The U.S. and the world economy still haven’t fully recovered from the global financial crisis that provoked the worst recession since the Great Depression. Nonetheless, a lot of effort already has gone into reducing the risk that we put the world through something like this again. In the US and elsewhere we’ve renovated our regulatory and supervisory structures in charge of making the financial system more resilient. We’ve made progress, but the job isn’t done yet. In particular, I’ll be concentrating on the institutions we’ve created to deal with risk to the entire financial system—so-called macroprudential regulation. And I’ll be contrasting the systems put in place in the UK, where I have been privileged to become a member of the macroprudential regulator—the Financial Policy Committee (FPC) – and the US where the Dodd-Frank law created the Financial Stability Oversight Council (FSOC). And I’ll be drawing some lessons you might consider as you shape your macroprudential institutions in China. 
Many failures in the private and public sectors contributed to the buildup of risks in the financial market and institutions that led to the financial crisis. But one was that no entity had a clear mandate to look at the financial system as a whole and the authority to act to mitigate the risks it saw developing; no one was responsible for identifying and taking steps to protect against the dangerous elements of procyclicality in the buildup of leverage, maturity transformation and complex interdependencies and interconnections in good times that left risk higher and more concentrated than private entities realized. Moreover, the resulting exposures to rare tail events—like a national decline in housing prices—left the economy and many innocent parties subject to externalities, or effects beyond those that could be priced into the market by private parties even with more complete knowledge and understanding.
Among the many responses to the crisis has been the establishment of groups to monitor systemic risk in the financial system and to do something to build resilience of that system to problems that might develop—macroprudential regulation. In US, the Financial Stability Oversight Council, FSOC, is a committee of microprudential regulators under the Chairmanship of the Secretary of the Treasury. In the UK, the Financial Policy Committee is a new committee established as part of a comprehensive restructuring of regulation. I am an external member of the FPC, that is, a member without a formal relationship to the Bank of England or to the microprudential regulators on the committee.
I am not an expert in the Chinese economy and financial system. But it does seem to me that you are facing some very necessary but very challenging developments from the perspective of preserving financial stability as you deregulate the financial system and liberalize the capital account. Deregulation and liberalization are necessary to improve the allocation of capital toward more efficient and more innovative firms, afford households and businesses more and sounder alternatives for both borrowing and lending, reduce the role of regulatory arbitrage in allocating credit and risk, and gradually integrate the RMB into the global financial system consistent with China’s growing importance in the global economy. All these changes must support and facilitate the difficult transitions in the real economy from reliance on credit-fueled capital investment and on exports toward more consumption.
But preserving financial stability as this process unfolds will require close cooperation among all the regulators and supervisors and the monetary authorities, and it will require all of them to view their own actions in the context of the stability of the overall system. As I’ll be relating, the U.S. hasn’t always been as successful as we might have been in preserving economic stability as markets were deregulated and in embedding a financial stability ethos in all our regulators after the crisis. I’ll try to point out the problems and draw some lessons for China from the UK and US experience. It’s especially important to avoid a crisis as the liberalization process unfolds because containing crises usually entails more state support and increased moral hazard, which would interrupt or even reverse the rollback of state involvement that will be important to empowering private markets and improving resource allocation and public welfare in China.
What do we want from macroprudential authorities .
We want those authorities to be able to:
Identify legitimate risks to financial stability. This requires analytical capability and access to information. Where the information is not immediately available, the authorities need to have the ability to get it from other agencies and from the private sector.
Be willing and able to act on the risks they identify in a timely way. Such actions might be in the form of direct instructions to financial institutions or supervisors; or they might be in the form of recommendations to microprudential supervisors. In the latter case, it’s important that regulatory structures allow the recommendations to be accepted or rejected quickly and implemented expeditiously.
Actions or recommendations can be aimed at strengthening the structure of the system, the ability of institutions and markets to withstand the unexpected and to behave in less procyclical ways, that is, less prone to exacerbate the ups and downs of the economy. Macroprudential authorities should also have the ability and willingness to take countercyclical actions, to tighten when good times tend to lead to added risk and to ease when downturns threaten the supply of credit.
