Risks that non-bank financial institutions pose to financial stability

Thank you for inviting me to address this important conference. I was co-chair of a Task Force on Financial Stability, a joint effort of the Hutchins Center on Fiscal & Monetary Policy at the Brookings Institution and the Initiative on Global Markets at the University of Chicago Booth School of Business, that issued a report in 2021 on the very subject of this conference. Our report saw important threats to financial stability from nonbank financial institutions (NBFIs) and made recommendations to address those vulnerabilities. I thought a good way of organizing this introductory keynote would be to go back to that task force report and see what’s been done to address those threats, and what remains to be done.

I will start by highlighting what I think is a very unusual and high-risk financial stability situation we are in right now. We are experiencing, as we’ve seen over the last couple weeks, rapidly shifting frameworks for important government policy areas, like tariffs. These shifts, which can happen in a press conference or by posting on Truth Social, have created a huge amount of uncertainty about the evolution of the economy with an enormous range of possible outcomes. One goal of the tariff policy seems to be to completely re-order the global trading system, greatly reducing, if not eliminating, the longstanding U.S. deficit in trading in goods, with associated reshuffling of global supply chains. This re-ordering—partial deglobalization—might well also have implications for integrated international financial markets and the role of the U.S. dollar. We had a whiff of a flight from dollar assets just a week ago as the dollar fell, term premiums on U.S. Treasury bonds rose, and the stock market fell.

Moreover, we’ve got a fiscal policy problem in the U.S. in which neither party seems intent on stopping or even slowing the rapidly rising debt-to-income line, with potential implications for interest rates. And we’ve had challenges to Fed independence. All this changes week to week, and the trajectories for these policies next week and going forward is unknown. This is a period of heightened risk, fat tails, and distributions of economic and financial developments we didn’t even think we’d be looking at just a couple months ago. Given this risk environment, we should be building additional resiliency in our financial system. But at the same time there’s a very marked deregulatory push in the U.S that could erode existing resilience and certainly makes it harder to build regulatory or supervisory safeguards against emerging risks. We appear to be in a situation of high risk, possibly reduced resilience, and, in addition, questions about crisis response. In a private conversation with me, a visiting central banker recently wondered about international cooperation in a financial crisis when the U.S. seems to be waging economic war on the rest of the world. In a crisis we are accustomed to everybody everywhere globally communicating freely and cooperating. This person asked whether we can be sure that’s going to happen this time under these circumstances.

This is a fraught moment in financial markets with financial stability risk. Accordingly, this is a good time to be looking at non-bank financial institutions and financial stability. NBFIs are very important in intermediation. Many are subject to run and fire sale externalities. Our information about them is incomplete because in many cases they are not required to be transparent. And regulation, especially in the U.S., is fragmented and partial. I will start by reviewing a few high-level trends in NBFI intermediation to set the stage for the issues we will be dealing with over the day. And then I will review the recommendations our Task Force made in five areas: insurance companies, open-end funds, housing, central counterparties (CCPs), and regulatory structure and process.¹ And I will conclude with a few suggestions about what else the Task Force might be looking at if it were formed today, hoping to spark a discussion among conference participants about emerging risks. 

This chart shows quite a flat trajectory for NBFI assets as a proportion of total assets—a bit of a surprise given how much discussion there is of the rising importance of NBFIs in financial markets. To be sure, this calculation might be sensitive to what is included in the numerator and denominator, but the roughly unchanged proportions since 2000 are consistent with a similar chart for the U.S. published by the Financial Stability Board, while the proportions are rising elsewhere. But look at the level of the lines—75% NBFI, 25% banks. This is a huge sector, and by any measurement more important than banks in intermediation in the U.S.

Although the total proportion is flat, NBFIs are picking up market share from banks in an important market—the financing of nonfinancial businesses, shown on the second chart.  Undoubtedly, that reflects, at least in part, the very rapid growth of private credit in recent years; middle market firms are increasingly finding funds outside the banking system.

