2017 Financial Markets Conference—Policy Session 3: Managing the Managers: International Coordination of Financial Supervision

National supervisors had been strengthening their coordination of prudential regulatory policies before the 2007–09 crisis, and those efforts have been significantly enhanced. However, financial firms have mounted increasingly effective challenges to international coordination as firms resist tighter regulations with the support of some of their home country officials.


Fabio Natalucci: Okay, good afternoon. My name is Fabio Natalucci. I'm deputy assistant secretary for international financial stability and regulation at the U.S. Treasury. I'll be moderating this panel on issues related to international corporation financial regulation supervision.

I'm joined by Richard Herring, professor of finance at the Wharton School of Business, University of Pennsylvania, who will be presenting a paper entitled "International Coordination of Financial Supervision: Why Has It Grown, and Will It Be Sustained?"

And we have two discussants. Barbara Novick is vice chairman of BlackRock and Marc Saidenberg, principal, financial services, at Ernst & Young.

So what I'll do [is] I'm going to make a few brief remarks to frame some of the issues that we'll be discussing in the session, and then I will move on to the paper discussion.

In recent years, we have made significant progress, globally, toward the timely and consistent implementation of postcrisis G20 financial regulatory reforms. The Financial Stability Board [FSB], which coordinates the standards-setting bodies, who are to operationalize the G20 financial regulatory reforms. Let's focus primarily on four pillars. One is making financial institutions more resilient; number two, ending the "too big to fail"; number three, making derivatives markets safer; and then number four, transforming shadow banking into resilient, market-based finance.

This regulatory reform agenda has helped create a more resilient global financial system, which is a crucial element to achieving the G20 objective of strong, sustainable, and balanced growth. Now, after 10 years since the beginning of the financial crisis, we believe it's clearly time to pause and assess both the progress as well as the cumulative effects, including both costs and benefits, of the financial regulatory reform. In particular, it seems appropriate to evaluate whether the reforms have achieved their intended objectives, and whether there are meaningful unintended and undesirable consequences; to analyze the interaction between the various reforms (especially in light of the fact that reforms are being implemented at different times—both within and across jurisdictions); and, ultimately, to gauge whether a policy response is warranted, without negatively affecting the original objectives of the reforms.

In the United States, the Department of the Treasury is leading a review of domestic regulations to ensure that they are aligned with the core principles that the president has laid out in February. We have embarked on an extensive outreach with both private and public stakeholders, and we expect to issue a report in early June. The FSB is also developing a comprehensive framework for the policy implementation evaluation of the effects of the G20 financial regulatory reforms—an effort in which the Treasury Department has had an active role. A consultation paper setting out the main elements of this framework has been published on April 11, and comments can be provided to the FSB by May 11. So there is still time to provide comments.

This international exercise is an additional useful tool to review the effects of the financial reforms, along with our domestic assessment. As we do this, an important goal of our international engagement is to promote high standard financial regulations, by bringing other jurisdictions up to our own standards, to lessen threats to global financial stability, and to support a level playing field for internationally active firms—that's diminishing the scope for regulatory arbitrage. Ultimately, the objective is to foster economic growth, safeguard the stability of U.S. financial institutions and financial markets, incentivize financial innovation, and promote competitiveness in global open financial markets.

The G20 and the FSB have played an important role in fostering global financial stability following the financial crisis, and they continue to provide a forum for international engagement on issues related to financial regulatory reforms.

Now, with that, I'll pass on the baton. Professor Herring.

Richard Herring: Thanks very much, Fabio, and thanks to Larry Wall, and the staff at the Atlanta Fed, for inviting me—this is a really interesting conference. When I drove in last night, I was a little skeptical about why it took so long to get to a conference venue. Then I opened my drapes this morning, and quickly understood. It's really quite lovely.

I think my title should have been more like, "And Now for Something Completely Different." The subject is going to be "supervision," which is really pretty far removed, at least at first glance, from monetary policy and the other items that have really—QE and so forth—that have dominated the agenda. As a very quick overview, I want to talk to you about how and why cooperation grew up in the financial sector. I'm going to focus most particularly on banking, because that's where it started first, and I think has gone farthest. There is some information in the paper about IOSCO [International Organization of Securities Commissions] and IAIS [International Association of Insurance Supervisors], but I will not address those to any great extent of my remarks. And then I'd like to talk about how all this changed after the recent financial crisis, and then offer some speculations—which are just that—about what the future for reform is going to be.

Let's start with the motive for international cooperation—and I'm not going to spend a lot of time with this because I hope we can assume it's self-evident—but usually it's an agreement that by cooperating we can avoid or mitigate some negative externality. And I think this has generally been the pattern of development—it's usually after something happens, at last, rather than anticipating something bad will happen—but nonetheless, that seems to be what convinces people there really is a value in cooperating.

But of course there are also efforts to achieve positive externalities—things that we think we could make better by cooperating—those tend to be more difficult to take on because everybody has a slightly different view of the costs and benefits, and the prospects are almost always uncertain. Well, there are some factors that facilitate cooperation that you can pull from the political science literature—and, in fact, game theory comes up with very similar kinds of conclusions. One is that it's easier to agree, the smaller and more homogeneous the number of countries that you need to agree. It's also easier to agree if there's a broad consensus on policy objectives and the potential gains from cooperation—and underlying all this is that we're dealing with sovereigns. All of this has to be voluntary. There's no way to compel anyone to join in.

The deeper the international agreement on the probable consequences of policy alternatives also facilitates cooperation, the stronger the international infrastructure for decision making plays a role. And this one I think is particularly important in the case of banking supervision: the greater the domestic influence of experts who share a common understanding of a problem and its solution—and I would argue that in the early days of banking supervision, the bank supervisors actually had a remarkable amount of discretion over the kinds of policies they could pursue in the supervisory domain. In fact, many of the first activities of the Basel Committee required no legislation at all.

Well, I think if you look at these factors you can understand why cooperation happened first in banking. First, you could actually convene a group of countries that controlled 90 percent of the cross-border banking flows by simply putting together the G10—and G10 was always 12, because when you talked about banking you had to include Luxembourg and Switzerland—but that group of countries basically gave you a pretty full picture of what was going on cross-border in the world at the time. The incident I'm going to refer to next gave everybody a very keen sense of the costs of failing to cooperate, and I think it was not too hard to agree on the first steps. It got increasingly more difficult as they took on more ambitious goals.

The other thing that was very useful is that in the early days, the Basel Committee could build on the infrastructure that had been in place for several decades for the G10 central bankers that were meeting quietly in Basel—which was a very useful place out of the spotlight; I think there's still no direct flights from the U.S., so it's kind of out of the way, but a very civilized place, and there was a well-established forum. And bank supervisors, I think as much as any other kind of supervisor, tend to share a world view and a regulatory culture, and have considerable scope for discretion.