And we want the authorities to take responsibility for patrolling the regulatory perimeter—for identifying potential risks that can’t be dealt with under the existing legal structure and to make recommendations to the executive and legislative bodies for new legislation to fill the gaps.
Acting on risks will require a considerable degree of independence from short-term political pressures, including those exercised by regulated firms themselves directly on the regulators and indirectly through their influence on the state.
Be able to interact productively with the microprudential and monetary policy authorities. For the most part, macroprudential policies will involve using the microprudential tools—like capital and liquidity requirements—but with a macroprudential overlay that takes account of the entire system. As such, those tools generally will be implemented by the microprudential authorities, who need to understand and buy into the financial stability objective. The cooperation of those authorities will also be necessary for analyzing the strengths and weaknesses of the system and for getting data and other information to deepen that analysis where necessary.
The conduct of macroprudential and monetary policies each may have important effects on meeting the objectives of the other. The macroprudential and monetary policy authorities each need to understand the strategies the other is pursuing and how they might evolve. And under some circumstances they may want to work in tandem, with explicit understandings about the circumstances under which reliance on macroprudential tools to protect financial stability will be paramount and the circumstances under which monetary policy may need to be adjusted, perhaps because macroprudential tools are not adequate to address building risks.
Weigh the costs and benefits of proposed actions appropriately. Precise, quantitative cost benefit analysis is unlikely to be possible for proposed macroprudential actions, but, as in all economic decisions, a cost-benefit mindset is required to enhance public welfare. There are considerable benefits from avoiding crisis and collateral damage to the economy, from avoiding the need for state support to limit any such damage and from reducing the moral hazard of public support and erosion of market discipline. But there are potential costs as well if macroprudential regulation raises the costs of intermediation unnecessarily or slows the liberalization of markets.
Macroprudential regulation in the United Kingdom: The Financial Policy Committee.
The UK approached building an institution to meet these challenges for macroprudential policy as an element in a root and branch restructuring of the whole financial regulatory structure. The system is built around three new entities. The FPC; the Financial Conduct Authority (FCA), with responsibility for consumer protection, market functioning and the supervision of some financial firms, such as asset managers; and the Prudential Regulatory Authority (PRA), the microprudential regulator and supervisor for banks, other depository institutions, insurance companies, and major investment firms. In addition, the Bank of England plays a major role in financial stability in the UK, housing and staffing the FPC and PRA, but also with broad responsibility for analysis and policy to promote stability. Although they are within the BoE, the FPC and PRA have separate boards or committees ultimately responsible for the policies to meet their objectives, each with external members, like myself. Our job is to bring our expertise to bear in formulating policies but crucially also to bring diverse views to the table.
The Financial Policy Committee has 10 members: the Governor of the Bank of England, the chief executives of the PRA and FCA, three others from the Bank, and four external members. A Treasury observer attends FPC meetings as a nonvoting member. Our primary objective is to identify, monitor, and take action to remove or reduce systemic risks with a view to protecting and enhancing the resilience of the UK financial system. The FPC has a secondary objective to support the economic policy of the government, including for growth. To achieve these objectives we can make recommendations to just about anyone—other regulators, the Treasury, private industry associations. Recommendations to supervisory agencies can be on a “comply or explain” basis. Implementing any recommended changes in regulation would be subject to the usual public comment period and be ultimately under the control of the agency or organization. But we also have “powers of direction” –direct orders we can give for certain defined changes in regulation. To date, those powers encompass the countercyclical capital buffer under Basel III and sectoral capital requirements pertaining to categories of real estate lending. We also asked for direction powers over the leverage ratio; although this was not granted initially, we are now working on a framework for the role of the leverage ratio within capital requirements to be published for comment, which should rekindle the discussion of the FPC’s authority in this area. And we have noted that we will consider whether to ask for direction powers for bank liquidity requirements and for margin requirements in securities financing markets once international standards have been agreed.
