The Supervisor’s Role in Identifying Emerging Financial Instability: Leveraging Confidential Supervisory Information to Identify Risk in the Financial System

It's truly an honor to speak here at the Global Risk Regulation Summit organized by RiskMinds.  As I'm sure all of you would agree, these conferences are great opportunities to learn about current and developing ideas and perspectives among all participants in our financial sector, and also a great opportunity to meet new contacts and reconnect with colleagues and fellow practitioners in the industry.  Also important is the global nature of this conference.  As the interconnectedness of the financial sector increases, our world becomes smaller, and conferences such as these that bring together industry participants from around the world are particularly valuable.  I am delighted to be part of such a conference.  And in particular, I am pleased to speak on the topic of the future direction of regulation, as this is a topic that has implications to both supervisors as well as to supervised firms, and the broader financial system.

 As the title of this session implies, I would like to provide you with my perspective on the role of supervisors in identifying emerging financial instability.  In my remarks today, I will focus on three primary themes.  First, I will discuss the evolving role of supervisors and the need for supervisors to take a broader perspective on the scope of our responsibilities.  I will then discuss an aspect of our role as supervisors that provides us with a unique advantage in identifying systemic risk. That is, the use of confidential supervisory information in the tools we develop. And I will provide an example of such a tool that was developed at the Federal Reserve Bank of Cleveland.  And lastly, I will discuss the importance of global collaboration among supervisors as a key to effectively supervising the overall financial system and preserving financial stability.  Of course, the views I express here are my own, and do not necessarily represent the views of the Federal Reserve Bank of Cleveland or the Federal Reserve System.

The Evolving Role of Supervision

For the last several years since the financial crisis, most speeches and presentations at conferences such as this one have begun with citing the financial crisis as the basis or impetus for the research work or perspective.  In my case, as I think about the evolving role of supervisors in the current environment, the financial crisis merely confirmed the direction our role as supervisors needed to evolve.  The true impetus was a strategic planning initiative of my Supervision and Regulation function at the Federal Reserve Bank of Cleveland in 2007 – prior to the financial crisis.

At that time, a key strategic initiative we were focusing on related to the enterprise-wide risk management programs of the financial firms we supervise.  Specifically, we were discussing how we, as supervisors, should go about ensuring that financial firms had effective enterprise-wide risk management programs to help them identify and mitigate risks.  Incidentally, this is an initiative that is still important to us, and one that we continue to make progress on.  But it was during these discussions of risk management that we decided to turn the mirror on ourselves, so to speak.  In other words, we asked ourselves, "As part of the central bank of the United States, if we were the risk managers, what is our enterprise?"  Well, we quickly came to the conclusion that, as a central bank, our enterprise is not limited to just the financial firms  we had specific responsibility for supervising.  It is much broader than that.  As a supervision function that is part of the central bank, our enterprise is actually the financial system, and it is our responsibility to identify and manage the risk in this enterprise that is the financial system.

This is not a change or revelation that has come overnight.  On the contrary, this is an evolution that has progressed over a number of years, dating back to even when I began my career over 30 years ago as a bank examiner.  At that time, the term "bank examiner" was very fitting in that it accurately described our role and the scope of our activities.  We examined entities that were specifically banks, which engaged in very specific, traditional banking activities.  This included deposit taking, lending to primarily local customers, and limited investment activities.  As such, our own activities were similarly limited, in that we examined these banks as of a particular point in time.  At the conclusion of the examination, examiners would issue a report on the condition of the bank as of that date.  Unless significant issues were identified, the next examination of that bank would not occur for another year or more, and this was adequate given the limited activities and risks of the banks.

With the introduction of technology as a catalyst, the financial industry began to change, and the pace of this change increased.  Banks would introduce new products and services, and perhaps even expand their geographical presence, at a much more rapid pace.  Changes occurred so frequently that it was no longer adequate to wait from one examination to the next to assess the condition of a particular bank.  Offsite monitoring techniques were developed to augment the point-in-time examinations.  This ongoing "supervision" of banks was characterized by more frequent and sophisticated analysis of financial reports submitted by the banks.  At larger and more complex banks, teams of examiners were dedicated to the ongoing supervision of specific firms.  This ongoing supervision provided for more timely identification of changes in bank strategies and activities that could result in greater risk.  The role of the bank examiner had evolved to that of a "bank supervisor" who not only examined the banks, but also monitored them on an ongoing basis.

Of course, the pace of change in the banking industry did not abate.  On the contrary, the pace of change only increased, and was fueled in part by a piece of legislation in the United States known as the Gramm-Leach-Bliley Act in 1999.  The Gramm-Leach-Bliley Act empowered banking organizations to engage in a wider array of financial activities.  The line between banks and nonbank financial firms began to blur, and bank supervisors had to adapt accordingly.  Bank supervisors were required to view these entities differently and assess a broader range of potential risks.  By necessity, the bank supervisor had evolved into a "financial institution supervisor"--still engaged in continuous supervision of a single firm, but with the recognition that that single firm was no longer a traditional bank.

