The possible need for insurance of money market mutual funds is of particular concern given the vulnerability of those funds, in particular, prime institutional funds, to contagious runs due to the money-like nature of their liabilities. Totaling $2.6 trillion in liabilities in the second quarter of 20151 compared to bank liabilities of $13.6 trillion,2 money market funds are major buyers of short-term capital markets instruments including short-term Treasury debt, ABCP and secured repo, primarily those issued by global banks.3 However, institutional prime funds make up a minority of this total, or around $910 billion.4 In its 2010 report on money market fund reform, the President’s Working Group on Financial Markets (PWG) suggested insurance as a reform option for money market funds.5 Discussion of this option has largely laid moribund since that time due to the unclear case for its need, the complexity and cost of providing insurance, and a political environment hostile to any further support for “Wall Street.”
At the outset, however, one must ask whether there is a current need for insurance. First, runs in the crisis were focused on prime institutional funds. Total prime funds now stand at approximately $1.4 trillion6 compared to $2.18 trillion7 before the crisis—they are shrinking in value. Additionally institutional prime funds—with assets of $910 billion in the fourth quarter of 20158—are only a fraction of total prime funds. Institutional prime funds are expected to shrink even further after the SEC’s requirement that these funds have a floating NAV comes into effect. Fidelity has already announced that it will move its investors in prime institutional funds into fixed NAV government funds where there is virtually no risk of failure.9 The case could be made for insurance for retail prime funds even though they did not run in the crisis, because retail investors might run in a future one; perhaps the quick provision of the Treasury guarantee, only three days after the Reserve Primary Fund “broke the buck,” explains why retail funds did not run.10 In principle if we think there is a need for deposit insurance for retail deposits to avoid bank runs, one would think one needed the same insurance for retail money market fund investments. One might also argue that the floating NAV and authorization of redemption fees and gates will stop contagion in money market funds—but this is not well founded, as we discuss in part IV. The centrality of money market funds to modern financial intermediation and the powerful influence that investor confidence in the integrity of MMF investments exerts on the stability of financial markets was displayed during the financial crisis of 2007 to 2009, when serious runs on money market funds impaired the orderly operation of the commercial paper markets and propelled contagious knock-on runs up the chain of intermediation. Whatever risk does remain of runs in the money market funds post-crisis might be reduced by insurance. Birdthistle (2010) suggests that, alongside other reforms, MMF insurance organized privately or publicly would help to offset this risk.11 Private insurance might be impractical, however, because, owing to contagion, loss-causing events are unlikely to be small, isolated occurrences that insurers can comfortably manage. Instead, in the event that one fund breaks the buck, contagion-induced outflows might cause countless other funds to do so, leading to losses that are too large for insurers to bear.12 Given the potentially extreme losses that private insurers could face, some kind of public insurance may be a more feasible option, at least for the tail risk of large losses.
However, there are serious countervailing arguments against providing any insurance to money market funds. For example, ICI contends that the size and complexity of contemporary MMF portfolios would make a comprehensive insurance system impracticable, drive outflows from depository banking, and create moral hazard.13 In the ICI’s view, providing federal insurance to MMF investments would siphon cash from traditional bank deposits, causing “disintermediation [and] significant disruption to the banking system.”14 Capping the guarantee, as in the depository insurance context, would leave room for runs by investors with uninsured exposures in excess of the cap.15 On the other hand, exempting money market funds from explicit insurance, as they are today, may encourage investors to shift short-term funding from deposits into money market funds because those investments will continue to benefit from an implied public guarantee (as we saw in the crisis) without internalizing its costs. And clearly industry would prefer an implicit guarantee that they need not pay for since any explicit insurance fees would be passed onto investors. The 2010 PWG report noted several critical issues that would need to be addressed about the provision of insurance to MMFs.16 These issues included moral hazard concerns, the dramatic expansion of the role of regulators with limited bandwidth, and insurance pricing difficulties.17
Considering the issues regarding pure private or pure public insurance, MMF insurance could be provided by a hybrid system that combines private insurance with a public backstop. This approach, also discussed by the ICI, could entail three levels of loss absorption.18 For example, the MMF would be responsible for losses on the first 0.5 percent of fund assets, while a private insurer covers the next 2.5 percent. The public backstop would kick in after losses exceed 3 percent of fund assets, hence limiting government exposure to extremely high losses and capping potential losses to private insurers. Such a system of government support would resemble the federal backstop created by the Terrorism Risk Insurance Act (“TRIA”), which provides public reinsurance coverage to insurance companies facing claims related to declared acts of terrorism. While private insurance companies may have concern about heightened federal regulation (insurance companies are primarily state regulated) under such a hybrid system, the TRIA model suggests additional regulatory burdens may not be a necessary component. TRIA explicitly preserves the jurisdiction and regulatory authority of the States with only minor exceptions.19 While an original version of the bill in the House of Representatives included a provision to impose increased capital requirements through tax deductions, ultimately this provision was removed from the final TRIA legislation.20 However, despite the potential for a hybrid system modeled after TRIA, insurance companies may not have the capacity or desire to allocate sufficient capital for the private portion, even with the capped exposure.21