The UK system appears to be well-designed to meet the objectives outlined above. Each of the Financial Policy Committee, the two supervisory agencies and the Monetary Policy Committee with its inflation objective has a well-defined primary objective related to aspects of the overall objective of macroeconomic and financial stability. And each Committee has tools adapted to its objective. Each is populated by personnel with relevant expertise. Overlap among committee membership enhances communication and understanding of policy interactions, while external members guard against group think. The microprudential regulators on the FPC embrace the financial stability mandate of the committee and they have been able to help shape the macroprudential overlay to assure that it can be effectively implemented.
In practice, the FPC has concentrated on a few issues. First and foremost has been to encourage the rebuilding of the capital of the UK banking system, building it after the crisis and holding the system to a higher, more resilient standard than the one that had proven so inadequate in 2007 and 2008. Working with the microprudential regulators, we required banks to buildup capital to ensure it was adequate against a more forward looking recognition of losses than required by accounting standards, a more realistic assessment of costs for conduct redress and a more prudent approach to risk weighting. Currently we are working with the PRA to initiate simultaneous stress tests of bank capital built around a macroeconomic stress scenario.
A second major thrust has been around information. The FPC has made requests to the PRA and FCA for added information that would help us judge systemic risks. And we have worked with the microprudential regulators and the British Bankers Association to increase the amount and usefulness of information released publically so that markets can better judge financial health, raising the odds on market discipline reinforcing financial stability.
Finally, the FPC has been called on to bring its perspective to bear on certain policies of other entities, with the potential for inducing modifications of those policies where they might pose threats to financial stability. For example, in 2013 the UK government initiated special programs to stimulate housing construction and to help first time home buyers. Using its judgment on the effect of these programs on financial stability, the FPC is expected to opine periodically on whether changes in the terms of these programs would be desirable, and it is expected to give its views on the overall impact of the programs on financial stability when it is time to consider renewing them after three years.
With respect to monetary policy, when the MPC initiated the first phase of its forward guidance about the persistence of very low interest rates, it gave a financial stability “knock out” to the FPC. If the FPC found that the low rates threatened financial stability and it had unsuccessfully tried to deal with these risks with available macroprudential approaches, it would notify the MPC and that notification would “knock out” the promise to wait until the unemployment rate had at least slipped below 7 percent. Such a notification would not automatically entail a rise in the bank rate, but it would sharpen the discussion about the relative costs and benefits of the stance of monetary policy. Although, with the unemployment rate passing through 7, percent that first phase has been passed and the financial stability knock out no longer applies, the division of responsibilities between the FPC and MPC continues; the FPC will continue to monitor the financial stability consequences of very low rates and use its macroprudential tools to counter any emerging problems, allowing the MPC to concentrate for now on achieving its primary and secondary goals for inflation and growth.
What’s appealing about this use of the FPC is that it brings an independent view on the effects of a given policy on financial stability from a specialist group—and one that does not have an ownership interest in the policy itself. The FPC has tools targeted specifically to financial stability. In addition, the communication between the FPC and the Treasury and MPC is open and transparent.
I don’t want to hold up the UK institutional arrangement as ideal for all countries in all circumstances. It vests a lot of authority in one institution—the Bank of England, which has a good deal of independence. This raises questions around effective accountability, already a difficult issue for financial stability. And it will be challenging to be assured that alternative perspectives are being brought to bear on difficult issues when so much is in the Bank of England. It is early days; I’m optimistic, but we don’t know how this will work out over time.
Macroprudential regulation in the US: the Financial Stability Oversight Council.
Overview. FSOC has the same objectives as the FPC: identify systemic risks and find ways to mitigate them through action or through recommendations to regulatory agencies or to the Congress and the administration. But unlike in the UK, the Congress, in Dodd-Frank, left the current regulatory structure in place and grafted the macroprudential authority on top of it. Moreover, it’s a very fragmented regulatory structure in which multiple agencies have responsibilities for the same or closely related entities and markets.