Well, what began it all was—and this is a pattern I'm going to emphasize throughout—a response to a crisis, and this was the hugely disproportionate impact of the closure of a small German bank, Bankhaus Herstatt, in 1974. It was headquartered in Cologne. It was not publicly traded, but the German regulators found that it was massively insolvent and it was fraudulently concealing losses that were more than half of its balance sheet of assets. So they did what was traditional and typical to do for a strictly domestic bank: they shut it at the end of the business day in Germany (which I think was about four o'clock), and that was in the middle of the clearing cycle in the United States.

Well, because Herstatt had been massively overtrading in foreign exchange markets, there were about $680 million worth of spot transactions for which other banks had paid out Asian and European currencies, expecting to get dollars at the end of the day. It was a spot transaction, so in fact most banks didn't even think of it as having credit risk. But they would come to another realization quite suddenly when the German authorities told the clearing correspondent bank of Herstatt to freeze all of the accounts. And they did, and the disruption of the foreign exchange market—at that time, the dollar-Deutsche mark market, which was the most important foreign exchange market in the world—extended for about six weeks as people tried to adjust, to deal with this new perceived risk. It also spilled over to the euro-dollar market, and caused lots of dislocations there—including a number of bank failures.

Incidentally, I have grayed out some of these things that were lessons that were not taken up for a very long time, but came back to haunt us after 2008. These are all issues having to do with closing down an institution, which we should have known about but simply refused to think about. Once the central bank governors started to look at it, they realized that, really, that was not their interest—or, I would argue, competence, although I doubt they thought of it that way. But they did think it was very useful for their bank supervisors to at least get to know each other and to think about how they might coordinate and cooperate in bank supervision. This was a huge change because even within Europe, bank supervisors tended not to know each other. Banking supervision was a strictly domestically focused operation at the time, even though many of the institutions they were looking at had cross-border operations that were often somewhat significant.

So the initial problem was, "Well, if we are going to cooperate beyond a 'meet-and-greet,' how do we figure out what to do?" Obviously, there were potential gains, because if they were really working at cross purposes you could stifle cross-border expansion of banking altogether, and we'd already seen an example where they got in each other's way.

Their first act was the Basel Concordat, which was a very plausible place to start because, since you were dealing with institutions that had a dual life—you both had a presence in a foreign country and so you had a parent supervisor who was worried about what the office was doing, but you also had a host country who was vitally interested. And unless the two coordinated their actions, you could have some very dysfunctional results. So they set out in the Concordat to simply allocate responsibilities: who was responsible to whom, and how should they cooperate.

The first principle, which turned out to be one that continued to haunt them, was that no foreign banking establishment should escape competent supervision. And that was, I guess, easy for them to agree to. The second was how you split up the supervisory responsibilities, and, in principle, the host was supposed to look at liquidity—the underlying assumption was that it's a local bank, it would be dealing in local currencies—so the host country would have a better view of what's going on, and the parent was given primary responsibility for the supervision of solvency.

Of course, to make these joint responsibilities feasible, you had to have an agreement to share information across borders. You had to have some sharing of information—certainly information relevant to understanding the business, solvency considerations, and so forth. And this was a real problem, because in many countries it was a crime—you could go to jail if you were a banking supervisor and said anything about one of your banks to a counterpart abroad—so that was a big leap.

The last thing, which was not stated in the original Concordat because it required legislation in Germany, was to look at these banks on a consolidated basis. I want to look at both of these latter two things because they've been a continuing source of problems.

Sharing information is very, very difficult. Most of the "at-the-border" barriers have been eliminated, although there are certainly still several barriers against exchanging certain kinds of information that relate to privacy in any way. But even when those are gone, you have enormous barriers that are quite understandable. Managers tend to withhold bad news from their supervisors as long as possible, because they're afraid they're going to lose their job or because they're afraid they'll have regulatory sanctions—and there's always wishful thinking that if you can only work on it a little longer, it may right itself and just go away.

So it tends to be a while before bad news in the bank gets to the supervisor. The supervisor also tends to be a little slow in passing it on, because in so doing you're really giving up some of your scope for discretion in dealing with the problem, and I think they too share the hope that it could be self-correcting. This is a problem if you have other supervisors looking at different parts of the bank that are functional supervisors, because there again you may be a little reluctant to share with a domestic supervisor, who is looking at a different part of the bank. You may argue that it's irrelevant, that you're on top of it, and you don't really need to share.

The other problem, of course—and a much bigger problem—is sharing it across borders, because your counterparts abroad have different legal obligations. Upon learning about something unpleasant, they may have a legal obligation to act. You're always concerned about leaking information, and especially bad information that might result in a run. So there is a tendency—despite getting rid of the "at-the-border" barriers—to really have problems on a continuing basis about sharing relevant information. And remember, this is the important time to share—when you have a situation under stress—and that's the time it's least likely to happen.

There are also issues with consolidating supervision. It was a practice that the Americans and the Brits insisted on. The Germans were very reluctant because they had actual legislation in place that actually made it illegal for them to inquire about what was going on—in the Luxembourg banks in particular—and that had to be lifted. The idea behind this is that if you really think of these international entities as operating on an integrated basis in which they can, without any real friction, transfer capital and liquidity from a surplus unit to a deficit unit, then it doesn't really make sense to look at it in pieces. You want to look at the institution as a whole.

And indeed, that would probably be a better system for the world. It would be more efficient, people could distribute their capital and liquidity more efficiently—but that's not the world we live in. In fact, it's sort of a delusion to think you can do that in the event of stress, because one of the first things that countries do when they're concerned that an action taken abroad may well harm their residents is to put a ring fence around it one way or another, and that is an issue that has continued to plague the efforts to coordinate things in times of stress.

So, in effect, institutions will ultimately have to meet standards, not only on a consolidated basis, but increasingly on a standalone basis as well. The first test—and I apologize, this is in terrible taste; but this is unfortunately a real photograph. Let me hasten to add, the supervisors did not hang Mr. Calvi, it was clearly the Mafia and the Freemasons. But nonetheless, this was the end of Banco Ambrosiano. The leading character fled to London, and unfortunately did not make it.

But what was at stake in this case was principle number one in the Concordat. And Banco Ambrosiano had situated itself in gaps in the regulatory and supervisory system to evade detection altogether. It had set up a subsidiary in Luxembourg that was majority-owned, but in Luxembourg it was not registered as a bank. It was a secret holding company, so the Bank of Italy could not properly question what was going on there, yet the bank in Luxembourg held basically two different banks—in two jurisdictions, I might add, that were not really famous for the quality of their banking supervision. And most of the business was done through these two entities.

So in effect, they had been able to evade any kind of regulatory oversight at all. The impact was to revise the Concordat—again, sort of showing the pattern in which we have a negative externality that pops up, they try to adjust coordination protocols to take it into account. And this one was basically a hidden rebuke to the Bank of Italy, saying, "Look, if this institution is not classified as a bank by the host country and you think it should be, then you should either shut it down or demand to supervise it." So, that was the hope.