Although the proportions are flat between banks and nonbanks, banks have been increasing their involvement with nonbanks. This chart shows the percentage of bank lending that is going to nonbanks. That spike at the end is lending to broker-dealers; even if it turns out to be a bad or one-time data point, the trend is clear. Even though banks are losing share in business loans, they are, to some extent, compensating for that by doing more lending to the NBFIs who are in turn lending to businesses. As a consequence, banks are becoming more exposed to problems that might arise in NBFIs; the financial system is not as well diversified as it might appear on the surface.

For each sector the Task Forced examined, we began by identifying the externality that might result from problems in that sector, justifying additional regulation or other reforms. For life insurance, we worried about two forms of externality. One was the availability of credit to businesses because insurance companies play a key role in funneling credit to a wide variety of businesses. The other was the availability or terms of insurance itself; if insurance companies are impaired and can’t write insurance, or they tighten the terms to make up for their exposure to financial losses, the ability of households and businesses to manage their risks will be impaired, and they will cut back spending. Our recommendations in 2021 emphasized stepped-up stress testing of the insurers and the state insurance guarantee associations, and more transparency on the risks insurers are taking, including in their increased use of captive reinsurers offshore through which they are arbitraging regulations and taxes. And we proposed giving authority to the Federal Insurance Office, to standardize the stress tests and make sure they take account of macroprudential externalities.  

The National Association of Insurance Commissioners (NAIC), which is a coordinating body for state insurance commissioners, has made some recommendations to address several of these vulnerabilities, including liquidity stress tests and greater transparency on off-shore re-insurance.   But these are recommendations to the states and not every state has adopted them; for example, only 35 jurisdictions have taken up liquidity stress tests.² And the stress test itself does not have a systemic macroprudential add-on. As an organization of state regulators, the NAIC naturally is not a fan of federal oversight and abolishing the Federal Insurance Office is part of its 2025 legislative priorities. 

Meanwhile, insurance company business models evolve toward greater risk. On the asset side of their balance sheets, they have become more involved in private equity and private credit, which are less liquid, more complex, and harder to value (mark to model rather than mark to market) than marketable securities. And they have become more reliant on nontraditional liabilities, like borrowing from credit markets and the Federal Home Loan Banks. And a growing proportion of their capital comes from private equity, not known as patient investors. On balance, we have a difficult economic and financial situation, insurance companies moving in more risky directions, and regulatory oversight not keeping up with these developments.

For open-end bond funds, the externality arises from the liquidity mismatch of such funds, which promise more liquidity to investors than is often inherent in the assets they hold. For example, the funds may be holding junk bonds that are traded, if at all, in very illiquid markets, but offer overnight redemption to investors. That liquidity mismatch is exacerbated in prime money market funds, where they promise to give a dollar back for a dollar in, even if the value of assets has slipped some. The liquidity mismatch provides a first mover advantage in stressed markets with the resulting run sparking a fire sale by the funds. Our recommendations included adopting swing pricing for both prime money market funds (MMFs) and other open-end funds such that prices at redemption would be adjusted lower to take account of the effects of a surge in redemptions, reducing or eliminating the first-mover advantage. We also recommended breaking the regulatory link between gates on withdrawals on MMFs and the fund’s liquidity, which had exacerbated first mover advantage for those funds. 

Swing pricing has not been adopted for any of the funds because it was so difficult to implement in the U.S.—much more difficult than in Europe. But for prime money market funds, a lot has been done to address the issues threatening financial stability. Liquidity requirements have been raised; institutional MMFs are required to impose redemption fees when withdrawals exceed 5%; and the link between gates on withdrawals and liquid asset holdings has been eliminated. Other open-ended funds are subject to more reporting on liquidity to help people better understand the risks of the funds they’re investing in, but nothing material has been done to mitigate the run risk arising from the first mover advantage of getting out of a bond fund while the market is going down. That particular systemic risk still exists and is likely growing in importance as mutual funds diversify into private equity and private credit. 

Housing and mortgages, of course, have been an important cyclical amplifier in the U.S. They were at the center of the financial crisis of 2007-09 and were a source of systemic problems even before then—for example, in the failure of many savings and loan associations in the 1980s and early 1990s. One externality is the fire sale risk that can arise when people are in default on their mortgages, and either they or the lending institution must sell the house. Those sales drive down prices exacerbating the housing market decline and associated reductions in spending and employment. As in 2007-09, disruptions can originate in the mortgage market, reducing the availability of credit broadly as well as for housing. 