Another strand of criticism urges that insurance of money market funds would confront traditional depository banking institutions with burdensome competition. Since MMF portfolios contain generally high-quality, liquid, readily marketable securities, insurance premiums charged to MMF institutions would presumably be lower than the rates applied to conventional banks, which often transform deposits into longer term, illiquid, and thus riskier loans. Insured MMF instruments would then pass through a portion of this cost advantage to investors in the form of higher yields relative to traditional deposits, encouraging customers to migrate out from depository banking institutions into lower cost (but equally secure) MMF shares. This could decrease bank lending in favor of more direct finance through the markets in which the money market funds invest. Indeed the PWG notes, “Limits on insurance coverage (perhaps similar to those for deposit insurance) would be needed to avoid giving money market funds an advantage over banks.”22 Indeed this is a fundamental problem—having caps on deposit insurance with no caps on MMF account insurance. One could solve this problem by capping MMF insurance levels or providing unlimited insurance to bank deposits. We would tend to favor the first alternative recognizing that insurance is only part of the answer to contagion. Given a strong lender of last resort, partial insurance should be sufficient.
Despite the potential competitive advantage that liability insurance might confer upon the money market industry, many in the industry continue to oppose it, believing that it would represent the first step toward full capital-based regulation of the money market funds.23 For example, Hanson, Sharfstein, and Sunderam (2014) have proposed adopting money market fund capital requirements as a way to mitigate run risks.24 It is hard to see, however, why capital regulation would necessarily follow from insurance. Money market funds will not become insolvent like banks—their assets are more liquid and less risky—and thus do not really require capital. At most, liabilities may not be worth par. The purpose of insurance for money market funds is only to prevent runs not to insure solvency.
Recent consideration of the designation of asset management companies as SIFIs,25 which would be regulated by the Fed and be subject to enhanced supervision and capital requirements, is clearly not the answer. First, it is hard to see why asset management companies themselves pose any systemic risk or how the designation of asset management companies could prevent runs on the money market funds they manage. Second, the risk of runs on money market funds cannot be countered by SIFI designation of particular funds, if there is a problem it is one that is industry wide. A 2013 report by the Office of Financial Research (“OFR”) suggests that operational or risk-management failures at a particular asset management firm (including valuation problems, fraud, or even reputational damage26 could potentially trigger a run on that firm’s funds. Whatever the likelihood of such an event, the OFR fails to demonstrate the transmission mechanism whereby the idiosyncratic failure of a single fund or manager could spill over into the broader market.
The true competitive impact of expanded insurance on the banking and money market industries requires more detailed study of the appropriate cost and pricing of insurance, before any firm conclusions can be drawn. Given all of the difficulties of extending an insurance regime to money market funds, a more practical approach might be to make clear funds are eligible for lender-of-last-resort support either directly or indirectly through the banks. Liquidity support may be sufficient by itself to stop runs.
Contagion is an extremely serious problem that has endangered our financial system, economy, and, ultimately, our polity in the crisis. Before the Federal Reserve and FDIC were created to contain financial crises, contagion wreaked havoc on the US economy. For example, the panic of 1893 and the panic of 1907 both resulted in economic output reductions of over 15 percent.27 Contagion during the most recent crisis was successfully halted by a combination of the traditional remedies, LLR (as extended to nonbanks), and deposit insurance. But in the aftermath of the crisis, in an anti-Wall Street and bailout frenzy, these powers were significantly curtailed. The Federal Reserve is now a weaker LLR, as can be seen in a comparison with its peer central banks. Calls for additional weakening of the Federal Reserve’s power continue from both political parties.28 According to Federal Reserve Vice Chair Stanley Fischer, such restrictions are “a very high price” to pay for hypothetical concerns about moral hazard.29 We need to not only restore the Fed powers taken away by Dodd–Frank, we need to strengthen the Fed’s power so its strength as lender of last resort will insure its powers may never have to be used.
Given the increased presence of short-term liabilities outside the banking system, we must also consider expanding insurance to deal with the increasing risk of runs outside the banking system. But the scope and pricing of such insurance pose great difficulty. It may be the case that in the future we may have to rely even more heavily on the LLR power to nonbanks, without insurance and guarantees, to stop contagion.