Of all the policymaking holes revealed by the financial crisis, few were as glaring as the dearth of reliable data on the health of the financial system. In response, Congress authorized creation of the Office of Financial Research (OFR), a unit of the Treasury Department. Its job is to find ways to improve the quality of financial data and to help policymakers perform more sophisticated analyses of the financial system.
The Senate confirmed Richard Berner as the OFR’s first director in January 2013. He comes to the job honestly, with a career steeped in financial stability and economic analysis, beginning with the Federal Reserve Board and including stops at a forecasting firm and Wall Street investment banks. Berner knows financial data and he knows we don’t have enough good data—yet.
Berner was a keynote speaker at the Federal Reserve Bank of Cleveland’s joint conference with the fOffice of Financial Research on financial stability analysis on May 30, 2013. Mark Sniderman, the Cleveland Fed’s executive vice president and chief policy officer, interviewed Berner at the Federal Reserve Board in Washington, DC. An edited transcript follows.
Sniderman: Thank you for your willingness to sit down and join me today.
Berner: Thank you for having me.
Sniderman: How did the Office of Financial Research (OFR), a brand new office, come to be a part of the Dodd–Frank Act, and what’s the OFR’s mission?
Berner: In the wake of the crisis, we saw that there were very significant gaps in our knowledge of how the financial system worked and in our ability to measure financial activity. To fill those gaps, Congress created the Financial Stability Oversight Council (FSOC) and the OFR, both within the Treasury. The OFR’s mission is to serve the needs of the FSOC in assessing and monitoring threats to financial stability and to improve the quality and scope of financial data for the FSOC and its member organizations.
Sniderman: This might not be such a great analogy, but the image of a roving linebacker comes to mind. The idea that there’s someone to spot areas of vulnerability within the financial system. Why didn’t Congress give existing agencies more authority to accomplish this?
Berner: It’s not a bad analogy. It’s not our mission to duplicate or replicate what other organizations in the FSOC are doing. Rather it is to fill in the gaps.
It’s exciting, exhilarating, and a huge challenge because you’re going where nobody’s gone before. And hopefully boldly, as the saying goes.
Sniderman: What is it like to have to start up an office—to start essentially from scratch—like you are doing with the OFR? How do you get organized? Where do you start?
Berner: It’s exciting, exhilarating, and a huge challenge because you’re going where nobody’s gone before. And hopefully boldly, as the saying goes. Boldness is necessary because you need to have a vision of what you want to accomplish. The congressional statute provides the framework for how the OFR should be set up, what its mission is, and so on. But within the lines of the statute, there’s a lot of room for interpretation and inserting a personal vision.
In consultation with the Treasury and, more broadly, in consultation with the other members of the FSOC, whose needs the OFR serves, we started to develop that vision for how it would function. It requires a culture of collaboration, engagement, and accountability. The OFR also needs trust within the organization and from the other agencies of the FSOC to function. And this all takes time. This is not something that can happen overnight. It’s not exactly like the pyramids, but nonetheless, it’s something that does require patience and persistence to really put together and stick to that vision, particularly when others have a lot of doubts about whether you’re going to get there.
I happened to run into Alice [Rivlin, first director of the Congressional Budget Office] shortly after I took this job and I asked, “What advice would you give me to set up the OFR, since you did the same thing, in effect, at CBO?” She said, “Just do it.” That was her advice. Worked for Nike, worked for her, works for me.
Sniderman: Are you trying to illustrate why it is ultimately in your partners’ best interest to work with you?
Berner: Absolutely. If you want to try to persuade people to collaborate with you, then there has to be something in it for them. And that’s true whether it’s collaborating on analytical projects where we have common interests or collaborating on projects that require us to share data with each other.
The typical question from a supervisor might be “What’s in it for me?” The answer might be “We can look across the entire financial system and there may be data in the not-too-distant future that you’re going to want and that we can share with you.” That’s part of the give-and-take that goes on in trying to forge those relationships and to build not only the trust but also the currency with which you can make those relationships work. That “win—win” idea is something that we try to infuse through building the OFR and building relationships with the FSOC.
Sniderman: Have you come across any helpful models, either in other U.S. industries or internationally? Or is this really a one-of-a-kind effort?
Berner: It is one of a kind, but there are other things or other entities that have some similarities to what we’re doing. Take, for example, the Congressional Budget Office (CBO). That was started from scratch in 1974 to serve the needs of Congress, to have a particular mandate, to do analysis of budgetary issues for the benefit of Congress, and to come up with quantitative metrics by which they could score particular bills and initiatives.