This is the only bank organization chart I know that is also a work of art—this is to be found in the Whitney Museum—but it is Mark Lombardi's interpretation of the actual management control structure of BCCI [Bank of Credit and Commerce International]. This was actually appropriated by the CIA after 9/11 because there are several links here to Al Qaeda and Osama bin Laden, in fact. So they were trying to take it apart to see what he had found out, but he was obsessive about this. He had all sorts of diagrams that did this kind of thing. This is simply to say the structure was incredibly complex and almost impossible to disentangle, and it, too, had managed to escape any kind of competent supervision.

They'd done it in a very similar way. They headquartered themselves in Luxembourg, but all of their banking activity was done outside Luxembourg, so Luxembourg had no interest in supervising it, and the host countries didn't want to take on the task of supervising it all. They further fragmented oversight by hiring two different accounting firms to look at each part of it, so nobody had an overall view. It probably started out as a legitimate bank, but when they got into huge losses in London, they diversified their business plan and they became a full-service bank, in I guess the most appalling way you can imagine—they developed sidelines in arms dealing and drug running, and virtually every criminal syndicate, and intelligence organization, in the world had an account there.

Well, they did go bankrupt, and it showed there was yet another hole in the Concordat. They were unable to supervise everybody, so this amendment sort of put some teeth into it. It said, "Look, if the host country determines the home country supervisors are not performing competent consolidated supervision, then they can impose restrictive measures—including forbidding the institution to operate at all." At the same time, the U.S. stepped up its oversight of foreign banks quite sharply.

Parallel to these efforts to figure out how to allocate supervisory responsibilities, you had another stream of work in the Basel Committee that was certainly equally important, but maybe a little less obvious. It was an attempt to harmonize supervisory frameworks. And this was important because when the whole process began, you had very different attitudes about what a supervisor could or should do. You had disagreements over whether they should have onsite examinations, and a whole variety of things. So what the Basel Committee tried to do—I think believing that unless they had some kind of convergence in supervisory structures, they really couldn't get on with the agenda—was to reach a consensus on the core principles of banking supervision. That's been supplemented over time with a whole string of papers about—I won't say best practices, but certainly sound practices—in a variety of aspects, from managing derivatives to oversight of the trading book, and so forth.

These have become increasingly important over time because they have been served as sort of the benchmark against which the FSAP [Financial Sector Assessment Program] exercise, by the IMF and World Bank, judges the financial sectors of all countries. So, these are standards set by Basel that have become highly relevant to any country in the world who is subjected to the financial sector oversight protocol.

It has also become even more important after the crisis, when we are now having biannual surveys of just what the gaps are. Which countries do not meet them? Are they in process to correct their deviation? It's important to point out, however, that everything the Committee does lacks legal force. You can't take anybody to court. The best you can do is hope the process results in some sort of moral suasion, and more recently it's become sort of a "name-and-shame" arrangement, which has its limitations, but may be the only way you can proceed.

Of course, the accord on capital adequacy was by far the most ambitious thing they've tried to do, which was to set a common minimum requirement. I won't go through the details, because that's a whole other three or four hours that would certainly bore you to tears. But the important thing to note about this is that it had two objectives. It was in response to, in this case, I think not an actual event, but the perception that the banking system had taken on huge amounts of risk exposure relative to their capital. This was the beginning of the LDC [Latin American debt crisis] debt crisis. Most of the large U.S. banks had lent more to these countries than their entire capital, and regulators were concerned. But when they began to raise capital standards unilaterally, they found that their banks were quickly losing market share to foreign banks that played with different rules.

So from the very beginning, there were two objectives in forming these regulations. One was the prudential objective, to make the system safer and sounder. But at the same time, there was a "level playing field" objective, which may be inevitable, but really complicates things incredibly. Virtually every odd detail in the Basel Accord—in Basel I, II, and III—is that it was an accommodation to try to somehow level out the playing field, because of the strong objections of one country or another. So you ended up with very peculiar things, like defining capital for Japan to include 45 percent of unrealized gains on stock and real estate, or in the U.S., to allow us to count loan loss reserves in excess of the usual.

So that was another effort.

The next main crisis was both a bank failure—but this brought IOSCO into the picture: this was the collapse of Barings in 1995. It showed that all of this sharing had really not conveyed the relevant information to the right parties at the right time. You had problems in the British specialized regulator actually collaborating and sharing information, and you had huge problems between the home and host country authorities.

But the important thing that I'd like to emphasize here is that the losses at Barings' Singapore office were all from supposedly arbitrage transactions between the futures exchanges in Osaka and Singapore—huge amounts outstanding that the trader had managed to conceal for a number of years, but became so large after the earthquake hit Japan, in Kobe, that you had an abrupt fall and the loss was so large that he was unable to cover it, so he fled, trying to evade prosecution, and Barings had no choice but to go into bankruptcy. They did try to appeal to the Bank of England, but the Bank of England wasn't moved.

The problem here was that nobody quite understood how bankruptcies were to be held by exchanges. And when they began to pull out the agreements about loss sharing, a number of important players said, "Gee, you know, I'd really rather withdraw from this exchange than be obliged to absorb some of the losses from Barings' fraudulent behavior." And it led to a situation, very briefly, where people thought you could have a contagious collapse of securities markets around the world.

Luckily, it didn't happen, because they were able to sell Barings for a pound sterling to ING, but the threat was really quite appalling, and the response was that the supervisors of derivatives exchanges in 16 countries met at Windsor Castle and agreed to the Windsor Declaration, which was more or less the securities counterpart to the Basel Accord. It was about how you'd share oversight and share information, and it even went into details of having an on-call schedule, which was a bit odd except that one of their problems in collaborating is they couldn't figure out where the Swiss representative was—he happened to be having a picnic in the mountains—but it's now like an emergency room so that you now know who you can call if something happens.

IOSCO wasn't the convener of this, but they were a participant, and these principles got built into IOSCO very promptly.

The next issue, which also touched securities—and I'm really suggesting how the agenda of IOSCO, which was originally about enforcement cooperation, came to think more and more about looking at stability and systemic risk.

I won't belabor the points in Lehman. I think you all know them all too well. One thing I would like to note, however, is that the total number of employees at Lehman Brothers was actually fewer than the number of people employed in compliance at Citigroup these days, to give you an idea of just how massively the situation has changed. The Lehman structure was almost a perfect poster child for what not to do if you want to get a firm ready to go through bankruptcy. Chapter 11 had to be applied in a hurry. You had an incredibly chaotic bankruptcy, absolutely no preparation—still the largest bankruptcy, thank God, that we've ever gone through, and it was a complete mess.

The fallout from Lehman Brothers seemed to justify the presumption that most other authorities had made, which was it's much better to do a bailout than to actually resolve one of these large cross-border institutions. The result was that countries were engaging in bailouts on a massive scale—Andy Haldane estimates that Europe and the United States allocated about 25 percent of world GDP to back up the banking system. Luckily they didn't have to pay it out, but it was still a huge cost in terms of the strain on fiscal capacity. And probably also one should think about—but it was not the sort of thing anybody dwelled on at the time—the moral hazard it was setting up for future rounds, because once you start it, your protestation that you would never bail out a bank becomes less and less plausible.