Changes were made to the mandates of a number of existing agencies to take care of problems identified as having contributed to the build up to the crisis or as potential weaknesses in the financial system—for example in the derivative markets. And a new regulator was created to protect consumers as they interact with the financial sector. But for the system as a whole, of the existing agencies, only the Federal Reserve was given broad new responsibility, and even that was focused mainly on systemically important bank holding companies and financial market utilities, with the addition of, so far a few, nonbank companies designated as systemically important by the FSOC and therefore subject to special oversight.
So monitoring and devising actions to deal with systemic risk rests importantly with FSOC. FSOC has ten voting members: eight regulatory agency heads, one independent member with expertise in insurance, and the Secretary of the Treasury, who chairs the Council. Importantly, these are the heads of the agencies, not the agencies themselves and the heads can’t always speak for the whole agency when it is run by a board. FSOC also has 5 nonvoting members, including two more from the insurance sector.
FSOC can and has identified a variety of risks to financial stability in its annual report. It has some powers to act on those risks. It can identify systemically important financial institutions for regulation by the Federal Reserve; it can make recommendations, and when it recommends more stringent regulation and heightened safeguards to regulatory agencies it can do so on a comply or explain basis; it can make recommendations to the Congress on problems it sees with the perimeter or reach of regulation. But it does not have the authority to act on its own—except for the SIFI designation. For example the countercyclical capital buffer is under control of the Federal Reserve for bank holding companies and of the Comptroller of the Currency for national banks.
Without question, FSOC is a step forward in the US in dealing with systemic issues in the financial markets. Most macroprudential tools work by tweaking microprudential tools and FSOC has all the heads of all the microprudential regulators sitting on it for discussions of systemic issues and how they might be dealt with. This should foster understanding of how each agency’s actions can affect stability. Moreover, by all reports, FSOC has helped to foster coordination and cooperation among these regulators—much more so than existed in the past. However, the early years of FSOC have also revealed structural issues and shortfalls in its ability to accomplish the objectives we identified for macroprudential regulation, with possible implications for the structure of Chinese macroprudential regulation.
First, it is composed of many independent regulators, each of whom has mandates focused on particular institutions or markets. The legislation for each agency limits its objectives and the reach of its regulations. For example, the Securities and Exchange Commission is tasked with protecting investors and maintaining fair, orderly, and efficient securities markets, with a focus on getting adequate information to all investors at the same time. Achieving its objectives may well be necessary for financial stability, but they are not sufficient for the system, even in the areas under its jurisdiction. No agency has an explicit objective for maintaining financial stability—for taking into account the macroprudential add-on to microprudential oversight. The Federal Reserve comes closest, but there are limits on the scope of its authority. Macroprudential regulation will require cooperation and coordination across many agencies in the context of shared goals for the system. FSOC has made progress but isn’t yet as effective as it needs to be in the balkanized US regulatory system.
Second, there are gaps in regulation among FSOC and its regulators that could interfere with efforts to reduce systemic risk. For example, pieces of the shadow banking system—the chain of securitization vehicles, securities financing and short-term funding through commercial paper and repurchase agreements outside the commercial banking sector--can be regulated as they come in contact with the regulated sector, but not on their own.
Third, a consequence of regulatory fragmentation is the difficulty of obtaining and using data from a variety of agencies, cutting across institutions and markets. Data sharing is hard and complex, requiring strong protections for confidentiality of information obtained by a particular agency for supervisory purposes. But sharing data will help to spot problems developing, help inform the oversight of closely related agencies, and assist in identifying data gaps or inconsistencies.
