Macroprudential policy is financial policy intended to safeguard the financial system as a whole, as opposed to microprudential policy, which is intended to ensure the safety and soundness of individual financial institutions. Macroprudential policy can be static – time-invariant steps designed to increase the resilience of the financial system – or dynamic – efforts to lean against building financial imbalances or cushion the blow when those risks crystallize, such as tightening or easing prudential requirements.
To execute dynamic macroprudential policy as originally envisioned by the designers of the current official approach, it is necessary to identify financial imbalances that are building up as they develop. To do so, macroprudential policy is informed by assessments of financial stability. But recent errors call into question whether such foresight is possible, suggesting caution is necessary in basing policy decisions on that foresight. Financial stability assessments failed to identify the risks that led to the banking crisis in spring 2023; in addition, the Fed made the incorrect official (although not internal) judgment in 2014 that leveraged lending was a financial stability risk and the Fed’s failure to see the scale of the risks that caused the Global Financial Crisis call into question whether such foresight is possible. Meanwhile the severe but completely unforecastable COVID financial crisis demonstrates that building-risks-followed-by-crisis is not the only way a crisis happens.
The primary tool for conducting dynamic macroprudential policy in the United States is the countercyclical capital buffer, which is set by the Federal Reserve Board. The CCyB is an addition to the capital requirements of large banks that can vary between 0 and 2½ percentage points. The buffer is intended by the Basel Committee and by U.S. regulation to be raised when financial stability risks are building and lowered when those risks recede. The buffer was adopted in 2016 but has never been used, in part because the Federal Reserve has not identified in advance any periods of building financial risks that it judged warrant a positive CCyB.
With the CCyB unworkable as designed, two possible uses for the buffer have become popular among macroprudential theorists and architects. First, some have proposed simply raising the buffer during economic expansions and lowering it during recessions, even though Basel Committee was explicit that the buffer should not be raised and lowered in response to the macroeconomic cycle and the U.S regulation does not discuss such use. Second, some have framed lowering the CCyB as a way to create a “usable” capital requirement buffer – a zone that a bank’s capital ratios can fall into during periods of economic weakness without the bank cutting back on lending.[1] Although the post-GFC capital framework is full of such buffers, they have proven to be unworkable because banks will not allow their capital ratios to fall into them. Whatever the merits of using the CCyB this way, it is important to note that it is illegal for banking agencies to redefine the purpose and operation of regulatory requirements without putting those requirements out for new notice and comment.
This note discusses the failure of dynamic macroprudential policy and the problems with the new contemplated uses of the CCyB. In particular, it addresses three issues. First: it considers whether and, if so, in what circumstances, the government should use prudential tools to achieve strictly macroeconomic aims. For example, should the banking agencies raise commercial bank capital requirements to reduce inflation if inflation is too high? Second, after setting aside using prudential tools for business cycle management, it reviews whether the government can effectively implement dynamic macroprudential supervision to achieve prudential objectives. Doing so depends on 1) the government being able to judge correctly when systemic risks are building; 2) the government not identifying things as a financial stability risk that aren’t; and 3) crises being typically the consequence of building risks. Experience since and leading up to the GFC calls these two preconditions into question. Third, it discusses the problems with using the CCyB as a replacement for the buffers built into the post-GFC capital requirements that have proven unusable, especially if any reductions in the CCyB are combined with limits on banks’ distributions to shareholders.
Common Misunderstandings About Macroprudential Policy
There are two common sources of confusion about the term macroprudential policy. First, the “macro” in macroprudential does not mean the policy is meant to influence or be influenced by the macroeconomy. Regardless of how many times it is repeated that the CCyB is intended to be raised in good times and lowered in bad times, it remains false. The Basel Committee guidance on how the CCyB should be set says explicitly:
…the buffer is not meant to be used as an instrument to manage economic cycles or asset prices. Where appropriate those may be best addressed through fiscal, monetary and other public policy actions. (p.3) (emphasis added)
The United States has adopted a Basel-consistent policy. Regulation Q states that the Board will raise the countercyclical capital buffer (CCyB) when “systemic vulnerabilities become meaningfully above normal” and reduce the CCyB “when the conditions that led to its activation abate or lessen.” Consequently, the regulation states that the Board will monitor a range of information about financial market developments to set the CCyB; it does not state that the Board would monitor the state of the economy.