Even with these bailouts, of course, we've had huge impacts on the real economy. At the time, officials were keenly aware that this was by no means the worst case. Many of the global systemically important banks were much larger—trillions, not billions—had much more extensive interconnections with other banks in the system, had much more complex inter-affiliate transactions, more diverse lines of business, more complex organizational structures, and more extensive international involvement. If they were confronted with a larger problem, no one thought there was any possible way to resolve.

So the lack of a plausible resolution policy had really led to these massive bailouts, because when supervisors got to the brink, no matter how much they'd argued they would never bail out a bank—when you looked into the abyss and saw chaos or having to deal with the political consequences of making a bailout, they almost always chose to bail out.

Well, the G20—and this was important because the G10 had previously been making most of the...or the G7. The G20 met to talk about what to do in the wake of this crisis. The first meeting was in the fall of 2008. In the spring meeting, it actually turned to regulation quite specifically, but the G20—the group of leaders, political leaders from the most important countries in the world, in this context—agreed on a reform agenda. This really changed the whole political legitimacy of the coordination processes. Before this, it had all been agreements among specialists that hadn't really had—they were all unelected, and it didn't have the sort of political legitimacy that a group headed by the heads of state would.

The G20 basically had two different prongs in its strategy. One was to strengthen prudential supervision and regulation—and some of you know how this has been firsthand. It wasn't different in kind, but it sure has been different in intensity and quantity. What was totally new, however, was the position of the resolution policy because, after the experience of Lehman and the theory about why bailouts had been so large, the G20 was absolutely determined to make sure that there was a credible way to resolve any one of the G-SIBs [global list of systemically important banks]—and that's pretty much the situation we're now living through, and it is altogether different. It's long. It's very, very different from the other reforms, which were basically just intensifying what Basel was already doing.

This is hard, though, because the G-SIBs had grown enormously since the 1990s. They had management structures that were usually unified and basically agnostic about legal structures—most banks didn't even do accounting by legal entity, but in the event of a crisis it's the legal entity that has to go through some kind of a resolution. The cross-border complexity implies that at least two authorities have to be involved in any kind of resolution decision, and that's going to be complicated. It creates a big communication challenge. And since most G-SIBs have entities in scores of countries, that is a huge coordination problem.

In addition, the cross-sector complexity—the fact that most G-SIBs don't do just a banking business, but surely have a leg in the securities business, often in insurance, and frequently in a lot of other things; in fact, by number of subsidiaries, that's the bulk of the activity—implies that you're going to have to deal not only with the host country bank supervisor, but specialist supervisors in lots of other cases. So, quickly you get to a ballroom full of supervisors that somehow have to be consulted.

The understanding of this moved resolution tools to the very top of the G20 reform agenda. And to carry out this agenda, the G20 created a new institution—partly because none of the old coordinating bodies had really acquitted themselves with great glory in the crisis, so there was a credibility problem, but more importantly because the scale of the reforms they wanted to put in place spanned well beyond the domain of any one of the specialist coordinating bodies.

And so the Financial Stability Board was created—it was technically a repurposing of the Financial Stability Forum, but altogether different—and it set out as one of its first tasks the key attributes of a resolution regime for financial institutions. This was a huge stretch because when they began, most of the members didn't even have a resolution regime, so this was a stretch goal to be sure. And it has been a difficult thing to accomplish. They have made massive efforts. We now have crisis management groups for every G-SIB that is an informal way for authorities to cooperate in overseeing an institution. And in addition, it is a way for trying to assess the way in which firms had adjusted to the new regime.

Everybody, of course, is required to file a resolution and recovery plan—or living will, as it's commonly known—but it also highlighted the limits to international coordination. And the brute fact is that when trouble arises, national authorities are inevitably going to place greater weight on domestic objectives and protecting domestic residents, in particular, than cooperating abroad. And even if the key attributes were adhered to by all the member countries, you're still stuck with some very fundamental asymmetries: asymmetries of resources—some supervisory institutions are much better financed, have much better human capital and ability to deal with things than others; huge asymmetries in financial infrastructure—the quality of accounting, the quality of any kind of press oversight of banks; and huge asymmetries in exposure—a firm that may be very systemic in one country may in fact be not so big in another.

So there is going to be, inevitably, an attempt for the host country to look after its own. That means that no matter what you agree to in the key attributes, you're going to have to think about a scenario in which ring fencing will dominate. And as a result, we've seen major countries actually set up structures that will be ready-made for ring fencing. That's exactly what the U.S. has done with the intermediate holding company structure. Britain has done something similar in implementing the Vickers Committee reforms, and the EU is threatening to do much the same. Others are doing it more informally by pre-positioning capital and liquidity.

Will international coordination continue? Certainly the velocity of regulatory change is beginning to slow, which is I'm sure a relief to everyone in the room. And it's imposed enormous costs on regulators and regulatees. A colleague of mine has recently done a paper looking at the resources that are allocated now in compliance with these rules, and he's come up with a figure that is mind boggling, but indicates just how far we've gone. He's counted more employees in financial institutions that are focused on compliance than the total number of all the policemen on the beat in all of the cities in the United States. It's a very odd societal choice, but, nonetheless, that's where we are.

So I think it's high time to pause and take stock, and ask: are regulations fit for purpose? Could we have accomplished the same objectives at lower cost? Should we have placed greater reliance, as I would have liked, to a disclosure and market discipline? I don't think the FSB work plan, as set up, is going to hit that. I wish it would. Why is it that we didn't get to a mutual recognition regime?—which, I think was the hope. And, have we attempted to coordinate too many details of financial regulation and supervision, or have we in fact attempted to coordinate the wrong details? My proposal, for a mistake, would have been allocating a lot of time to trying to establish a net-stable funding ratio, instead of standardizing data definitions and regulatory reports—which would at least lighten the compliance burdens in the world.

As memory of the crisis recedes, it's inevitable that the enthusiasm for regulatory reform will fade. The factors that explained growth earlier on are also going to explain why it's inevitable that coordination will be more difficult looking ahead. You now have to convene a much larger group of countries to have the same part of the world financial system represented. The Basel Committee now totals 28 members, but there are continuing pressures for access because decisions made in Basel are now having important impacts on a lot of countries that are not represented around the table.

I think the urgency of cooperation seems less strong than it did in the wake of the crisis. In fact, some people are inclined to say, "Let's declare victory and start rolling back." And you're beginning to see disagreements about whether this is all a good thing or not. It's frequently, I guess, focused on Basel III, where there are a substantial number of critics who claim that Basel III reforms have in fact slowed the recovery and contributed to economic stagnation—so you don't have the same agreement about how to move forward.

