain below), nowhere does American fund regulation expressly require this pattern. This is why a handful of mutual funds were able to combine ownership of their funds and managers and still comply with the ICA for decades before they finally separated funds and managers.108
Another theory is that the separation of funds and managers arises from the tendency of many investment funds—particularly private equity and hedge funds—to organize as limited partnerships.109 This theory says that the reason private equity and hedge fund investors have few control rights is that in limited partnerships, anyone who exercises control becomes a general partner and has to give up limited liability. Fund investors may not like the limits on control, the argument goes, but they accept these limits as the price they must pay for limited liability.
This theory is incorrect, because if shareholders and managers did not like limits on control, then they would simply avoid these limits by organizing their funds as LLCs or statutory business trusts. These forms give limited liability to all investors without any restrictions on control. And they still achieve the same pass-through tax treatment as limited partnerships.110 In fact, many private equity and hedge funds actually use the LLC and business trust forms.111 Notably, when funds adopt these forms, they deliberately write their operating agreements or trust instruments to include draconian restrictions on fund investors’ control rights that resemble the restrictions that appear by law in limited partnerships.112 This pattern suggests that limits on control are desirable for their own sake, and not merely as instruments for achieving limited liability.
Other readers might argue that the success of the Vanguard mutual fund management company, which is now one of the nation’s largest,113 casts doubt on the value of separating funds and managers. Vanguard has a unique ownership structure that might appear at first not to involve the separation of funds and managers.
The Vanguard management company is a Pennsylvania corporation and each of the Vanguard funds is a distinct legal entity. Vanguard is unique, because the management company has issued its common stock to the funds that the management company serves, rather than to investors. In other words, Vanguard is technically owned by its funds. Vanguard thus claims—with the widespread agreement of academics and the public—to be a kind of client-owned co-op or mutual company.114
On first impression, therefore, Vanguard might appear to do away with one of the two key features that define the separation of funds and managers. Although Vanguard clearly uses separate entities to hold investment assets and management assets, it does not exactly construct separate ownership for each of those entities. Vanguard nominally combines the ownership of the management company and the funds by placing the management company’s ownership in the hands of the funds.
Vanguard is worth thinking about because, as noted above, Vanguard’s founder, Jack Bogle, is a vigorous critic of the separation of funds and managers.115 And Vanguard itself has built its branding partly on the notion that its corporate structure uniquely aligns its interests with those of fund investors.
In reality, however, Vanguard is not meaningfully different from any other mutual fund management company. Practical circumstances and contractual devices deprive Vanguard’s fund investors of any meaningful residual control and residual earnings rights. In economic reality, therefore, Vanguard investors are not truly the “owners” of the management company any more than the fund investors in any other mutual fund complex are.
Vanguard fund investors’ right to residual control over the management company is meaningless because fund investors have redemption rights. Fund investors will never actually use their rights to vote in the management company, because they will almost always prefer instead just to redeem. In other words, fund investors will never exercise control rights over the management company for the same reasons they will never exercise control rights over the funds.116
Vanguard’s fund investors also have no meaningful right to claim the management company’s residual earnings. At the time the management company starts its funds, the management company designates the fund’s board. The board then causes each fund to sign agreements drafted by the management company that expressly give the management company essentially unlimited discretion to allocate costs among the various funds.117 The board of the management company can allocate costs without regard to the number of shares each fund owns in the management company and without regard to the amount of costs each fund actually generates.
As a practical matter, this system deprives the funds and their investors of any real entitlement to Vanguard’s profits. Vanguard can arbitrarily determine the financial benefit any particular fund receives from the management company by simply manipulating the portion of costs the fund must pay and the portion of resources it receives. Vanguard rarely distributes profits—it claims that its costs exactly equal its revenues—but even if Vanguard did distribute profits, it could offset or enhance any particular fund’s dividends by simultaneously allocating the fund an increased or reduced share of costs.118
This makes Vanguard very different from true customer-owned co-ops and mutuals, such as agricultural and grocery supply co-ops. In these organizations, dividends must be paid out at regular intervals according to strict formulas based on share ownership and patronage.
It is thus clear that Vanguard is not truly owned by its customers. So then who are its real owners? That is not so clear. One possibility is that Vanguard might not have any owners. It may be a kind of autonomous commercial nonprofit, sort of like a hospital. To be precise, Vanguard is not actually a nonprofit under the tax code or state law. But it purports to operate its funds “at cost” and so may functionally approximate a nonprofit. Alternatively, perhaps Vanguard is functionally owned by its employees and senior executives. Vanguard’s employees and senior executives apparently exercise a high degree of influence inside of Vanguard, and it is possible that Vanguard’s profits are being paid to its senior executives in the form of large salaries and other perks, even though these executives lack formal residual earnings rights. More formally, Vanguard’s employees participate in a profit-sharing program that actually grants them residual earnings rights.119
To be clear, the point of this discussion is not to say that there is some scandal in Vanguard’s structure. To the contrary: Vanguard’s fund investors actually benefit from Vanguard’s de facto separation of funds and managers. Vanguard’s fund investors would be much worse off if they truly did own Vanguard. If the fund investors exercised meaningful control rights over the management company, then investors in the various funds would use their control rights to tear Vanguard apart by fighting over conflicted resources. And if the fund investors had real exposure to Vanguard’s profits, then they would be forced to bear the risks of those profits. Thus, the lesson of Vanguard is that the separation of funds and managers is so useful that not even Vanguard can live without it.