The second confusion is closely related. Section 616 of the Dodd Frank Act says that the banking agencies should make capital standards “countercyclical.” However, the “countercyclical” in the CCyB, as described above, and the “countercyclical” in the Dodd-Frank Act have different meanings. The cycle that the CCyB is meant to counter is the financial cycle – periods of excessive credit growth, excessively narrow risk premiums, or undue risk-taking by financial institutions – followed by financial crises. The cycle the Dodd-Frank Act is concerned about is the normal business cycle: it specifies that, to be countercyclical, “the amount of capital required to be maintained by an insured depository institution increases in times of economic expansion and decreases in times of economic contraction.” The Fed seeks to comply with this requirement by, for example, assuming a larger increase in the unemployment rate in its stress test scenarios when the unemployment rate is low than when it is high, but not through the setting of the CCyB.
The concept that the government could execute dynamic macroprudential supervision grew from reflections on how the GFC could possibly have been prevented. The crisis had its roots in the deterioration in lending standards on home mortgages that had begun several years before. The subsequent loss rates on mortgages extended in the few years before the crisis ended up much higher than those on mortgages made earlier or later. Many of those involved in designing the CCyB judged that, if the government had acted to rein in that deterioration, the crisis, and attendant catastrophic costs, could have potentially been avoided.
Should the Government Use Prudential Tools to Achieve Macroeconomic Objectives?
Traditionally, the government uses monetary or fiscal policy to influence the macroeconomy and prudential requirements to influence risks to the financial system (bank safety and soundness and financial stability risks). However, monetary and fiscal policy can influence risks to the financial system and prudential requirements can influence the macroeconomy. There is, of course, no controversy over the appropriateness of using prudential tools to achieve prudential aims or using monetary and fiscal policy to achieve macroeconomic aims. To clear the way for clear thinking on the use of dynamic macroprudential policy tools, it is necessary to consider if it is appropriate in conceivable circumstances to use prudential tools to manage the macroeconomic cycle.
There is a large literature devoted to the question of whether monetary policy should be adjusted to address risks to the financial system.[2] In particular, when the central bank sets interest rates very low, investors may be incentivized to take on more risk to earn a higher return. If so, and if the additional risk-taking is a material threat to the financial system broadly, the question arises as to whether the central bank should raise interest rates at least somewhat to reduce risk-taking. There is a broad consensus around the view that governments should use prudential tools to reduce problematic risk-taking to the extent possible, aiming monetary policy exclusively at macroeconomic outcomes. Monetary policy should be adjusted to influence risks to the financial system caused by low rates that for some reason cannot be addressed by prudential tools, and even in that situation, the case for adjusting monetary policy is weak. It is weak because the cost of higher interest rates in terms of lower GDP and lower employment occurs with certainty and continuously whereas the benefits in terms of a reduced probability of a financial crisis or a less costly crisis are only relevant for unlikely outcomes.[3]
What has only rarely been discussed is whether prudential tools should be used to achieve macroeconomic outcomes. For example, suppose the bank capital requirement is established at its optimal level – that is, the marginal public and private benefit of raising requirements in terms of a reduced probability of a financial crisis and a lower likelihood of bank failures equals the marginal public and private cost in terms of more costly credit and lower deposit rates. Suppose further that there is no macroprudential reason to raise capital requirements – the risk of losses by banks and the risk of a financial crisis are normal. But the unemployment rate is below the NAIRU and the inflation rate is above the central bank’s target of 2 percent. Should the prudential authority raise the capital requirement to slow economic activity, raise the unemployment rate, and lower inflation?
If the economy can be slowed and inflation reduced using monetary or fiscal policy, it would be suboptimal to change the capital requirement. The capital requirement is at its socially optimal level, so it would reduce social welfare if it were increased, a needless cost if normal macroeconomic tools are effective.
