Thank you for the opportunity to speak with you. As we approach one year after the closure of Silicon Valley Bank, I will begin by sharing two approaches to assessing financial stability. Then I will discuss the current outlook for financial stability and close with some brief thoughts about financial stability considerations in the policy context.
Before I go further, I will note that the views I express today are my own and do not necessarily reflect those of the Federal Reserve Bank of New York or the Federal Reserve System.
Financial Stability and the Economy
When winter approaches, we don’t know when it will snow, or how much snow we will get. While we hope for flurries, we prepare for a blizzard, designing buildings and infrastructure to hold up under the weight of heavy snow. Similarly, the work of financial stability is to understand how our financial system will respond to adverse shocks.
Theoretical models show that higher leverage can make the economy more vulnerable. In these models, an economic shock can trigger a self-reinforcing loop as margin calls impair credit provision. Reductions in credit lower asset prices and depress economic activity. Consistent with these models, empirical evidence suggests that financial crises lead to more severe economic downturns and usually follow credit expansions that were accompanied by asset price booms.1 It is this nonlinear response that creates financial stability concerns and that we seek to gauge in the financial stability outlook.
The outlook for financial stability is inherently hard to capture. First, like predicted blizzards, shocks do not always happen. Or, if they do, the government might intervene. In these cases, we may not observe the amplification from financial system vulnerabilities that the models suggest. Second—fortunately—in advanced economies, we have experienced few financial crises over the last few decades. But while we do not get many blizzards in New York City, we still invest in snowplows, as the consequences of being unprepared are severe. For that reason, we monitor the financial stability outlook to understand the extent to which the financial system has the potential to amplify a shock, if one were to occur. Then we work to mitigate financial system vulnerabilities that would lead to more amplification.
I will discuss two ways of measuring the financial stability outlook.
Statistical Approach
One approach to quantifying financial stability risks is to estimate the probability of very bad outcomes in real economic variables such as GDP and inflation using quantile regressions that measure the relationship of these variables in bad times instead of just on average. We know from economic research pioneered at the New York Fed that growth can become vulnerable as financial conditions change and that the probability of a large decline in real activity in the lowest quantiles rises as financial conditions tighten.2 Each month the New York Fed posts estimates of the Outlook-at-Risk on our website, forecasting the conditional distribution of average real GDP growth, unemployment, and inflation.3
One way of visualizing the financial stability outlook is to look at the size of the tail of the GDP growth distribution. While we hope to avoid shocks that land us in the tails of the distribution, when the bad tail is wide, the economy is vulnerable. Figure 1 is based on the most recent release of Outlook-at-Risk, posted February 21. The dashed vertical line shows the blue-chip consensus GDP forecast. We estimate that the conditional distribution given current financial conditions of the left tail of GDP growth, graphed in blue, is in line with historical standards. That’s similar to the unconditional distribution, graphed in red. This means that the left tail is near its historical average and the probability of a large fall of real GDP over the next four quarters is similar to past levels. You can see that visually, because the conditional distribution lays almost on top of the unconditional one. This compares favorably to the conditional distribution as of December 2022, shown in Figure 2, in yellow. There, the bad tail, measured as the 10th percentile, was much wider.
Financial Stability Vulnerabilities Approach
A second approach to capturing hidden risks to economic outcomes is examining patterns in the financial system historically associated with the amplification of negative real shocks both in the U.S. and internationally. The Federal Reserve’s Financial Stability Report is anchored in this type of monitoring. It tracks several vulnerabilities associated with the 2008 global financial crisis and other crises.4
The first vulnerability is high asset prices relative to fundamentals. When prices are high relative to fundamentals—presumably reflecting investors’ willingness to take on risk—the possibility of sudden declines in prices is inherently higher.
The second vulnerability is business and household leverage. As I mentioned earlier, when borrowing by businesses and households is high, there can be sharp declines in investment and spending if there are declines in incomes or the value of assets securing loans falls.
The third and fourth vulnerabilities arise from the financial sector, from both its leverage and the extent of maturity and liquidity transformation. These interrelated factors can lead to concerns about financial institutions’ ability to absorb losses or funding withdrawals when shocks happen. This can reduce credit availability or lead to fire sales, causing additional price declines and further stress for markets and financial institutions.
Financial Stability Vulnerabilities Outlook
I will turn to the current outlook for financial stability with respect to these vulnerabilities, starting with asset prices. Prices across real and financial assets are high relative to fundamental values and historical averages. Residential real estate prices are high relative to rents, and commercial real estate prices have been high relative to rental income even as demand in some office markets has collapsed. Equity market valuations remain high and corporate bond spreads on the low side.