Our recommendations focused on damping the feedback loop between foreclosures and house prices by altering the waterfall of remediation as prices go down and people are underwater; on creating a liquidity backstop for mortgage securities guaranteed by the government or by a government-sponsored enterprise; and strengthening nonbank mortgage servicers.

With respect to the damaging feedback loop, we have seen a very positive development. When COVID hit, the waterfall of what happens when a borrower can’t make mortgage payments was changed. It became forbearance first for as long as 12 months, restructuring second, and foreclosure and sale only when borrowers could not meet the terms of the restructured loan.   And that ordering has been made permanent for loans that flow through Fannie Mae, Freddie Mac, and the Federal Housing Administration. This should greatly reduce the systemic risk from defaults and fire sales in the housing and mortgage markets. But we don’t have a permanent liquidity backstop for mortgage securities; the Federal Reserve has stepped into the mortgage-backed securities market, but only after it became seriously dysfunctional and as part of an effort to stabilize a broad swath of markets under severe duress. And non-bank mortgage servicers remain a vulnerability. The Financial Stability Oversight Council (FSOC) published a study that identified weaknesses and made some recommendations, but most of these were for legislation, which isn’t going to pass. Non-bank mortgage services are state regulated, and they remain vulnerable to default and liquidity pressures. 

These issues persist at a time when the Federal Reserve itself has identified house prices as a systemic risk, judging, for example, from a price-to-rent ratio that was quite elevated relative to history.³ If, as many people fear, we are looking at a period of weak growth or even recession and high interest rates to lean against tariff-induced inflation, house prices could start to decline, and that will test the mortgage industry. In sum, despite important progress, housing remains a systemic risk originating in NBFIs.

Central counterparties (CCPs) or clearinghouses are critical financial market utilities. A problem in a CCP would echo through the financial markets in a number of ways. It would disrupt key risk transfer markets, impairing the ability of users to manage their risks, and it could threaten the functioning of the clearing members called on to back the CCP—and in so doing could threaten other utilities where these firms do business. Procyclical margining practices at CCPs protect the CCP, but place demands on liquidity that can amplify price movements if users are forced to sell assets in size. 

Our recommendations to address these externalities included making margins less procyclical and taking steps to limit the likelihood of severe problems at CCPs and to limit the spillover of any problems to other players in the financial markets. Those steps included stress test and loss allocation rules that incorporate systemic effects; better resolution planning; more CCP skin in the game to better align CCP and clearing member interests; and greater Fed involvement in oversight to bring a macroprudential perspective. 

None of those recommendations were well received by the industry or its regulator. The industry asserted that our concern was unnecessary; there was no systemic risk, they had it covered. The Commodity Futures Trading Commission pushed back hard against greater Fed involvement.  Those guys don’t know much about this; we know what we’re doing, don’t worry.

But I can’t help worrying. The Financial Stability Board, the Basel Committee on Bank Supervision and other international coordination panels, recognizing the centrality of CCPs in financial stability, have paid a lot of attention. They have issued two sets of recommendations. One was aimed at the CCPs but did not include a recommendation to alter margining practices to make them less procyclical, despite evidence that has contributed on several occasions to sharp disruptive movements in asset prices, especially in the Treasury market. Instead, it focused on increasing the transparency of margining practices, on the thought that it would help users anticipate margin calls and meet them in a less disruptive manner. The recommendations were numerous and would increase transparency among users at every level.⁴ And they made recommendations to the users to improve management of liquidity risk at CCPs and other markets where they faced margin or collateral requirements that could shift with price movements or increases in volatility.⁵ In the U.S., CCPs have added clearing members to their risk committees, which will help those members understand and anticipate margin developments.  But that is only a start on the ten recommendations to improve transparency, so margining is still procyclical and transparency is incomplete for many users. Nothing has been done to put more CCP skin in the game to protect clearing members and align the CCP’s interests with those of its members. Liquidity stress tests have been enhanced, but it is my understanding that they still don’t incorporate potential systemic issues, like an extended market disruption as we experienced in 2008 and 2020. CCPs remain a point of vulnerability. 