However, suppose monetary and fiscal policy are not available. For example, in a recent paper, Pablo Hernandez de Cos, then Governor of the Banco de Espana and Chairman of the Basel Committee on Banking Supervision, observed that prudential tools might be useful for achieving macroeconomic stability in a country in the Eurosystem, in the case where adjusting fiscal policy was not viable, because the countries do not have an independent monetary policy.[4]
In that case, it might be optimal to raise capital requirements to slow the economy, raise unemployment, and reduce inflation. However, there are also costs associated with raising capital requirements, because (by assumption) requirements had been set optimally. Moreover, there are costs associated not only with capital requirements being tighter than optimal, but with changing them around over time because doing so makes capital planning more difficult for banks and requires banks to maintain a buffer over minimum requirements all the time. Furthermore, there are lags between when capital requirements would be changed and when economic activity would respond, and the relationship is highly imprecise. For all these reasons, the optimal adjustment to capital requirements is likely small. If there are fixed costs to the financial system of varying capital requirements at all as opposed to keeping them constant, the marginal benefit to raising capital requirements may well be insufficient to warrant incurring those fixed costs, in which case, capital requirements should be left unchanged.
When Hernandez de Cos considers the prospect of the CCyB being positive in normal times so that it can be reduced in recessions, he only contemplates the entire stack of capital requirements being increased in normal times, stipulating that “overall capital levels are to be preserved in periods of stress.” But as long as capital requirements had been set optimally, raising them in normal times so that they can be lowered in periods of stress is almost surely suboptimal assuming that stress periods are relatively rare compared with normal periods. The cost of having requirements that are too tight would be incurred most of the time with the benefit of lowering it to zero in periods of stress only accumulating when there is stress. Moreover, because the CCyB would have been zero in times of stress anyway, the only benefit of the new policy could come from the change in the CCyB, not the level. One is reminded of the following scene from The Office:
DWIGHT: I can raise and lower my cholesterol at will.
JIM: Why would you want to raise your cholesterol?
DWIGHT: So I can lower it!
Furthermore, when the prudential authority reduces the CCyB to stimulate the economy, it will likely also limit capital distributions by banks. For example, De Nederlandsche Bank (DNB, the central bank of the Netherlands) states “[when the CCyB has been reduced] we will emphasize – as in March 2020 – that the released buffer is intended to support lending, not for extra dividend payments or share buy-backs.”[5] In March 2020, the DNB stated “It is paramount that banks use this capital to support lending, and not to distribute dividends or fund share repurchases.”[6] If the CCyB is lowered in difficult times and banks are prohibited from disbursing capital when the CCyB is lowered, then bank stocks will have the undesirable characteristic of paying off in good times but not in bad. Because such a policy would make bank stock less attractive, banks’ cost of capital would be permanently higher.[7]
While it may seem obvious, therefore, that prudential tools should not be used to achieve strictly macroeconomic objectives, it is worth noting that a couple generations ago the opposite conclusion seemed obvious. A half century ago, monetary policy was seen as influencing economic activity primarily through changes in bank balance sheets – money on the liability side and credit on the asset side. For example, reserve requirements required banks to hold some fixed percentage of deposits as cash or deposits at the Fed. Reserve requirements were very similar to the current liquidity coverage ratio (LCR) requirement, which compels banks to hold fixed percentages of bank liabilities and assets as reserves and other liquid assets. Raising or lowering reserve requirements and thereby lowering or raising loan supply was then regarded as a perfectly normal way to manage the macroeconomy. Raising and lowering the LCR to influence the macroeconomy therefore has a certain surface legitimacy. While managing the macroeconomy with capital requirements is not exactly the same, it is close enough that it can’t be simply dismissed.
Nevertheless, it has been decades since varying reserve requirements, as opposed to varying interest rates, has been seen in the United States as the appropriate way to conduct monetary policy. Presumably the same reasons why using reserve requirements to manage the economy fell out of favor, including the breakdown of the relationship between monetary aggregates and the economy, also apply to using the LCR and capital requirements to do so.
Macroprudential Policy for Macroprudential Objectives
If we can stipulate, therefore, that prudential tools should not be adjusted to achieve macroeconomic objectives – tightening in good times and easing in bad times – what about the prospect of varying prudential tools to achieve macroprudential objectives? The failure of macroprudential supervisors to see in advance the tinder that led to the spring 2023 banking turmoil is just the last of a string of episodes that cast doubt on the ability to conduct effective dynamic macroprudential policy.