We don’t have to defend the policies. In fact, it’s our mandate to evaluate the policies.
Sniderman: That’s a particularly apt example. It is pretty remarkable that for about 40 years, CBO has clearly been able to maintain a reputation for bipartisan objectivity.
Berner: Exactly. And that goes back to the first director, Alice Rivlin. I happened to run into Alice shortly after I took this job and I asked, “What advice would you give me to set up the OFR, since you did the same thing, in effect, at CBO?” She said, “Just do it.” That was her advice. Worked for Nike, worked for her, works for me. Obviously there are mid-course corrections as you grow and there are nuances as you evolve; but having some core principles, values, and a foundation for the mission is really key.
It is interesting that Congress set us up as an office that doesn’t have any policy responsibility. The objectivity factor is really important because it separates us from the policymakers who do have that responsibility. We don’t have to defend the policies. In fact, it’s our mandate to evaluate the policies, particularly stress tests, but also to look at the effectiveness of the policy tools we’re developing to achieve financial stability. The objectivity and the integrity of our research are core principles that I tried to lay out from the start and that we want to really build into the culture of the organization.
Sniderman: Let’s move back to economics. There was a time when most economists said the best regulator was market discipline. Has the financial crisis proved those economists wrong? Has market discipline lost its standing in the world of big ideas, or is the opposite the case? Has the pendulum swung too far with too much regulation? How do we balance these ideas?
Market discipline is great when it works the right way, but you have to provide the right incentives.
Berner: In some ways, I think that dichotomy is a false one because market discipline really hinges on the incentives you put in front of market participants. If they have the wrong incentives to take on too much leverage or risk, or if our policies and regulations depress the price of risk, they’re going to want more of it. In the buildup to the financial crisis, there were incentives for people to write credit and liquidity puts (an agreement in which one firm agrees to provide a counterparty cash or equivalent), to write options, and to make funding really cheap, which all contributed to its severity.
One of the three key mandates of the FSOC is to restore market discipline, which implies it was lacking. The reason it was lacking was people were provided with the wrong incentives. Market discipline is great when it works the right way, but you have to provide the right incentives: to do the right thing; to self-insure; to recognize that if they’re going to take on more risk, they may take losses in accepting that risk; and that they understand how to analyze, to measure, and to manage that risk. Those are all very important factors in thinking about how market discipline should work.
There was an assumption that if we had economic stability, then financial stability would follow…But it didn’t quite work that way.
Sniderman: Has there been an evolution in how both economists and regulators think about financial stability?
Berner: There has. While there was some attention paid to examining the financial system as a whole prior to the crisis, after the crisis that focus is almost universally embraced. There was also an assumption that if we had economic stability, then financial stability would follow—like in the discussion of the Great Moderation, where reducing business-cycle volatility was thought of as a way to promote economic stability. But it didn’t quite work that way.
Not paying attention to financial stability needs and being complacent led people to assume we didn’t have to worry too much about it because we had a benign period. From my perspective, that’s exactly the time that we should be vigilant and watchful for signs that people might be taking on extra risk because they get paid to do it. There has been a lot of good research recently that supports the claim that when volatility is low and the price of risk is low, people are going to take on more risk, whether it’s through leverage or maturity transformation or other means. That’s when you start to see a buildup of risks in the system. So there’s been a change in thinking in that respect.
Sniderman: And maybe where the risk–return tradeoff was thought to be better?
Berner: Indeed. If you didn’t have capital requirements, if you didn’t have all these things that are restricting returns and that are building in cushions of safety, the returns are higher, but so are the risks. So how do we adapt the idea that we need to limit those risks and the build-up of those risks in markets? In the paper, we talked about the idea of minimum haircuts in repo markets, which are key for funding securities financing—so-called securities financing transactions. The implementation can be tricky, but it’s something worth looking at.
Sniderman: Recently, you’ve researched how a systemic regulator might operate to avoid a replay of the financial crisis period. One thing you and your co-authors identified was a problem with fire sales in magnifying the crisis. How do you think we could head those off in the future?
Berner: Fire sales result from two things. One is that investors get complacent and assets become overvalued, so people pile into assets because they’re paid to do that. And second, they may try to enhance returns using leverage that adds to risk and to the resulting unwind of that over-valuation. Eventually a shock comes along, particularly if leverage or risks that are asymmetric in nature have been used. Then, all of a sudden, the price of the assets goes down, the risks get unwound, participants might be subject to margin calls if they’ve embraced leverage significantly. All those things magnify the impact of the initial shock. That’s the origin of fire sales. So we must determine how to limit the incidence of fire sales or, on the other side of the balance sheet, runs.