Turning to leverage, Figure 3 shows the historical path of aggregate business and household leverage relative to GDP. The blue line, nonfinancial business debt to GDP, has exceeded historical trends since well before the pandemic, but appears to have peaked. Interest coverage ratios, the ratio of operating income to interest, also appear to have peaked as fixed-rate debt is refinanced and earnings growth moderates. Despite the theoretical risks of business debt overhang limiting new investment, business credit booms historically have not had the same effect on tail risks to the macro economy as have consumer credit booms.5
In contrast to high leverage in the business sector, , household sector balance sheets, graphed in the black line in Figure 3, are less levered than average, in aggregate. Note that these graphs are on different scales. Most household debt consists of fixed-rate mortgages, limiting the pass-through of higher interest rates. Mortgage credit risk has remained low, as much of this credit has been issued to prime borrowers and homeowners have a significant amount of home equity.
However, we are seeing increasing signs of stress for lower-income Americans. Delinquencies in auto and credit cards loans rose in the second half of 2023, with the increases highest for borrowers with auto and student loans as you can see in Figure 4 in the light blue, red and yellow lines.6 While this trend does not bode well for the prospects for consumption of these younger, cash-strapped borrowers, financial stability implications are limited because this type of debt is a relatively small share of total household debt in the economy. Therefore, the risks of amplification through unexpected credit losses to financial institutions are low.
Finally, most financial crises have emerged when shocks are amplified by fragilities in the financial sector. The New York Fed measures vulnerabilities in the banking system using models of capital and liquidity.7 We revised these analyses last year by marking all securities to market value.
Bank capital vulnerability, shown in the top left of Figure 5, remains relatively flat. However, measures of liquidity stress, shown in the bottom left, have increased since early 2022, driven by a shift from liquid to less-liquid assets and from stable to unstable funding. Fire sale vulnerabilities, in the top right, are also higher, capturing similarities in bank assets. The run vulnerability index in the bottom right measures the fraction of runnable funding a bank needs to retain in a stress scenario to prevent insolvency. Our measures suggest a moderate increase in systemic vulnerability compared to the low levels of the previous ten years. In aggregate, systemic vulnerability remains below the high levels preceding the 2008 crisis, especially because the largest banks tend to be less exposed to capital shortfalls, fire sales, liquidity mismatches, and run risk compared to smaller institutions.
Thinking about vulnerabilities that can turn snow into an avalanche, many times, nonbank financial institutions are the source of issues. Tighter monetary policy has spurred flows into money market funds, including prime money market funds, which continue to have liquidity risks that make them structurally vulnerable and at risk of fire sales. Funds that seek to maintain stable net asset values such as retail prime money market funds are particularly susceptible to sharp changes in interest rates and asset prices. Since the October 2023 Financial Stability Report, assets under management in prime money market funds increased by $82 billion, driven mostly by inflows into retail prime funds.
As this audience probably knows, identical risks exist in the stablecoin sector, where investors can similarly treat their holdings as akin to cash. Researchers at the New York Fed have estimated that redemptions for stablecoins accelerate significantly when their prices drop below 99 cents, reminiscent of the “break-the-buck” threshold for money market funds.8 Additionally, even funds without stable net asset values, such as open-end corporate bond funds, present fire-sale vulnerability in the face of large price changes because they offer daily liquidity despite holding less liquid assets.
Financial Stability Vulnerabilities in 2024 and Beyond
Both the statistical approach and a consideration of factors associated with financial stability vulnerabilities offer reasonably comforting pictures.
That said, I am keeping my eyes on four key financial stability vulnerabilities for 2024 and beyond.
First are pockets of risk in the banking sector. One lesson of the banking stress in March 2023 is that aggregate measures of financial stability, which weight the largest banks according to their sizes, may mask vulnerabilities at some individual banks. While the banking industry overall is sound and resilient, changes in interest rates have negatively affected the value of long-duration assets and the impact of revaluations of commercial real estate has been unevenly distributed.
Office commercial real estate is an area of ongoing supervisory monitoring. While bank supervisors continue to work to ensure that banks have adequately reserved for credit losses, higher-than-anticipated credit losses may require banks to increase loss allowances, reducing bank capital. Silicon Valley Bank was an outlier along many dimensions, including the scale of its mark-to-market securities portfolio losses relative to capital and its large share of uninsured deposits. Ensuring appropriate risk and liquidity management by banks with a variety of exposures continues to be important.
Looking at banks’ value, we must consider not just their assets but the economic value of banks’ deposit franchises. In the last two decades, bank profits have increased with tightening monetary policy as bank profitability benefits from lagging adjustment of deposit rates to rising rates. This cycle is no different—New York Fed research has found that bank deposit betas have generally been on track with past monetary policy tightening cycles, with betas on total deposits approaching 0.4 using data through the end of 2023.9 However, the value of the deposit franchise can disappear in a run and so needs to be turned from bank profits into bank capital when banks are realizing losses.