In terms of the GFC experience, it was widely known that lending standards on home mortgages had deteriorated and that poor loans were being packaged and sold as private subprime mortgage-backed securities. What was missed was that the securities, as well as certain other securitization products, were being funded through chains of intermediation in the shadow banking system by runnable liabilities. Fed staff (myself included) judged that the worst that could happen was some investors would make losses on subprime mortgage-backed securities; staff did not foresee that a run on the shadow banking system would result in the worst financial crisis since the Great Depression.[8] That is, even if there had been a macroprudential supervisor at the Fed with macroprudential tools at her disposal running up to the GFC, she would not have taken action.
The converse situation occurred in the mid-2010s with leveraged loans. Standards and terms had slackened on leveraged loans that were originated by banks but not held by banks. However, the loans were funded by stable structures that were not subject to runs such as CLOs. As discussed in Baer and colleagues (2017), the Fed issued stringent guidance to supervisors to stop banks from originating loans that were ineligible for the banks to hold on their own balance sheet, pointing to risks to the financial system as a reason because the loans clearly did not represent a safety and soundness risk to banks.[9] Subsequent events, as well as analysis by the GAO, concluded that the loans were not a systemic risk.[10] Instead, as discussed by Baer and colleagues, the guidance was motivated by a desire of FOMC leadership to convince the Committee to keep monetary policy stimulative despite concerns by some Committee members that the easy policy was pushing investors to take on high levels of risk.[11] Because of the guidance, businesses were denied credit based on unwarranted conclusions.
By contrast, the COVID financial crisis in 2020 was unambiguously systemic and the government appropriately eased bank regulations in response. But the crisis was also unforecastable, again a situation ill-suited for dynamic macroprudential policy as originally envisioned.
The bank crisis in the spring of 2023 has cast further doubt on the potential for macroprudential supervision. The conditions that gave rise to the bank troubles are textbook examples of tinder for a financial crisis. Silicon Valley Bank was funding investments in long-term risky assets with short-term, cash-like, unreliable funding. Its failure sparked a classic bank run that brought down Signature and First Republic. For example, the Fed’s financial stability webpage states:
Elevated funding risks can occur when financial institutions raise funds from the public with a commitment to return investor money on short notice, but those institutions then invest much of the funds in illiquid assets that are hard to sell quickly or in assets that have a long maturity. This liquidity and maturity transformation can create an incentive for investors to withdraw funds quickly in adverse situations, commonly referred to as a “run.”
Why, then, did none of the U.S. financial stability monitors see it coming? It could be because none of the instruments that the monitors were monitoring was designed to capture the risk. Nelson and Passmore (2001) “Pragmatic monitoring of financial stability” lists many indicators that the Federal Reserve keeps tabs on to keep track of financial stability including risk spreads, market liquidity measures, and option prices.[12] With respect to banks specifically, they list growth in business loans and terms and conditions on business loans as loan performance, capital ratios, and profitability. None of these measures would pick up the interest rate risk in SVB’s securities portfolio or the unreliability of SVB’s funding.
In addition to these measures, Nelson and Perli (2007) “Selected Indicators of Financial Stability” also list some broader indexes of financial stability conditions that the Federal Reserve had developed.[13] In particular, they list a financial fragility measure and an estimated probability of multiple bank failures. While updated readings from the indexes are not available, the similar financial stress index calculated by the St. Louis Fed judged that stress was low on the eve of the failures, and given the sanguine readings from equity markets, the failure probability would probably also have been judged to be low.
Adrian, Covitz, and Liang (2015) “Financial Stability Monitoring” provide additional aggregate measures of financial stress – CoVaR, Srisk, and DIP, as well as various market measures of bank condition, but all the measures are focused on GSIBs and so would not have picked up worries about SVB and some other regional banks.[14] They also point to supervisory stress tests as a means to identify risk, but the stress tests do not assess funding reliability, nor, normally, do they measure interest rate risk. In their discussion of bank liabilities, the authors classify deposits as stable funding and group deposits with longer-term debt and equity without breaking out uninsured deposits.