Fourth, FSOC is likely to face particularly difficult challenges in the area of countercyclical macroprudential policy. The institutional makeup of FSOC and its focus on systemically important institutions and on broad weaknesses in markets and institutions seem better suited to building through-the-cycle resilience than to leaning against developing asset mispricing and growing leverage and maturity mismatch in the expansionary phase of a cycle. Some degree of independence from short-term political pressures will be required to try to calm the party down as asset prices rise and credit expands and to reduce the severity of the hangover when the party is over. Yet, the FSOC is chaired by the Secretary of the Treasury who represents the administration in power. And the interaction of countercyclical macroprudential policy and monetary policy is particularly difficult with the Secretary of the Treasury chairing FSOC.
Fifth, FSOC itself has very limited tools to deal with structural or countercyclical macroprudential risks, and recommendations to other regulatory agencies might take considerable time to be implemented, coming into effect when they could be ineffective or even procyclical.
Some possible lessons for Macroprudential regulation in China from the US experience.
I’m going to draw these lessons not only from the issues with FSOC I just identified, but also from the effect on financial stability in the U.S. of deregulation of deposit rates and development of alternative sources of intermediation, as these are high on the agenda for liberalization in China. As I noted at the beginning of my talk, deregulation and liberalization only heighten the need for strong macroprudential oversight.
- Deposit rate deregulation is necessary for financial stability and essential to promote efficient allocation of resources, but the process of deregulation can also threaten stability. The U.S. deregulated deposit interest rates in discrete and limited steps in the 1970s and across the board in the early 1980s. As in China, we found that when deposit rate ceilings were in place efforts by savers to realize more market-based rates of return and efforts by depositories to retain these funds resulted in disintermediation and distortions. But deregulation wasn’t smooth. Some depository institutions—especially the S&Ls—couldn’t afford to pay higher rates because their asset portfolios were locked into longer-term assets with relatively low yields. We made the mistake of letting them diversify into other line of businesses to bolster profitability; those businesses were even riskier and ultimately increased the cost of remediation. The implosion of the S&Ls, together with problems in the commercial real estate market that hit commercial banks, resulted in a squeeze on credit availability that impeded the growth of the economy in the early 1990s.
Deposit rate deregulation in China will require careful micro and macroprudential oversight of the institutions and the system. Depository institutions will need the capital and the risk management skills to operate in a more market-based environment. Weak institutions that can’t afford to pay market-based rates on their sources of funds or can’t meet the new higher standards for capital and management should be resolved early or they will be tempted to engage in actions that will only make their eventual resolution more expensive.
- Alternative sources of intermediation services make the financial system more resilient, but they can also give rise to new sources of vulnerability. The U.S. has a more highly diversified financial system than many other countries where universal banks dominate. Access to credit through securities and securitization markets has kept credit flowing even when, as in the late 1980s and early 1990s, the banking system was troubled. But a number of forces came together in the 2000s to support the origination of increasingly complex and opaque securitization instruments that were highly vulnerable to unexpected economic developments, like a decline in house prices. These instruments embodied considerable leverage, and far too often holdings of long-term assets were being financed with short-term liabilities. When house prices began to fall, doubts emerged about the underlying value of these instruments, but it was difficult for market participants to know who was exposed, and the result was contagion throughout the financial system with very adverse economic consequences.
It is important to encourage the development of securities and securitization markets, but it is equally important to make sure the instruments and their risks are transparent and understood by market participants, well supported by the necessary capital and liquidity, and that the incentives of issuers and investors are aligned with society’s broader financial stability objectives.
Although deregulation and liberalization should reduce incentives for many forms of regulatory arbitrage, others may persist or even spring up anew in response to the sequencing of deregulation or the residual regulatory burden on some markets and institutions. The macroprudential authority must patrol the regulatory perimeter, remaining alert for the possibility that alternative markets are simply a product of regulatory arbitrage rather than soundly based diversification of intermediation channels and alert to the potential for systemic risks to migrate or be amplified in alternative markets, even when there is a good rationale for their utilization.