When people are being paid to take on more risks, then it’s very tempting to ignore what those risks might be in the future because of the short-term gain.
Sniderman: Do you think that some of this is because in some cases people realize that they may have taken on excessive risk but think that they’ll be able to get out of the door faster than everybody else?
Berner: That’s part of it, sure. But I think it’s important that, if you work in financial markets, then it’s understood that people in theory look ahead and manage risks well, but you always have to remember what they’re being paid to do. When people are being paid to take on more risks, then it’s very tempting to ignore what those risks might be in the future because of the short-term gain. That’s a very important psychological or cultural factor that we always need to be mindful of.
So what do we need to do to make the system more resilient and to limit the buildup of risk? One of the things we wrote about in the piece that you’re referring to is the fact that there was a bank-centricity to our prudential regulations. We focused on capital requirements and we focused on liquidity requirements in banks or insured depository institutions. We didn’t focus so much on what was going on in markets, but of course the regulations we had in those institutions prompted the migration of financial activity outside of those institutions toward the so-called shadow banking system that, properly defined, involves the creation of money-like liabilities.
There’s the old adage that bankers make their best loans in the worst of times. And the flip side of that is they make their worst loans in the best of times. Same story.
Sniderman: A lot of stress indexes actually show relatively low financial stress. Is this precisely the time to pay closer attention?
Berner: That’s exactly right. Because of the work of Hyun Shin, Hughes-Rogers Professor of Economics at Princeton University, and others, we’ve discovered that low volatility reduces the price of risk, which is a key ingredient in option pricing. When volatility is low and spreads are narrow, it gives people incentive to take on more risk. That’s exactly when you should be watching for a buildup of leverage, a buildup of maturity transformation, a buildup of risk in other ways. That’s exactly when we need to be more vigilant. There’s the old adage that bankers make their best loans in the worst of times. And the flip side of that is they make their worst loans in the best of times. Same story.
Sniderman: At the OFR, you’ve laid out a framework for analyzing threats to the financial system. Can you point to one or two areas in the framework that are sometimes overlooked and why you think they’re so important?
Berner: One of our goals is to fill the gaps in the data and try to find out where we need to improve both their scope and their quality, so we can do the analysis in a way that has integrity and rigor. Before the financial crisis, we clearly couldn’t do that. We didn’t have enough high-quality data to do that analysis. So data and data standards are key areas for improvement. If you assume your parameters because of insufficient data, that may dictate what the outcome is. That’s important because often you draw those parameters from the historical pattern. History may not be replicated in the future. There may be regularities, but there may also be new things that arise and need to be accounted for.
One of the things that we try to emphasize, which is very difficult to deal with but which we need to think about hard, is that the financial system is constantly evolving and changing. It’s evolving and changing incentives in response to changes in regulation. Recent changes to regulation will likely change the way that people seek returns and how they manage their risk. These issues have some fundamental uncertainty attached to them, so they may be unknowable to some extent, but if we think hard about them and talk to market participants about what they’re doing, then we can understand better where the system is going, as opposed to where it’s been. I don’t think we can predict financial crises and I don’t think we know where the next shock is going to come from. Our goal is to make the system stronger, but in order to make the financial system stronger, we have to understand where those vulnerabilities are.
Sniderman: Let me focus on the issue of data, which is one of the topics at this conference. During your career, you collected data from private companies both as a government analyst and as a private-sector employee. Did you find that data were not always as easy to find within these private companies? Were there issues with these data?
Berner: Completely. The data were not well organized; the scope of the data was not sufficient to do the kind of analysis that was needed. People would quickly ask me to serve on risk committees or asset liability management committees in some organizations and we constantly needed more data, better data, more comprehensive data in order to do the job right.
Sniderman: Do you find that business units working on the same kinds of things didn’t necessarily share the same knowledge? Following the financial crisis, do you think some firms found they didn’t really quite understand themselves and what their own situation was?
Berner: Absolutely. And that’s why I stress in my current role how important it is to standardize data, so we can compare like with like when data come from different parts of an institution or from different parts of the financial system. So that is a big gap, and we’re trying to fill it.