One index that at least had the potential to see SVB’s risk coming is the New York Fed’s run vulnerability index (see Choi and colleagues 2018). That index includes uninsured deposits as unstable funding.[15] However, as reported in Kovner (2024), the index for banks on average on the eve of SVB’s failure, using the methodology Choi and colleagues were then applying, indicated vulnerabilities were at a record low.[16]
Perhaps because they were looking in the wrong direction: the Board’s November 2022 financial stability report does not mention any of the risks whose crystallization brought down SVB. For example, the discussion of leverage in the financial sector did not mention interest rate risk or unrealized losses.[17] The discussion of funding risk stated that the risk for banks was “low”. Although the report does include uninsured deposits in a graph of “runnable money-like liabilities” it does not include them in a measure of bank wholesale funding.
Equity market investors were also fairly upbeat about SVB prior to March 2023. While SVB’s stock declined substantially in 2022, likely reflecting in part the slowdown in the venture capital business, it had increased by about 40 percent over the months just before it closed. Its P/E ratio was 10, in line with the average for other banks. That said, Autonomous Research in November 2022 expressed concern, shifting the stock to “neutral” from “buy,” noting “…the numbers in black and white are a bit of a wake-up call: SIVB’s tangible book would be negative if HTM securities were included.”[18]
Cutting the CCyB to Create a Usable Buffer
Having failed in its original objective, the objective of the CCyB has been reimagined among macroprudential thought leaders and policymakers. The post-GFC capital regime consists of a minimum requirement and several buffers – the capital conservation buffer, the GSIB surcharge (if applicable) and the countercyclical capital buffer (if nonzero).[19] In general, a bank whose capital level falls below the sum of the minimum and the buffers is prevented from making capital distributions to shareholders. Oddly, the consequences of falling below the minimum are not spelled out in the Basel standard or the U.S. implementation but are no doubt bad. The intent of the design was that banks should be willing to allow their capital levels to fall into the buffers during transitory periods of actual or expected losses or model-induced increases in risk-weights without cutting back on lending, lessening the inherent procyclicality (in the macroeconomic sense) of capital requirements.
Not surprisingly, however, the buffers have proven unusable. That is, banks have been unwilling to allow their capital levels to fall into the buffers. Instead, cutting the CCyB from some positive level has come to be seen as a means to, in effect, create a usable buffer. That is, the “buffer” is the difference between where the CCyB was and where it ends up. That view is reflected, for example, in the commonly heard complaint that the United States had not established a positive CCyB going into the COVID pandemic based exclusively on the desirability of being able to reduce the buffer even though such use has no grounding in the rule establishing the regulation.
Using the CCyB to replace the CCB unfortunately ports over the limitations on distributions on shareholders that apply if a bank falls into its CCB. As with cutting the CCyB to stimulate the economy, discussed above, the objective behind such a restriction is so that a bank uses the “buffer” to support lending, not to buy back shares or pay dividends. But whereas the limitations would apply to an individual bank that falls below the CCB, they would be applied to the entire banking industry whenever the CCyB was cut, even to banks whose capital levels remain above where the CCyB began before the cut.[20] A core tenet of modern finance theory is that investors will require a lower return on assets that pay off in bad times rather than good times because the asset is a desirable addition to a portfolio. In this reimagining of the CCyB, bank stocks become the opposite – ones that can only pay off in good times and not in bad times. Not surprisingly, an informal poll of bank investors found that such a policy would lead them to require a higher return on bank stocks – three-quarters of the investors indicated such a ban would cause them to reduce their valuations of banks stocks by 5 percent or more.[21]
Moreover, the current set of capital requirements – minimum plus buffers and CCyB at zero – was calibrated to be optimal in normal times with banks dipping into the buffers such as the CCB in stress times. A positive CCyB is only optimal if risks to banks are materially higher than normal.[22] Therefore, if a cut in the CCyB is intended to replace existing buffers, then optimal calibration would have a positive CCyB replace the CCB not be added to it. Nevertheless, over four-fifths of the European jurisdictions that implemented a positive neutral CCyB did so by raising the overall level of capital requirements.[23]
The bottom line, though, is that the Federal Reserve should not and cannot unilaterally change the intended purpose of existing capital requirements and the mechanics of how they will be adjusted. Those requirements, including their intended use, were established through a notice-and-comment rulemaking process. If buffers don’t work, the entire original framework doesn’t work and needs to be reconsidered. If the agencies intend to instead switch to a regime where capital distributions by all banks, irrespective of their condition, are banned in bad times, that radical new policy also needs to be proposed and discussed and voted on before being adopted.