This positive explanation for the separation of funds and managers has far-reaching normative implications. It can tell us about the overall efficiency of the separation of funds and managers, the basic purposes of investment fund regulation, and several ways in which we might improve fund regulation.
First, we can now see clearly that the separation of funds and managers is generally a good thing. This insight may seem trivial, but it may actually be this Article’s most important contribution. As noted above, a distressing number of commentators and policymakers, particularly in the world of mutual funds, believe that the separation of funds and managers is some sort of scandal. That is why Vanguard has so aggressively touted the benefits of its ostensibly “mutual” ownership structure. The theory of this Article suggests that we can set these concerns aside.
Additionally, we can now discern one of the ICA’s core functions. The almost universally accepted wisdom about the ICA is that its primary function is to regulate a particular kind of asset—i.e., securities. I would argue, however, that much of the ICA’s function is also to regulate a particular kind of agency relationship or organizational structure that tends to be associated with securities investing—i.e., the separation of funds and managers.
To be clear, I express no opinion for present purposes about whether the ICA is on balance good or bad. The point is just that, whether one likes the ICA or not, one can only make descriptive sense of it by acknowledging that one of its central functions is to address the separation of funds and managers. The ICA is not explicit about its focus on the separation of funds and managers, but the focus nevertheless appears implicitly throughout.
To see this, try a thought experiment: imagine how the ICA would operate on a fund that invested in securities, but did not separate funds and managers. We might ask whether in such a scenario the ICA’s peculiar requirements would seem sensible or not. The answer is that they would be absurd.
Consider sections 10 and 17 of the ICA. These sections create an elaborate (but ultimately not very restrictive) set of rules to address conflicts between a manager’s various funds and between a manager’s funds and its other lines of business.120 Practitioners and scholars of mutual fund regulation tend to regard these sections as the heart of the ICA.121 These sections may or may not make sense for funds that separate funds and managers, but for funds that do not separate funds and managers, these provisions are pointless. Without the separation of funds and managers, there would no point in regulating conflicts among a manager’s multiple funds, because the management of multiple funds would be impossible.
Similarly, the ICA contains several provisions that maintain the integrity of asset partitioning between funds and management companies. For example, the ICA requires fund assets to be deposited with third-party custodians unaffiliated with either the funds or the managers.122 If funds merely invested in securities without also separating funds and managers, these provisions would make little sense. The point of these provisions is to address the fact that investment funds are dominated by external entities with their own distinct assets.
The ICA also requires contracts between funds and their managers to be written down, rather than merely agreed upon orally.123 This requirement implicitly assumes that a fund will have a single external investment adviser with whom a contract can easily be reduced to writing, which is only the case when a fund separates funds and managers.
To be clear, the ICA’s implicit focus on the separation of funds and managers was not the result of an explicitly articulated legislative effort. The historical actors who drafted the ICA were no more explicitly aware of the central importance of the separation of funds and managers than modern observers are. Nevertheless, like modern observers, the ICA’s drafters had a deep implicit awareness of the separation of funds and managers and the problems and opportunities it creates. The ICA’s drafters were career bureaucrats and industry actors who lived the experience of this pattern every day.124 It is thus no surprise that as they attempted to address investment funds’ unique problems, they produced a system that—as a purely practical matter—focuses heavily on the separation of funds and managers.
Nevertheless, the ICA drafters’ failure to understand the separation of funds and managers explicitly has left serious shortcomings in the statute. The most important is the statute’s definition of an “investment company,” which is the ICA’s term for an investment fund. This definition is important, because it determines which companies are subject to the ICA. Note also that the way we define the term “investment fund” is important in the study of economics as well as in law, because it determines which enterprises we conceptually treat as “investment funds” for purposes of comparison and study.