- The sting is in the tail. Small to moderate shifts in expected returns aren’t going to endanger financial stability, but when outcomes are drawn from the tail of the distribution, most participants won’t be prepared and that’s when crises occur. That’s why stress tests of capital and liquidity are so important as a macroprudential tool. The tests must be stressful—reach well into the tail of potential outcomes—and they must be transparent about the stresses and about the results. Transparency will enable market participants to judge the stability of individual institutions and bring market discipline to bear to reinforce financial stability. Transparency will also enable the public and other elements of the government to judge the performance of the macroprudential regulator and the strength of the system.
- Macroprudential regulation will need to be strong and comprehensive and that in turn will require buy in by all the microprudential regulators. Macroprudential regulation works in large part through microprudential regulation, but with an extra overlay to take account of systemic concerns and externalities. The cooperation and understanding of the microprudential regulators is therefore essential to effective macroprudential regulation. They need to take systemic risks seriously and see where their actions fit into the safety of the overall system. And they need to be ready to share data with other authorities that have an interest in building the resilience of the system.
When addressing this problem for the U.S. I suggested that every member agency in FSOC be given an explicit objective to protect the stability of the financial system and that each be required to share data in response to well-supported requests from other regulators.  I don’t know whether this approach would work in China, but I suspect it would help if every agency involved in regulation was required to cooperate in protecting the system.
- The organizational structure of macroprudential regulation must be adapted to the particular circumstances of each jurisdiction, but the macroprudential regulator should include the central bank in a major role and the macroprudential regulator should be able to take unpopular actions. History, the political system, and financial structures that evolved over time may make it not possible to undertake the sort of complete re-organization of regulation that included creation and integration of the macroprudential regulator in the UK. But however the systemic authority is put together, two requirements for effective macroprudential are important to highlight. First, it must include a major role for the central bank. The central bank is the liquidity backstop for the financial system as the lender of last resort; its operations and responsibilities put it in close contact with a broad array of institutions and markets; its responsibility for macroeconomic stability gives it the expertise and incentive to see how the financial system interacts with the economy. Second, the systemic authority will need enough independence from short-term political pressures to be able to act to correct structural weaknesses and to lean against rapid credit expansions, especially when times are good and risks appear low.
My own recommendation for FSOC was to keep the Federal Reserve in a central role, but to enhance independence of FSOC by replacing the Secretary of the Treasury as its head with an independent leader unassociated with any particular agency and with an arms-length relationship to the political process.
- Macroprudential regulators should have tools they can exercise directly—or at a minimum be able to make recommendations to others that carry considerable weight and can be implemented quickly if necessary. When the macroprudential authorities see a threat to financial stability—whether structural or one building up with rising asset prices or rapid increases in credit—they need to have a tool kit of actions they can take to deal with it. Ideally, the ideas for and implementation of those actions can emerge largely by consensus out of discussions with and commitment to financial stability of the microprudential authorities. But the macroprudential regulator may also need to be able to act quickly itself or to elicit an especially speedy response from the microprudential regulators under some circumstances, such as when risks are especially threatening or building fast. In the UK, the FPC has control over the countercyclical capital buffer under Basel III and also sectoral capital requirements. FSOC’s tools are more limited, and my recommendation for FSOC was that it be able to fast track some recommendations to the microprudential regulators—reduce the time to implementation--and that FSOC be required to make recommendations on a clearly macroprudential instrument, the countercyclical capital buffer.
I don’t know what the right formula is for the Chinese system, but some special powers for the macroprudential regulator acting in that capacity would help to align responsibility and authority, and in so doing increase accountability and the ability to protect financial stability.
I believe deregulation of the Chinese financial system will increase both the stability and the productivity of your economy over time. But with those rewards come risks. You will need a strong, effective, oversight of the financial system, with considerable emphasis on the stability of the whole system, as you free institutions and people to allocate savings, capital, and risk based on market signals. I’ve tried to draw out some lessons from my experience in the U.S. and U.K. about what those risks might be and how to structure systemic oversight to raise the odds on effective protection against the downsides. I hope this has been helpful; you have a lot of interested observers, myself very much included, as you go forward.
 See my Kennedy School talk referenced in footnote 1 for my recomeendatons for FSOC.