Conclusion
If dynamic macroprudential supervision is not feasible, what should be done instead?[24] It might seem apparent that the alternative would be to make static bank regulations more stringent. However, that conclusion is inconsistent with the analysis that went into the calibration of the current regulatory regime. The calibration of each of the components of the capital requirement stack as well as the Basel Committee’s cost-benefit analysis of the level of requirements overall were not based on an expectation that the CCyB would be positive before periods of financial stress allowing for easier static requirements in normal times. Instead, the solution may simply be to set aside the idea that a macroprudential policymaker can see crises coming and stick with bespoke responses to each unique crisis when it occurs.
[1] For example, on January 31, 2025, the European Central Bank and European Systemic Risk Board stated that the foremost benefit of having a positive CCyB in normal times was to create the availability of “releasable buffers.” “Using the Countercyclical Capital Buffer to Build Resilience Early in the Cycle” p. 21. https://www.ecb.europa.eu/pub/pdf/other/ecb.jointreportecbesrb202501~f5cf374d79.en.pdf
[2] For an overview of the literature, see English, W. (2022). Monetary Policy and Financial Stability in Handbook of Financial Stress Testing edited by Farmer et al. Cambridge U. Press. For a concise and influential discussion of the issues, see Stein, J. (2013). Overheating in Credit Markets: Origins, Measurement, and Policy Responses; remarks at the “Restoring Household Financial Stability after the Great Recession: Why Household Balance Sheets Matter” research symposium sponsored by the Federal Reserve Bank of St. Louis, St. Louis, Missouri, Feb. 7. https://www.federalreserve.gov/newsevents/speech/stein20130207a.htm.
[3] Svensson, L. E. (2016). A Simple Cost-Benefit Analysis of Using Monetary Policy for Financial Stability Purposes. Progress and Confusion: The State of Macroeconomic Policy, 107.
[4] Hernández de Cos, P. (2024). The role of macroprudential policy in the stabilisation of macro-financial fluctuations. Conference on Financial Stability/Banco de Portugal, Lisbon (Portugal), 2 Oct. 2023. Documentos Ocasionales/Banco de España, 2403.
[5] https://www.dnb.nl/media/gd1m1mps/analytical-framework-for-setting-the-countercyclical-capital-buffer-in-the-netherlands.pdf
[6] https://www.dnb.nl/en/general-news/news-2020/dnb-temporarily-lowers-bank-buffer-requirements-to-support-lending/
[7] For further discussion see Covas, F. and B. Nelson (2020). A Few Observations on Professor Stein’s Remarks Last Week at the Brookings Institution. Bank Policy Institute. https://bpi.com/wp-content/uploads/2020/06/A-Few-Observations-on-Professor-Steins-Remarks-Last-Week-at-the-Brookings-Institution.pdf.
[8] For example, in a briefing for the FOMC in June 2005, Fed staff concluded “To sum up, neither borrowers nor lenders appear particularly shaky. Indeed, the evidence points in the opposite direction: borrowers have large equity cushions, interest-only mortgages are not an especially sinister development, and financial institutions are quite healthy.” https://www.federalreserve.gov/monetarypolicy/files/FOMC20050630meeting.pdf p. 11.
I provided a similar assessment to the Board at a briefing on financial stability toward the end of 2006.
[9] Baer, G., B. Nelson and J. Newell (2017). Why Leveraged Lending Guidance is Far More Important, and Far More Misguided, Than Advertised. Bank Policy Institute, Nov. 14. https://bpi.com/why-leveraged-lending-guidance-is-far-more-important-and-far-more-misguided-than-advertised/
[10] Nelson, B. (2021). Turns Out Leveraged Loans Aren’t a Systemic Risk After All. Bank Policy Institute, February 8. https://bpi.com/wp-content/uploads/2021/02/Turns-Out-Leveraged-Loans-Arent-a-Systemic-Risk-After-All.pdf. Government Accountability Office (2020). Agencies Have Not Found Leveraged Lending to Significantly Threaten Stability but Remain Cautious Amid Pandemic. December. https://www.gao.gov/assets/gao-21-167.pdf.