Although we now have huge infrastructures to support international coordination, the national support for these entities is really beginning to fade. I think most important of all, though, is the domestic influence of experts is in decline. There's a widespread view that they simply failed to safeguard the international financial system before and during the crisis, and there is a growing contempt for experts that is associated with the rise of populism on both sides of the Atlantic—at least it didn't carry the day in France, but it was still a very strong proportion of the French voting public.

Bank supervisors and regulators now have, I think, a reduced scope for exercising discretion. Issues are no longer simply regarded as technical and relegated to the specialists. They're now highly politicized. Tt's difficult to think of a regulatory issue that can't be reconceptualized as some kind of covert way to redistribute power or income from one group to another, and there's always the abiding concern that it's somehow a hidden subsidy to Wall Street.

This is a real problem with regard to the new administration and Congress in the U.S. As you know, President Trump has vowed to "do a number on Dodd-Frank." It's still not entirely clear what that means, but his executive order—which is kind of a loosely drawn memo, with odd bullet points—does have one bullet point that appears to be directed at international coordination: the regulators are supposed to advance American interests in the international financial regulatory negotiations and meetings. I think our counterparts would wonder why that's a change. I think they believe we've been doing that pretty aggressively before, but nonetheless, that seems to be the new charge.

Even more explicitly, Congressman McHenry, who's the vice chair of the House Financial Services Committee, has written a very sharp letter to Janet Yellen saying, "Despite the clear message from President Trump, it appears the Federal Reserve continues negotiating international regulatory standards for financial institutions among global bureaucrats in foreign lands without transparency, accountability, or the authority to do so...this is unacceptable." So there are pressures that simply did not exist 20 years ago.

There's also a sense in which the Choice Act—the House proposal to reform Dodd-Frank—is going to pull the U.S. away from Basel standards, and it's particularly clear with regard to resolution policy. There's enormous anger in Congress about Title II of the Dodd-Frank Act. Part of it is the conviction of a number of people that, if you give any possibility for government bureaucrats to subsidize, they will, and so they wanted to try to take bureaucrats out of the decision altogether. There's also another group that believes Title II simply perpetuates the notion of "too big to fail." If institutions are too big to go through a normal bankruptcy, then why should we set up a special resolution procedure for them? Shouldn't we just make them smaller so they could be resolved in bankruptcy?

And then there's another group that has sort of an absolute ideological preference for rules-based court procedures, rather than giving administrative discretion to regulators. This is very odd in the eyes of the rest of the world. I think foreign supervisors view the FDIC and the Fed as much more reliable counterparts in negotiations than dealing with U.S. courts, which they regard as something of a crapshoot. The problem, too, is that it's impossible to cooperate. You have no idea who the judge will be in a bankruptcy, so there's no possibility of having advance arrangements, and the bankruptcy judge is going to have to make a judgement over a very tight weekend, if they're lucky. So there is a real problem in seeing how this can advance in a cooperative manner.

There are also strains in the Basel Committee. As you know, Basel III reforms have not been completed on time. There are huge debates between the EU—Germany in particular—and the United States over putting a floor on internal risk weights. We see it as basically an abuse of internal discretion; the Germans see it as a competitive advantage that we're trying to gain because they hold more mortgages on their balance sheets than we do. And underneath it all is not just the usual carping over a level playing field, but a real change in financial structure. We now rely incredibly heavily on the GSEs [government-sponsored entities]. So banks have put most of the mortgage risk out into the GSEs, and consequently don't hold much in the way of mortgages—the others do—and so it suggests a limit to standardizing rules if financial structures remain very different.

Finally—and I realize it's high time to get there—there is, I think, waning support for multinational institutions generally, and in financial cooperation, specifically. I guess we have to hope that we don't have another shock that demonstrates once again the cost of failing to cooperate.



Barbara Novick: I am with an asset manager called BlackRock. Not surprising, my remarks will be very focused on asset management—both what's been our experience with these international bodies and international cooperation, and what are some of our recommendations.

As an asset manager, I should start by saying we actually value financial stability. I describe it as we don't want to be driving down the road and having bombs going off on the side of the road. That's not good for us, or for what we do for our clients.

We have a global business. We're active in virtually every market around the world. We have clients in over a hundred countries. We even have offices in more than 30 countries, so very much global in nature.

I would say when I do an assessment of what's happened in the last nine years, we do have a system that is safer today. The question today is, is it too safe? Have we made it so safe that we have more compliance people than we have policeman? That we have more reporting than we have people who could accept and read and study those reports? And is all of this really necessary? Are there ways we can streamline it?

I've come to calling this "regulatory right-sizing." It's not about repealing rules; it's about looking critically, now that we have this huge body of them, and saying, "Where are the overlaps? Where can we cut some things back without losing on the safety?" And I think there's a lot of opportunities.

On a prior panel, Julia mentioned the fallacy of regulators believing in perfect market efficiency. Richard's paper actually suggests that human beings are just that: they are humans, they're mortals—that includes the market people and the regulators—and I think at the end of the day we all need to be more inclusive. We really need to listen to each other, listen to what the concerns are, listen to what the practical realities are, to be able to comply with some of the proposals.

So let's take a look just at the reporting for asset management. This is all post-crisis requirements. Now, I could argue very easily there was not sufficient reporting before the crisis. Don't worry, we have more than sufficient reporting now. And you step back and think about: is all this data useful? What is the point of having reports that go, even within a country, to two different regulators with slightly different definitions of the same fields, of the same information, reported in different ways all over the world?

It's really not very useful at all, and since there are many academics here, I think you'll understand, my point being, data is not the same as information. We need to harmonize the data. We need to figure out standard definitions—make the reporting the same—so that you can compare the data, you can contrast, you can look across time periods, you can look across jurisdictions. If you do it by rolling it up to a high level, you don't run into the issues about illegal sharing of data. But if it's not consistent, rolled up, it's meaningless. So we really need to get to the bottom of that.

Richard mentioned the executive order. Here are the seven principles. There's a lot of focus on, what is the U.S. going to do next? And I would say, when I look at the executive order, it reads a bit like "mom and apple pie." I think these are very nonpartisan. They don't really make a whole lot of issues about any particular regulations—they're simply common sense. And you might want to say, "Well, were we doing that before?" And I think I would argue not.

So this kicks off a multi-month study that, as Fabio mentioned, the Treasury is currently in the midst of. We've been told to expect multiple reports, starting in June, on different aspects of the information that they've been collecting. But they've actually done a very inclusive process. They've had I don't even know how many round tables, but I think a lot, as well as bilateral meetings, and they really seem to be very serious about understanding it, almost like a call for evidence like they did in Europe under capital markets union.

Likewise, last week there was a memorandum focused specifically on FSOC [Financial Stability Oversight Council]. As many of you probably know, MetLife challenged the decision of their designation in court. the judge's language in ruling that they should not be designated said that the designation was arbitrary and capricious. I think, as an American, I have problems with that.

And in asset management, of course, we feel very positive that there was a pivot to products and activities, but it was a common sense thing. If you think about money market funds, you would never say, "Take the top three, the largest three, and somehow regulate them differently than the others." If you take OTC [over-the-counter] derivatives, you would never take the largest dealers and regulate them differently from the others. It just doesn't make any sense.