Sniderman: Do you think that for some companies, the financial crisis was a wake-up call for them to improve their internal systems and data management, quite apart from any regulatory requirements?
Berner: I think that the crisis was certainly a wake-up call in that respect. And it was an important wake-up call in one other key respect: risk management practices. The position of chief risk officer (CRO) was created to help manage desk or business-unit activity, but they really didn’t look across the enterprise to manage risk across the whole company with all of its business units. I think the financial crisis really changed that, both in terms of the way we think about the best practices for risk management and in the kind of governance that’s used. The CRO now reports to the CEO and has a lot more power, which is a totally different perspective from the way that it used to be. So the crisis has galvanized people to think about risk practices a different way.
Sniderman: Can you discuss any particular initiatives the FSOC and OFR share? Is the OFR working on any special projects?
Berner: There’s agreement between the FSOC and OFR on the importance of some of the projects we’re involved in right now. For example, in the FSOC’s annual report this year, the first thing that was mentioned was vulnerabilities in short-term wholesale funding markets. Well, I happen to agree with that. And that’s why we need to improve the quality and scope of data there, and why we need to do more analysis.
However, there are some other things that we’ve discussed but don’t quite understand how they work. For example, short-term wholesale funding markets. Before the crisis, people were willing to lend unsecured bank-to-bank in markets without having collateral to back up the security of those loans. Now there is much less willingness to do that. That puts a premium on the use of collateral and it gives people incentives to trade one form of collateral for another in so-called collateral transformation trades. What are the implications of that? Do we understand how that is evolving in the financial system? You can dig under the surface of some of these broad problems and do some interesting work.
Sniderman: Let’s talk about cyclical and structural ways of thinking about stress. These almost seem like terms borrowed from labor economics. How is that a useful way of approaching financial stress?
Berner: I think it’s quite useful because, you know, there are different kinds of threats in the financial system that require different analytical tools to deal with. The classic example of a cyclical threat is a buildup of debt or a buildup of leverage in the financial system, whether it’s on the balance sheets of households, businesses, or financial institutions. The credit cycle is an inherently cyclical phenomenon. There are structural features that add what’s called pro-cyclicality to the financial system. They magnify the impact of a shock and push the system in one direction, making it more severe. That pro-cyclicality is a combination of these structural features and the cyclical result. In order to remedy that, the structural vulnerability needs to be addressed.
A good example of a structural vulnerability in our system is the runnability of money funds. Now a run, you could say, is a cyclical phenomenon. But money funds, under certain circumstances, will intrinsically be runnable because you’re promising—under current circumstances—a fixed, net-asset value, redeemable on demand for assets that are on the other side of the balance sheet and fluctuate in value. If there’s a shock and the value of those assets goes down, people begin to distrust your commitment to credibly make good on your promise and they’ll pull their money out. That’s a run.
In the buildup of the recovery phase or in the boom phase, there is a sense of market euphoria, a sense that people can get paid to take on more risk. But when the deleveraging comes—when the assumptions change and asset prices go down—the deleveraging is swift and the impact on the economy is sudden.
Sniderman: And this is why the FSOC has proposed certain structural reforms?
Berner: Indeed. It’s why the FSOC and now the SEC in its analysis of money funds acknowledges that it’s a risk. That’s why the FSOC’s and the OFR’s annual reports have all stressed this. In fact, we’ve looked at the degree to which there was risk in the money fund universe and we found that it’s probably more extensive than people realize.
We know that the financial cycle often takes longer to build than even an economic cycle. You can have a couple of recovery and recession scenarios during the buildup of risk in the financial system. In the buildup of the recovery phase or in the boom phase, there is a sense of market euphoria, a sense that people can get paid to take on more risk. But when the deleveraging comes—when the assumptions change and asset prices go down—the deleveraging is swift and the impact on the economy is sudden. We haven’t done a good job thinking of metrics to calibrate financial cycles, even if we look at debt-to-GDP ratios or other aggregate metrics, because there are so many dimensions. That’s something we can probably do more work on.
Sniderman: In the wake of Dodd–Frank, some critics complain regulation reform is happening too slowly. How should we judge the right pace of reform?
Berner: When I think about pace, I think that the real balance is between getting things done and getting them done right. As I think about how we went about setting up the OFR from scratch, some of the things we’re setting up in Dodd–Frank are also from scratch. We want to be thoughtful about the way we do it and we want to get it right. We want to avoid blind alleys. We would rather be deliberate and thoughtful to get it right than be hasty and get it wrong.