The current definition, widely accepted in both regulatory and scholarly circles, reflects the view that the essence of investment funds and their regulation lies in the peculiar nature of the funds’ assets. The ICA definition boils down to the notion that an investment fund is a company whose assets consist mostly of securities. More precisely, the ICA defines an investment fund as a company in which the ratio of securities to other assets is very high.125
The problem with this definition is that it conforms neither to most people’s intuition nor to the purposes of fund regulation. It simply includes too many companies that do not separate funds and managers.126
For example, most people’s intuition would tell them that a conglomerate, such as GE or PepsiCo, is not an investment fund. Under a purely assets-based definition, however, conglomerates are clearly investment funds, because their assets consist mostly of securities in their operating subsidiaries. One might reply that what differentiates conglomerates from investment funds is that conglomerates hold controlling majority stakes in their subsidiaries, while investment funds hold only non-controlling minority stakes. This is essentially the ICA’s solution to the conglomerate problem.127 But then what about private equity funds? By definition, buyout-style private equity funds hold controlling majority stakes in their portfolio companies, just as conglomerates do. And why exactly is the difference between control stakes and minority stakes so important anyway?
The assets-based definition also has a tendency to treat early-stage operating companies as investment funds, particularly in high-tech fields. Consider a former client of mine. The client was an early-stage developer of pharmaceuticals. It raised a large amount of cash through an equity offering to fund a long-term drug development project. Since the company intended to spend the cash slowly over several years on a long-term R&D program, the company parked the cash in corporate bonds until it was needed. But then, because the bonds were the company’s only assets with any accounting value, the company’s balance sheet consisted almost entirely of securities. From the perspective of the assets-based definition, the company was indistinguishable from a closed-end bond fund.
Microsoft—the grande dame of tech companies—faced this problem early in its life. Had the SEC not granted Microsoft a special ad hoc exemption by letter ruling, Microsoft would have been regulated as a closed-end fund.128 Indeed, the high-tech company problem is so serious that in 2003, the SEC adopted a special rule expressly exempting companies with substantial research and development expenditures.129
This “R&D” rule works as a purely practical matter, but it reflects no coherent principles. It is simply a patch. Indeed, the R&D company is symptomatic of a broader problem: the assets-based definition produces so many obviously wrong results that the only way to save it has been to cover it in layer upon layer of unprincipled exemptions, patches, and fixes. The definition of an investment fund in section 3 of the ICA stretches to more than three thousand words130 and its administrative lore has grown correspondingly gargantuan.131
The way out of this maze is to start defining investment funds by their structure as well as by their assets. Specifically, I propose that an investment fund should be defined as an entity that both (a) satisfies the assets-based definition, and (b) separates funds and managers. I call this approach the “structure-based definition.”
The structure-based definition offers important advantages. First, it would align the ICA’s application with its functions. As argued above, one of the ICA’s primary functions is to regulate the separation of funds and managers. It thus makes little sense to apply the ICA to companies that invest in securities but do not separate funds and managers. To issuers that do not separate funds and managers, the ICA can look baffling and incoherent.
The structure-based definition could also improve economic analysis by helping commentators to avoid confusion. For example, Ronald Gilson and Reinier Kraakman have compared Swedish closed-end funds to American closed-end funds and have puzzled over why the Swedish funds tend not to trade at discounts to NAV and tend to be more successful than their American counterparts.132 The comparison and the question are quite reasonable if we think that the essence of an investment fund lies in the nature of its assets, because American and Swedish closed-end funds tend to hold similar types of assets. However, if we think that the essence of an investment fund lies in the nature of its organization, the Swedish and American funds appear radically different and the puzzle is easily solved. Under my definition, the Swedish funds are not really “closed-end funds” at all, because they do not separate funds and managers. They employ their managers and own their operational assets directly, just like ordinary companies do.133 This fact has previously gone unnoticed because the separation of funds and managers has been underappreciated. But this fact provides perhaps the best explanation for the Swedish businesses’ ability to avoid NAV discounts. By holding their operational assets directly, investors in the Swedish businesses avoid the debilitating control restrictions that result from the separation of funds and managers and plague investors in American closed-end funds.
Perhaps the reason the assets-based definition has survived so long in spite of its obvious flaws is that, as noted above,134 investments in securities and the separation of funds and managers are powerfully correlated. This is probably because firms that hold a lot of securities often exhibit the traits that motivate the separation of funds and managers: these firms have enough liquidity to enable exit rights and clear performance evaluation; they tend to attract people who seek precise risk exposures; and they are amenable to economies of scope that can only be achieved by operating multiple vehicles with separate owners.
My proposed structure-based definition would nevertheless offer greater clarity and elegance than the assets-based definition, and it would ensure that the ICA applies only to the enterprises that it is well suited to regulate. The outcomes of the assets-based and structure-based definitions overlap substantially, but not completely, and so the assets-based definition can sometimes cause confusion about what precisely we are trying to regulate. To be sure, drafting and implementing the structure-based definition would be challenging. In particular, regulators would have to define precisely what it means to “separate funds and managers.” However, the structure-based definition would at least force us to ask the right questions.