So when you look at mutual funds, or you look at other products and activities of asset management, you quickly realize if you don't regulate it across the whole system—and I know I've stood on this platform in prior years and said the same thing—if you don't look system-wide, you don't actually reduce risk in any way. So, very important to take that pivot. And we do expect to hear more about FSOC and the designation aspect over the coming days. In fact, they're meeting right now, so maybe they'll be some smoke signals coming.

The Choice Act has been mentioned. We would say you should expect some legislative action in this new Congress, but you should also be looking very much at the regulatory space, where congressional action is not needed. So taking a quick look at Choice, led by Jeb Hensarling as the chair of House Financial Services—very importantly, this is not a repeal of Dodd-Frank. I hear sometimes people say, "Oh, the Americans. They want to repeal everything." In fact, it targets very specific aspects of Dodd-Frank—I'm not going to get into all the specifics, and which ones we agree or don't agree—but it also leaves quite a few of them alone, and in some cases it actually has harsher penalties than what was in the original act.

So it's something worth looking at. And if you don't want to read the whole act, I can understand—unless you have insomnia—but read a summary. It's actually very interesting how they go about looking at very specific aspects. That said, I think it'll pass the House, and I think you can see it as a beginning of a negotiating point with the Senate—nothing even close to the Choice Act is actually going to pass the Congress and go to the president. So, very important to look at that regulatory side.

As more people take their seats, we expect to see more activity. First is the CFTC [U.S. Commodity Futures Trading Commission]. Chair Giancarlo has already put out a white paper. He's given a number of speeches. He's particularly focused on technology—things like blockchain, how could that maybe change the risk in the industry; cross-border issues; and what are the SEF [swaps execution facilities] rules—which, as we know, are extremely complex. We have to give the SEC [U.S. Securities and Trade Commission] a little time. Chair Clayton was only confirmed last week, and took his seat at the end of the week. He hasn't yet written any white papers or given any speeches as chair. That would be pretty miraculous if he had. So I'm going to make some guesses, and they're just that. They are my personal guesses of what areas I think he will focus on.

Based on his testimony, I think capital formation is going to be a very important thing to him. Why do we have fewer IPOs [initial public offering]? Why do we have fewer listed companies? That's an area of clear focus. Likewise, I think he will look at the fiduciary standard issue—which we don't need to go into all the history there or where we are—but it's something that really the SEC should own. It's not something that makes sense for one kind of brokerage account versus another, but something that should be done across the industry.

I think he may revisit some of the rules for asset management. The rules themselves are reasonably good in many ways, but are extremely complex with very short implementation timeframes. And so I think even just a delay of some of the reporting requirements—and maybe some tweaking of some of the calculations and things in the rules—maybe as an order. Maybe we'll see that, maybe not.

There are certainly rumblings within the SEC, and within the FSB about ETFs [exchange traded fund]. Oddly enough, just before the crisis, there was almost an ETF rule. It came right up to the edge and then—change of administration, flash, crash. Lots of things happened. The rest is history. It never came to fruition. I think there is appetite in the industry, as well as with the regulators, to revisit that and try and come up with what I call a "vanilla rule." Why can't we have something very simple that makes sense for long-only securities in an ETF? Derivatives, leverage—all different, but at least have a vanilla rule.

And then, lastly, electronic delivery. We are in 2017, last I looked. Most of you I know have checked at least one if not multiple electronic devices during the sessions today, myself included. I'm just as guilty. I think we should get to the point where we're comfortable delivering shareholder information electronically. Everybody's got a computer or access to a computer if they're invested in mutual funds.

So, let's turn for a moment to the international forum. Think about how what I just said lines up with what's going on internationally—clearly a desire for consistent standards I think, is true across the board. I think there is a growing recognition amongst many of these bodies that it's gotten very complex—the FSB has actually put out a framework for comment about how to even assess what's been done and where we are, where we need to go forward. We certainly don't want to see regulatory arbitrage. We don't want to see contagion across borders. But we think that there's room for, let's just say, simplifying.

Mr. Mori from Japan, he identified 140 work streams between these groups in Richard's paper—he stopped counting at 100, but just, you know, Mr. Mori got to 140. So imagine, for asset managers, who have no history of dealing with the FSB or its predecessors, or virtually any of these groups other than IOSCO, and even that rather loosely. This has been just like a whirlwind, where you're not sure where to turn first because there's always a new proposal.

So, what are my conclusions? First is, I think global standards can reduce regulatory arbitrage and can be very good, so I would say we are supportive of that. As you can imagine from my "data is not information" [presentation], I think regulatory reporting begs for harmonization. Giving lots of information to lots of different people, in lots of different formats is incredibly wasteful. Maybe it cuts back on some jobs in compliance, but I think it would be money better spent differently.

There is a tsunami of regulation. I find it sort of interesting that even regulators are having trouble digesting all the new rules and being able to supervise the entities that they're responsible for, because they just can't get through all of it—they're doing ongoing training, they're really trying to stay abreast of it, but [that is] very difficult, and the cumulative impact is yet to be assessed.

So between the Treasury study, the FSB study, I think there'll be more opportunities to take that step back, think about what we're doing, think about what makes sense, and hopefully do it in what I call an inclusive process along the lines of what Treasury is doing now, and try and make sure that that practitioner perspective is factored in, because some of the things just aren't very practical to do and we'd like to flag those as early in the process as possible.

So, again, thank you for inviting me, and for the time today.


Marc Saidenberg: So I was joking with Scott before this started that I am probably in a—if not unique, then an atypical situation, that I can talk about supervisory coordination, having spent most of my professional career living and breathing supervisory and regulatory coordination. But I'm now neither a supervisor nor actually at an institution that is supervised by any of these agencies, so I may be in a position to speak a little more freely than some. I should also say, first, thank you to the Atlanta Fed. I very much appreciate the opportunity to participate. And on a totally separate note, I would also congratulate the Atlanta Fed on Raphael's hiring—notwithstanding the fact that he sits on the wrong side of the big game, still one of the great people around, so congratulations on that.

When you have the chance, please, I encourage you to take a look at Dick's paper. It is it is a tour de force in discussing the evolution of supervisory and regulatory coordination. And in the spirit of something I was taught about presentation skills—and also in the spirit of trying to self-edit—let me just start with the punch line, which is: I think we may be focused on the wrong thing if we're asking the question, "Is the presence or absence of coordination, or the level or nature of regulation, contributing to financial stability?"

Probably from my perspective, the most critical question—which is touched on in Dick's paper—is really the time consistency, in that what we see is that time-inconsistent supervision and regulation, including assumptions about coordination, contribute to financial instability and contribute to, and may actually act as a transmission mechanism for, shocks.