Finally, the Task Force took up the subject of regulatory structure and process—what could be done to improve the recognition of systemic risks in the future and to formulate and implement policies to reduce them. One recommendation was to elevate the attention to financial stability at all the agencies; every member of FSOC should have financial stability as part of its legislative mandate and an office to assess the effect of potential agency actions on stability. We also wanted to strengthen FSOC by making its analysis and recommendations less consensus driven. Our sense, confirmed by people who have been participants in FSOC deliberations, was that the need to find consensus among all these agencies tends to water down the identification of risks to financial stability and the recommendations to deal with those risks. We recommended that the Secretary of the Treasury alone should be responsible for the annual report that identified risks and made recommendations, with the other agencies responding to her findings and recommendations. We wanted to resuscitate and regularize the process for deciding whether NBFIs might be systemically important and therefore subject to macroprudential capital or liquidity requirements designed to mitigate their systemic externality. And we wanted to address information gaps more systematically by focusing the Office of Financial Research exclusively on data. 

Under the direction of Treasury Secretary Janet Yellen and Treasury undersecretary for domestic finance Nellie Liang, FSOC made great strides in living up to its mandate under Dodd-Frank of identifying risks to financial stability, promoting market discipline, and responding to emerging threats to the stability of the U.S. financial system. They formed working groups to identify issues and make recommendations to deal with them; they activated the systematically important financial institution (SIFI) designation process; and they had the Office of Financial Research begin collecting a lot of important new data.

But clearly a lot of these positive developments could be reversed in the Trump administration. Yes, elections have consequences. Democrats and Republicans have different views of how much regulation there ought to be, and that’s fine. The system can benefit from a fresh perspective every once in a while on what’s important and how to achieve those objectives. But it seems to me the swings in perception and policy have gotten larger over the past 10-15 years, and I am concerned that we could lose a core belief that government needs to administer the financial system to protect households and businesses from the effects of severe financial instability, even if there are different views on what that entails and how to do it. Moreover, in the current environment we are losing experienced people who have institutional knowledge and judgment that could help whichever crop are in power achieve their objectives efficiently and effectively with consideration for the longer-term consequences.

I conclude that we have made some progress in building a more resilient financial system on a number of these areas, but the job is far from complete, and some of these gains could be reversed. Many old vulnerabilities remain, and new ones are emerging when we will be experiencing a very volatile economic and financial environment. 

In this situation, you can strengthen the resilience of the system, or you can bolster central bank backstops if resilience isn’t there. But legislation has, deliberately, thrown up barriers to implementing lender of last resort or market maker of last resort for NBFIs. The Treasury must agree to any lending facility to backstop NBFI institutions or markets, and the Federal Reserve has come to depend on a Treasury backstop for lending or purchase of private securities if it is concerned about the value of the collateral. In COVID that worked well in an emergency originating outside the system, and indeed outside the country; Treasury readily agreed, and Congress allocated funds to make sure there was security, and the Fed was comfortable doing what it needed to do. But it’s not hard to imagine that getting this buy-in might not be so easy under other circumstances. The country may face circumstances in which resilience has been eroded by deregulation and the central bank backstop to stabilize the system in a crisis is less readily activated.   

Finally, if the task force were being formed today, I see three additional areas of potential stability risk it might want to explore. One is extremely elevated hedge fund leverage. Hedge funds forced into deleveraging by margin calls in volatile markets has been identified as a source of instability in Treasury markets in the past, and the threat persists. Another area is private credit, which has been growing very rapidly. Here the first step would be getting more information. Who holds it? What are the restrictions on withdrawal of financing—is there a run risk? What about this intersection between the banks and private credit? And stablecoins—Dick Berner reminded me that I once said about stablecoins that they were neither stable nor coins. That remains true. And while there’s legislation moving through Congress, where it will come out and whether it will be sufficiently restrictive in terms of the assets these entities can hold to promise a stable value against the dollar is very much up in the air.