One might wonder why the structure-based definition should contain any reference to assets at all. The “structure-based definition,” as I am proposing it, is actually a joint test of both assets and structure. But why not just structure? The answer is pragmatic: it seems unlikely that a purely structure-based definition could gain political traction. Nevertheless, the possibility of defining investment funds based solely on their structure warrants further exploration.
A further normative insight is that we should regulate open-end mutual funds differently—and more lightly—than closed-end funds. The ICA currently treats these two types of funds almost identically.135 This is particularly troubling because the ICA was drafted primarily to address problems in closed-end funds, but the amount of assets in closed-end funds is now dwarfed by the amount of assets in open-end funds.136
The ICA’s approach of treating both types of funds similarly makes perfect sense if we think that fund regulation is primarily oriented around the nature of funds’ assets, because open- and closed-end funds prototypically invest in the same types of assets. Once we understand that the fund regulation is oriented around organizational structure, however, the ICA’s approach becomes indefensible.
Open-end funds warrant lighter regulatory treatment, because investors in these funds are relatively well equipped to address most of the peculiar problems created by the separation of funds and managers. In response to inter-fund conflicts of interest and limits on control rights, for example, open-end fund investors can simply redeem. By contrast, closed-end fund shareholders’ ability to address these problems is extremely limited. They have neither meaningful control nor meaningful exit. Closed-end fund shareholders therefore require much stronger regulatory protection than open-end fund investors.
The ICA’s failure to distinguish between open- and closed-end funds probably has its roots in historical circumstances that are no longer relevant.137 Nowadays, closed-end funds comprise only a very small portion of the fund industry, but in the years leading up to the ICA’s passage in 1940, they comprised the industry’s core. They also attracted the greatest interest from Congress and the SEC in the 1930s and 1940s, because they had a far more egregious record of abuse and mismanagement than open-end funds did.138 The ICA is thus primarily designed for the closed-end fund industry and its myriad intrinsic problems and it is too restrictive for open-end funds, which comprise the bulk of the modern fund industry.
In particular, the shareholder governance structure imposed by the ICA is deeply inappropriate for open-end funds.139 The ICA mandates voting rights for shareholders in both open- and closed-end funds.140 In closed-end funds, these rights may be sensible, since closed-end fund shareholders do not have the option to exit, and therefore actually vote on occasion.141 These requirements are a useless and expensive formality in open-end funds, however, because in open-end funds, shareholders almost always prefer to redeem rather than to vote.
In arguing that open-end funds should be less regulated than closed-end funds, I do not mean to suggest that open-end funds should be free from regulation entirely. To the extent that exit rights in open-end funds are limited by practical constraints—such as investors’ lack of sophistication, information asymmetry, or taxes—we should regulate open-end funds to address these constraints. In other words, we should regulate open-end funds to address the same types of market imperfections that warrant regulation in other product markets. Clearly, however, the optimal regulatory scheme for open-end funds is much lighter overall than for closed-end funds.
This view of the separation of funds and managers suggests that the Supreme Court’s recent decision in Janus Capital Group v. First Derivative Traders142 was wrong, at least as a matter of policy. The Court held that because a mutual fund and its management company were separate from each other, the management company could not be liable to its shareholders under SEC Rule 10b-5 for false and misleading statements contained in the prospectuses issued by the fund.143 Justice Thomas’s opinion for the majority rested on the notion that the management company could not be said to have “made” the statements in the fund’s prospectus, because the management company lacked “ultimate control” over the fund.144
The foregoing analysis suggests that this outcome is rather silly, at least as a matter of economics.145 Although funds and managers are formally distinct, the distinction arises precisely in order to diminish fund investors’ control. By both practice and design, managers dominate their funds. Separation arises to partition assets and to limit control, not to give funds any real independence.
Shifting the focus of investment fund analysis from assets to structure accomplishes three things. First, it allows the construction of a descriptive framework that explains why the fund industry looks the way it does. It explains not only the basic fact of investment funds’ unusual organization, but also important details such as why private equity and hedge funds differ so radically from mutual funds in their control and conflict resolution structures and why closed-end funds persistently trade at discounts to NAV. We can now see that despite all of the skepticism that has been aimed at the fund industry’s basic pattern of organization, this pattern makes a great deal of sense.
Second, this shift in focus clarifies the basic functions of fund regulation and suggests how to improve this regulation. This shift provides insight about how to define investment funds, how to regulate open- and closed-end funds differently, how to regulate conflicts of interest between funds, and other matters.
Finally, a proper understanding of investment funds’ organization provides a unifying framework for the comparative study of different types of funds. To date, the fields of financial economics and law have produced remarkably little systematic analysis of how private equity funds, hedge funds, mutual funds, and other types of funds differ from one another and what the motivations for differences are. We can surely learn a great deal from comparison, but in order to do so, we must first understand what unites and defines these seemingly disparate enterprises and what might motivate their similarities. My hope is that the framework proposed here will help to provide that understanding.