Fast-forward to 2007, when we in Cleveland determined through our strategic planning self-assessment that we needed to continue to engage in ongoing supervision, but that our scope included not just financial firms, but also the financial system as a whole – our enterprise.  This view was confirmed by the financial crisis of 2008.  Before the financial crisis, supervisors focused primarily on the safety and soundness of the individual entities for which they were responsible.  Risks that spilled over or were transferred to the rest of the financial system were not closely monitored.  Unfortunately, as the financial crisis revealed, it turned out that these often made their way back onto the books of the banks.

The increased interconnectedness of the financial system facilitated risks from one sector of the financial system to spill over into other sectors of the financial system.  As such, it became abundantly clear that a strictly entity-based approach to supervision was not enough to ensure financial stability.  In addition to identifying and monitoring risk within an individual firm, supervisors must also pay attention to risks that exist more broadly in the financial system, many of which may ultimately find their way back onto the books in very creative forms.  In this way, financial institution supervisors must evolve into "financial system supervisors"--engaged in the ongoing supervision of individual financial institutions and focused on the risks in the financial system overall.

Financial System Supervisors:  A New Approach to Supervision

To effectively supervise the financial system, two dimensions of supervision must be expanded: scope and time.

Scope refers to the range of supervised entities, instruments, and practices.  The evolution of the financial system, particularly the interconnectedness among financial institutions, requires a more collective view and aggregate assessment of risks across firms.

An important tool to gain this aggregate assessment is called a horizontal review. Supervisors use horizontal reviews to provide a view of risk across a specified population of firms.  First, they identify a specific activity or area of risk.  Then, reviews at each firm are conducted simultaneously and the results are aggregated to provide supervisors with a “macroprudential” view of the activity or risk exposure.  This collective view allows supervisors to determine how the activity or risk across this population of firms may broadly affect the stability of the financial system.

As an example, in 2010, supervisors with the Federal Reserve looked at incentive compensation practices at the largest US banking organizations.  The review assessed incentive or bonus programs and looked for those programs that encouraged employees to take risks on behalf of the banking organization that were beyond the risk tolerance of the firm, or would result in substantial risk to the financial system.

The results from the review helped supervisors develop principles for incentive compensation practices that promote an appropriate consideration of risk and the alignment of incentive compensation practices with safe and sound practices by financial firms.  Horizontal reviews have also been conducted in the areas of capital planning and adequacy, mortgage servicing practices, and other operational areas.  Each review resulted in an aggregate view of risks to the financial system from these practices and supervisory guidance for firms engaging in these activities.

Beyond horizontal reviews, supervisors are looking closely at less regulated sectors of the financial system, what’s known in the US as the “shadow banking system,” which may introduce a host of instruments and activities that pose risk to the financial system. Even though some of them may not originate or reside in banking organizations supervised by the Federal Reserve or other federal regulatory agencies, shadow banking activities can introduce risks to the financial system that may ultimately have an impact on banking organizations. 

Time is the other dimension where supervisory approaches have evolved.  Traditionally, supervisors judged the condition of a financial institution based upon information as of a particular date.  This is useful, but only to a point.  At best, it provides a rearview mirror assessment of the firm.   A better approach includes a forward-looking perspective to understand how well an institution might hold up in any number of economic scenarios.

Toward that end, supervisors increasingly use simulation models to replicate the inter-relations among various balance sheet and income statement accounts of a financial institution.  Assumptions related to risk exposure, earnings performance, expense levels, and other factors are used as bases for these models, and the outputs are factored into the assessment of the overall condition of a financial firm.  The simulation models complement the historical trend and point-in-time perspectives of traditional supervisory approaches.

 A high-profile use of simulation techniques was first deployed in 2009, and was known as the Supervisory Capital Assessment Program, or SCAP.  Since then, and as a result of the effectiveness of the program, US supervisors have adopted the Comprehensive Capital Assessment Review program.  These reviews have included various scenarios and assumptions related to macroeconomic factors that would influence the performance of banking organizations.  The use of these simulation models provided supervisors with a better understanding of the potential loss exposure that would be experienced by financial firms in the future, should any of the assumed scenarios occur (such as very high unemployment, or a severe economic downturn).  Given the potential loss exposure revealed by the scenarios, firms were required to maintain levels of capital considered adequate to absorb the losses in those scenarios, and to continue to provide credit to borrowers as appropriate. In this way, simulation models have helped the financial system prepare for the worst in ways that rearview assessments simply couldn’t.  This forward-looking approach to analysis is an example of the evolution in risk assessment from the point-in-time reviews conducted by our bank examiners.

SAFE:  The Systemic Assessment of the Financial Environment

Another example of a forward-looking approach to risk identification and, in particular, risk in the financial system, is a forecasting model developed at the Federal Reserve Bank of Cleveland.  This forecasting model, called the Systemic Assessment of the Financial Environment, or SAFE, is an early warning system that allows for the identification of stress in the financial system six quarters in advance of the stress.  This design was specified in the model to allow adequate time to develop and implement appropriate policy measures to mitigate the stress once identified.  Of course, this specification is not unique to SAFE.  Other early warning models for systemic risk have been created that also allow time for the development and implementation of policy responses.