[11] As described in “Turns Out Leveraged Loans Aren’t a Systemic Risk After All,” the then Director of the Financial Stability Division, Nellie Liang, and the then effective Governor over Supervision Daniel Tarullo stated in FOMC meetings whose transcripts have since been released that leveraged lending was not a financial stability risk. The episode highlights yet another risk of macroprudential supervision – that the vagueness of “financial stability risks” allows them to be invoked to achieve unrelated objectives.
[12] Nelson, W., and W. Passmore (2001). Pragmatic monitoring of financial stability. BIS Papers, (1), 367-84.
[13] Nelson, W. and R. Perli (2007). Selected indicators of financial stability. Risk Measurement and Systemic Risk, 4, 343-372.
[14] Adrian, T., D. Covitz and N. Liang (2015). Financial stability monitoring. Annual Review of Financial Economics, 7(1), 357-395.
[15] Dong B., F. Duarte, T. Eisenbach, and J. Vickery, (2018). Ten Years after the Crisis, Is the Banking System Safer? Federal Reserve Bank of New York Liberty Street Economics (blog), November 14. https://libertystreeteconomics.newyorkfed.org/2018/11/ten-years-after-the-crisis-is-the-banking-system-safer.html.
[16] Kovner, A. (2024). The Financial Stability Outlook, remarks at the Forecasters Club of New York, New York City, Feb. 28. https://www.newyorkfed.org/newsevents/speeches/2024/kov240228
[17] Board of Governors of the Federal Reserve System (2022). Financial Stability Report, November. https://www.federalreserve.gov/publications/files/financial-stability-report-20221104.pdf
[18] Autonomous, “U.S. Banks, Fair to Focus on Fair Value,” 16 Nov. 2022.
[19] In the United States, the capital conservation buffer has been replaced with the stress capital buffer. The SCB is the maximum of the CCP (2½ percentage points) and the projected decline in the bank’s capital ratio in the annual stress test. If, as discussed, the CCyB were to replace the CCB, it would replace the floor on the SCB.
[20] In an October 2022 study on buffer usability, the Basel Committee on Banking stated: “A CCyB release could be accompanied by disincentives to distribute capital and by incentives to employ it for the intended objectives (eg supply of credit during stress and/or timely write-downs of non-performing loans). System-wide dividends and other pay-out restrictions can complement the CCyB by increasing bank resilience and loan supply.” p.21 https://www.bis.org/bcbs/publ/d542.pdf
[21] Nelson, B. (2022), Potential Costs of Banning Capital Distributions by Banks in Bad Times. Bank Policy Institute. June 16. https://bpi.com/wp-content/uploads/2022/06/Potential-Costs-Of-Banning-Capital-Distributions-By-Banks-In-Bad-Times.pdf
[22] See Nelson, B. (2016). Countercyclical Capital Buffer: Doing the Math. Bank Policy Institute. November 4. https://bpi.com/countercyclical-capital-buffer-doing-the-math/
[23] ECB and ESRB (2025) “Using the Countercyclical Capital Buffer to Build Resilience Early in the Cycle” p. 35. https://www.ecb.europa.eu/pub/pdf/other/ecb.jointreportecbesrb202501~f5cf374d79.en.pdf
[24] In the early 2000s, the staff of the Federal Reserve Board created what was called “the Survival Binder.” The first half of the binder was a list of things that could go wrong that would imperil financial stability, such as the failure of a housing GSE or a terrorist attack, and the immediate steps the Fed should take in response. The second half consisted of tools the Fed could use in a crisis such as emergency “13(3)” lending and the systemic risk exception to least cost resolution.
I have requested that the document be made public under the Freedom of Information Act, but the Fed has denied the request, stating that the 20-year-old document cannot be released because “…the document consists of recommendations for addressing various potential crises and is, by its very nature, deliberative and predecisional.” The document is different from the New York Fed’s “Doomsday Book,” which has been released under FOIA. If the survival binder were released, it would allow the public to better judge the extent to which financial stability risks can be foreseen.