So, the takeaway—and I'll come back to this at the end, in a few minutes—is, as we look at the set of regulations that we're discussing, rather than focus on the question of what we think is the best, or necessarily the "first best" choice, I would say it's more important to focus on, "What is the approach, both with regulation and supervision, that will be credibly maintained through a financial crisis?" Because Dick's paper does a great job of walking through some of the history of, what are some of the events—and he highlighted a number of these in his presentation—what are the events that contributed to the different waves of financial regulatory and supervisory coordination, or efforts to enhance regulation, efforts to enhance cooperation? And what's interesting, I think that he starts with the premise that the absence of coordination can lead to negative externalities.

What I think a number of us have seen—and I'll give you a few more examples, and it's highlighted by the examples that Dick gave—is that it wasn't the absence of coordination in these events that really contributed to sources of financial instability, but the common characteristic of each of these events is a case where a set of assumptions that were made that guided either regulatory or supervisory behavior turn out not to be true in the context of a financial crisis—or not to be able to be maintained in the context of a financial crisis. And some of the examples—questions of the assumed level of coverage of supervision and regulation, assumptions about the ability to coordinate in a crisis, assumptions in other cases about the sufficiency of capital and liquidity, and the source of any backstop for capital and liquidity.

In each case, one of the common themes that sits behind the examples that Dick gave, and what really drove it, is that there was what had seemed at the time (ex-ante) a reasonable expectation that something would happen, and then what turned out in the crisis was sort of similar to other things—you might say this about asset prices in other places—the nature of the way supervisory or regulatory cooperation may have contributed to the financial crisis was not what it was ex-ante, but—like asset pricing, or something else—which is a revelation that a set of assumptions that guided behaviors before the crisis could not hold when you're in a financial crisis. And it was the resetting of those expectations that really—and in some cases, the resetting of that behavior—that contributed to some of the financial instability.

So, let me highlight just a few cases, some of the examples that are not in Dick's paper. I have to confess, these are drawn from my own history. You won't be surprised—but just a couple things—some of them are examples of regulation, and some are supervision—where the takeaway, at least that I had, from looking back on these—sometimes not with entire bits of pride, I will start the conversation that way—is that there were things that we could have known beforehand, that if we'd really thought about them as, "How will this work in a crisis?"—that we may have come to a different conclusion than we did. And it was the nature of that revelation, and the change in behavior—that time inconsistency—that actually contributed, unfortunately, in some cases to financial instability.

So the first one—this is something that is familiar to many of us—is Basel Committee's definition of capital. There was a long period where a lot of work went into trying to find ways to make capital able to absorb losses when a banking organization failed, and there was also a long discussion—some of it was captured in an agreement the Basel Committee made on innovative capital instruments—by the way, those are two words that should never go together: innovative and capital. They just aren't—if you have to ask if it's capital, it's not. [laughter]

But, a long discussion about the efficiency and the innovation in capital—all of which made sense if you believed in a world where ultimately the value of capital was to stand in front of the deposit insurer when you were resolving a bank. Now, the reality is the world had moved on. Somehow we didn't fully process this—notwithstanding this conversation was only 15-ish or so years ago, 20 years ago, maybe, and what we found in 2008 is that all the market cared about—and, in fact, all supervisors really cared about—was tangible common equity and the ability of capital to absorb losses for a going concern.

Now, with hindsight, we all say, "Well, of course that's what matters. None of us want to consider the ability of these firms." The question should not be limited to, what happens to a bank in failure? But that's an example of where the ex-ante assumption that capital could absorb losses when it was needed, turned out to be not true. Capital can only absorb certain losses—or things that we call capital.

Another one, an example of supervision—two of us are in the room who sort of resemble the following remark—there were a good five years, if not longer—and Dick highlighted an aspect of this—where a lot of the focus on liquidity risk management was how to have effective integrated liquidity risk management and facilitate the, maintaining sufficient liquid assets at the top of an entity, at the holding company level of an entity so that you could downstream liquidity efficiently and wherever it was needed. It's a great theory, a really good idea. Failed miserably.

The issue was that the supervisory coordination that was assumed—that is, that you could rely on the home country supervisor to downstream liquidity into the entities where it would be needed, at the appropriate time, and not actually let a bank—in this case, an investment bank—take the liquidity out of the jurisdiction at the exact moment it was needed. Bad idea, it turns out.

So, discussions now, where there's a lot of very painful discussions—and Barbara touched on these—about trapping liquidity in different jurisdictions—while there is certainly an argument that it is a second-best solution to make it difficult to have liquidity flow around an organization, the point I want you to hopefully take away from my comments is if the reality of the situation is, in a crisis, liquidity will not be fungible, then the best thing we could do is create a set of expectations, through regulation and supervision, that the liquidity will be in the place that it's needed and not available to be freely flowing across the entity.

So it's a trade-off of, where most of us would sit here and say, "No, it's far more efficient to have the smooth-flowing liquidity to where it's needed. Fight the fires, limit contagion." The reality that many of us experienced in 2007, '08, '09, and since, is that's not what's going to happen, and so the recommendation I would have for all of you as you evaluate a question of policy, is really ask, what's the time-consistent policy, which may be assume that liquidity will be trapped, and therefore both consolidated host and home jurisdictions should act with that assumption, [is an approach that] may be more credible and do a better job of promoting financial stability than the approach that you would ex-ante assume that free-flowing liquidity would do so.

I'm mindful of time, so let me just give one more example. So another one—I love self-criticism, it's fun—the Federal Reserve and a number of other supervisors spent the better part of the 2000s, really the first half of the decade, convincing firms ("convincing" is a nice word), convincing the public and requiring firms to better integrate their risk management systems—that is, find ways to more efficiently aggregate exposure so you have a consolidated view of capital, liquidity, risk taking, many things. No argument that having a comprehensive view of risk is a good idea.

If doing so limits your ability to address the distress of a financial institution—limits the firm's own ability to address the distress of that financial institution—then that emphasis on consolidation may be somewhat backward. There are several large financial institutions—one that comes to mind—where their ability to actually respond to the financial crisis was severely limited because they had become so integrated that there was no concept of, which portion of this firm could we sell off to help recover the firm? Which portion of this firm could we actually choose, or could supervisors force, to actually reduce the size and complexity of the firm? All that was left as a supervisory option at that point was recapitalize the firm, guarantee assets, guarantee its liabilities, and continue to do that until, several years later, you might actually have the ability to resolve that firm.

So, the reason I raise all this is basically to come back to is what I hope you might agree is the takeaway—is that as we evaluate policy and the merits of coordination, in addition to asking the question, is it good policy, is it one that will promote financial stability on its own merits?—is to ask the question, is it a policy that will be consistent and credible through time? Is it an approach that will be consistent and credible through time? And that a source of financial instability is actually this time inconsistency. So I fear—and, going to some of the points that Barbara made—not that I agree with everything you said—but I would say I fear that much of the focus today, even from the FSB, of, "Let's review policy, and is it efficient?" will sort of ask the question, in today's market, could I be more efficient in the regulation and supervision of these firms? Could we better promote cooperation?