I’ll touch on financial conditions for a moment to note that the impact of a range of factors on bank lending supply combined with the potential for higher deposit betas at certain smaller banks means that smaller, bank-dependent borrowers have experienced tighter financial conditions than larger borrowers with access to capital markets.
A second concern emerging from higher rates is the acceleration of deposits into prime money market funds and stablecoins, both of which I mentioned earlier as sources of non-bank vulnerabilities. The structural risks of products that transform securities into cash is well understood and has been a source of official sector interventions in two of the last two episodes of financial market turmoil. The SEC has taken important actions to reduce vulnerabilities and investor incentives to preemptively run. However, there remains a substantial and growing amount of assets susceptible to large redemptions during periods of financial stress. This includes approximately $700 billion in retail prime money market funds, $700 billion in dollar denominated offshore money market funds, and more than $100 billion in stablecoins. It also includes more than $4 trillion in bond mutual funds.
My third area of focus is U.S. Treasury markets. The New York Fed actively monitors trading conditions across Treasury markets, which are the deepest and most liquid in the world. Treasury market liquidity has been somewhat strained across almost all measures for some time, including bid-ask spreads, market depth, and price impact. These measures also look somewhat worse from a longer-term perspective.10 On the one hand, the reduction in liquidity is exactly what we would expect given interest rate volatility, and trading conditions are orderly.11 On the other hand, lower-than-usual liquidity, even if we understand the reasons for it, means that a demand shock will have larger-than-usual effects on prices, and is more likely to trigger a negative feedback loop between sales, volatility, and liquidity.
Reducing vulnerabilities arising from Treasury markets and improving the resilience of market intermediation is a key focus of the New York Fed and other members of the Inter-Agency Working Group for Treasury Market Surveillance. The Federal Reserve has established the Standing Repo Facility and Foreign and International Monetary Authorities Repo Facility which support greater resilience. Interagency efforts have included work to improve data quality and availability through improvements to FINRA Treasury Trade Reporting and Compliance Engine data, and a study of all-to-all trading. The SEC has adopted rules to facilitate additional central clearing of Treasuries and to ensure that market participants who provide significant liquidity are registered with the SEC. We are also working to examine effects of leverage and fund liquidity risk management practices. We know snow can come every winter, so it makes sense to keep investing in boots and shovels.
Finally, a key vector for financial vulnerability is cyberattacks. These have been occurring at accelerating rates, so we must work to reduce their amplification though the financial system. That means not just leaving cyber to the technologists, but thinking about the real business impact of cyberattacks, vulnerable connections between financial institutions, and what changes to the financial system might mean for cybersecurity. For example, in a study I conducted with New York Fed colleagues, we estimate that a successful cyberattack on a designated financial market utility would double or triple the typical daily payments required for the same transactions, an amount of two to three times firms’ average reserve balances.12
At times when a significant amount of value is coming from deposit franchises where interest rates on total deposits have increased by less than half of the actual change in the fed funds rate, banks are more vulnerable to runs. This also means that banks are more vulnerable to cyberattacks, including misinformation that could spread rapidly and lead to erosion of the deposit base.
Turning to the Federal Reserve’s balance sheet, reserves are little changed since the middle of 2022, when balance sheet reduction began. This is due to reductions in take-up at the overnight reverse repurchase agreement facility (ONRRP). At its January meeting, the FOMC said that it will continue reducing its holdings of Treasury securities and agency debt and agency mortgage-backed securities, as described in its framework announced in 2022.13 The FOMC has noted its intention to continue operating in an ample reserve regime. Ongoing balance sheet reduction is proceeding well. Thus far, we have seen few signs that reserves are approaching the point where changes in supply have an impact on the fed funds rate. Reserves remain above ample. All else equal, a system with more reserves is less vulnerable to shocks than is one with fewer reserves; therefore, as ONRRP use declines and reserves begin to fall with balance sheet reduction, we will continue to closely monitor money market conditions.
Financial Stability is not a Goal in and of Itself
The business of financial intermediation is inherently risky. Banks engage in liquidity transformation: They take savings, offer daily withdrawals, and make loans. Our financial system has vulnerabilities.
The key to financial stability is understanding when vulnerabilities are so great that we are in danger of a nonlinear response in the economy. We contribute to a more resilient financial system by understanding and mitigating sources of these nonlinearities. Examples of this include the transition away from LIBOR to the adoption of more robust rates like the Secured Overnight Financing Rate (SOFR) and the ongoing work on Treasury market resilience. Robust bank supervision that ensures banks are prepared to facilitate access to credit even in a crisis also furthers this goal.
Conclusion
Achieving a strong U.S. economy and stable prices is paramount, and remaining aware of the impact of policy choices on the financial system is a key ingredient to maintaining the ability to execute policy. To close with the snow metaphor I began with, if there is a blizzard in March, we will be prepared to dig out quickly, plow the streets, and get back to work.