However, what is unique about SAFE is that it incorporates confidential supervisory information to inform its results.  This confidential supervisory information, or CSI, is derived from a combination of sources, ranging from examination results, proprietary tools of the Federal Reserve designed to anticipate ratings downgrades of banks, and even risk exposures of individual banks to particular factors.  This CSI is combined with publicly available information on financial firms as well as broader macroeconomic data to produce a forecast of financial stress based on risk, return, liquidity, and structure variables.  While I do not plan on elaborating on the specific details of this model, those who are interested can reference the November issue of the Journal of Banking and Finance1, which includes our paper detailing the model.

The Unique Added Value of Confidential Supervisory Information

My point in referencing our model is to emphasize the unique value that is added when confidential supervisory information is appropriately applied to forecasting models.  Both in-sample and out-of-sample testing confirm the effectiveness of the model in forecasting stress in the financial system.  In addition, testing of the model in the presence and in the absence of confidential supervisory information confirms the value of the CSI, where running the model incorporating the CSI consistently outperforms the model running without the CSI.

The ability to incorporate confidential supervisory information into financial stress forecasting models stems from the very conceptual origin of the model development.  Other systemic risk forecasting models that have been developed since the financial crisis have typically been developed in academia, or in the private sector, or even in the research arms of central banks – all areas where there is no access to confidential supervisory information.  SAFE, on the other hand, was developed in the Supervision & Regulation function of the Federal Reserve Bank of Cleveland, where CSI was accessible and carefully incorporated into the model.  This points to the value of having supervisors who think more broadly about their role as financial system supervisors involved in creating the tools necessary to fulfill this responsibility.  Supervisors around the world should view our role as not only preserving the stability of the individual entities we supervise; they should view our role as  preserving the stability of the financial system overall – the system of which the supervised entities are a part.  And in doing so, supervisors must incorporate the unique insights and knowledge we  have access to.  It is the confidential supervisory information that makes models like SAFE more robust and effective.

Having said that, I feel I must assure those in supervised financial firms that extreme care is taken to never compromise the confidential nature of supervisory information.  While there is clearly a strong benefit to leveraging CSI to the extent possible in order to more effectively forecast risk in the financial system, there are certainly risks in doing so as well. 

To address this, officials at the Federal Reserve Bank of Cleveland have worked with researchers at the US Treasury Department’s Office of Financial Research and with industry cryptography experts to explore various ways that confidential supervisory information and other sensitive information can safely be used and leveraged in forecasting models and other tools where clear benefits to the public and the financial system can be derived.  In addition to presenting alternative approaches to using confidential information, their research also addresses the need for balancing transparency and confidentiality when considering the use of this type of information.  Their research and findings can be found in their paper entitled, “Cryptography and the Economics of Supervisory Information:  Balancing Transparency and Confidentiality”2.

To effectively fulfill our role as financial system supervisors, the Federal Reserve Bank of Cleveland intends to further explore ways to use  confidential supervisory information. Our goal is to  at once preserve the confidential nature of this information while adding unique and significant value to the tools and measures that identify risks and preserve the stability of the financial system.

Global Collaboration Among Supervisors

While a systemic perspective and effective tools are necessary components, neither is sufficient  to effectively address risk to the financial system. A final element in a supervisor’s role in preserving financial stability is to collaborate effectively with our supervisory counterparts around the world.   The global and interconnected nature of the financial system and the entities that operate within it requires supervisors around the globe to work collaboratively to share information and insights in identifying emerging risks, and to take a coordinated approach to the development of macroprudential policies designed to address systemic risk.

A globally collaborative approach to developing macroprudential polices can ensure parity in treatment of internationally active financial firms, regardless of the country or continent in which a firm is based.  These macroprudential policies must also be developed with the recognition that financial markets are dynamic, and will adapt to the changes introduced to it, whether by an endogenous source or regulatory interjection.  Therefore, supervisors in every corner of the world should be prepared with a coordinated response to the manner in which the financial system will respond to the macroprudential policies.

In summary, let me reiterate that as the financial system evolves, so must the scope and perspective of the supervisors responsible for preserving the stability of the financial system.   Today’s supervisors must take a more holistic view of our responsibilities, and recognize that we must not only supervise individual entities, but we must also supervise the financial system in which those entities operate.  And in doing so, we must take a more forward-looking approach to risk identification and appropriately leverage all the information, data, and tools accessible to us.  And to the extent that many of the firms we supervise operate across the globe, we supervisors must take a collaborative approach to the identification of systemic risk and the development of macroprudential policies that address those risks to effectively preserve financial stability.

  1. Oet, M., Bianco, T., Gramlich, D., Ong, S., (November 2013), “SAFE:  An Early Warning System for Systemic Banking Risk” Journal of Banking and Finance, Issue 37, pp. 4510 – 4533
  2. Flood, M., Katz, J., Ong, S., & Smith, A., (September 2013), “Cryptography and the Economics of Supervisory Information:  Balancing Transparency and Confidentiality” Federal Reserve Bank of Cleveland, Working Paper #13-12