I think crisis management groups are a great idea that have—no disrespect to anyone in the room—blessedly little to do with actually managing financial crises, because when it gets to a point of crisis, the first thing that will happen is the openness that was exhibited in the CMGs [crisis management groups] today will immediately, for the very reasons that Dick described, will go away. Because there will be grave concerns about the information getting out to those who shouldn't have it, and so this question of, how do you evaluate the current set of policy options?—but really from the perspective of, what will be consistent through time?—I hope becomes part of the debate.

Thank you.


Natalucci: Okay, so we have a number of questions here. I'm going to start with one that's gotten the most votes. Does the complexity of Basel capital standards encourage financial firms to game the standards, with unintended consequences, and obscure the true state of the balance sheet?


Saidenberg: I don't think it's the complexity of Basel that encourages firms to do that. I think we've seen, through both very complex and incredibly simple rules—with no disrespect to any of the bankers in the room—the incentives to mask the financial condition of the firm are there. And so, the question really becomes—well, one question is, what enables firms to do it? And one of the things that's interesting in Basel, which started when I was there and has continued, is an interesting debate about consistency in Basel, which a lot of us have seen, and we see the conclusion on the view of consistencies—we need simple rules because they're more consistent.

So that only is true if you pick one-half of a definition of consistency. So the people who cite that more simple rules promote consistency start with the idea, which I think most of us agree, that if two firms are taking the same amount of risk, and have the same resources, they should report the same capital ratio. No argument, and you might say that—given the question, given the complexity of Basel—it enables firms to mask the risk taking they have, given the complexity the nature of the rules.

But that's a very one-sided definition of consistency. So without answering the question, I would say, what's the other side of the definition of consistency? Well, two firms that have very different risk profiles—different risk taking, different resources—should report different capital levels—that's another version of consistency, is that you should have a set of rules that allow you to differentiate the nature of the risk taking between two firms.

So I think the question sort of gets to a very one-sided answer, which says, well, if it's too complex then we won't get consistent answers. And what I would argue is that if it's too simple, you also won't get consistent answers because you won't satisfy the second side of that test.

Herring: I have a slightly different view. I certainly agree with Marc's observation, but it strikes me that the very complexity of the rules makes it more difficult to detect arbitrage. But I think this is a problem within firms, from time to time. If you look at the huge problems UBS had, it was basically employees gaming the regulatory capital rules within the system to disguise it from management. It sets up a very, very complicated incentive system that's simply very difficult for anyone to cut through in a way that gives you meaningful answers. I certainly agree that there's probably no more of it than there was under the simple Basel I.

It goes back, in a way, to an observation Peter Cooke, I think—the original chairman of the Basel Committee—made that we've got a choice: do we want to keep crime on the streets where you can monitor it? Or drive it into the dark alleyways where we don't know what's going on? It's a pejorative way to put the choice, but there is something of that element. Institutions will always try to arbitrage, and the question is, are we better off knowing how they're doing it, or making it so complicated that it's very hard to detect?

Natalucci: Barbara?

Novick: Since BlackRock's not a bank, I'm not going to opine on Basel.


Natalucci: Okay. So, regulators compete with each other and don't share information. Is it realistic to expect things to get better? And there's a related one here, which is, does the Basel consensus process encourage more regulation, as members are less willing to say "no" to each other so they can get what they want, in terms of the incentive of regulators?

Novick: There's definitely a competition amongst regulators. I would say some of it is natural because of the seat they sit in. So if I put on my hat—if I were a regulator, you can say you were one [motions towards Marc Saidenberg], but I haven't been—but if I were one, what's my biggest fear? Something wrong is going to happen, on my watch, in my area, right? So, by definition, they come at each problem from their perspective of making sure nothing happens on their watch, in their specific area. They're not necessarily thinking of it more holistically. They're not saying, "Well, if I need this data and it's really similar to what someone else needs—well, we should harmonize that." They're saying, "I just know what I need" and that's the highest priority.

So I think you do need some way of breaking through that, whether it's from international standards or at the national level. There has to be some mandate that comes down politically that says, "No. We're not going to have all this fractured data, we're going to get the benefits of harmonized data"—because that actually benefits all of them.

And I guess in economics it would be called some sort of weird commons problem.

Natalucci: Can I ask a variant of this, maybe? Is there a difference in competition among regulators from the different parts of the credit cycle? So, going into the financial crisis [versus] coming out of the financial crisis?

Saidenberg: So, I'm not sure how it changes over the cycle as much as... I think the degree of cooperation and coordination you get across supervisors is a very... There are a lot of things that drive it.

I'll take one that's entirely within the U.S., which probably isn't the nature of this question—for years the Federal Reserve and the SEC—or I should say, the Federal Reserve, the FDIC, and the OCC (the banking agencies) and the SEC—had very different concepts (and they still do, at times) about what the nature of supervision and regulation was. One was a question of largely a prudential conversation about safety and soundness, and the other was a question of enforcement.

And so there was a very odd absence of information sharing, where the view was if you take—there were Fed supervisors in a room. Most Fed supervisors think access to internal audit reports is a critical part of doing their job. While there are some differences in incentives, it is, within the firm, the group that has the most closely aligned incentives with the supervisors—and you should leverage the work they do.

And I think you'd also get most Fed supervisors to say, "We have absolutely no interest in having internal audit share its findings with the SEC, or with an enforcement arm of the SEC, because that will have the unintended consequence of actually leading internal audit to be reluctant to actually put its findings in writing, because if the SEC turns around and uses those findings against the firm, as part of an enforcement matter, then you sort of, you have the unintended...

So I think this question of information sharing gets down to one, and the behavior among regulators is, there are a lot of things that will drive this cooperation—or lack of cooperation, or information sharing—I mean, I do agree with Barbara that the more that we can standardize data, the better off we are. Having said that—and Jai Sooklal, who is here, and Jai and I had one of my favorite meetings while we were both still at the Fed. As we sat down, and there was a mandate from many people, including the industry, that the regulatory agencies should coordinate on their collection of information about funding and the nature of liquidity risk. Everybody in this room, I think, would agree that is a good idea. Why, then, five years later, it hasn't happened— well, the reason is, we sat down in a room together—in this case it was with the PRA [Prudential Regulation Authority]—and they said, "Here's how we plan to collect the data. Can you agree to collect it the way we want to?" It didn't go very well after that.

So, five years later, the PRA and the Fed still have different information collection on liquidity. And so it's really this question of you're talking about a number of very, in principle, aligned agents, but you could ask the question, there are many questions on data standardization in the industry, where you don't get standardization in the industry, for the same reason. Everybody would agree it's a good idea as long as the standardization occurs in the way that is most aligned with the way I currently maintain the data.

So I don't think there are any worse off—and Barbara, I agree with you: better alignment, consistent data standards would be a plus. I think it would enable systemic risk management more effectively, but some of the behavioral constraints that apply to the industry need to also apply to the public sector.

Novick: I used to have a dream that there'd be a single data repository. I've given up that dream. I'm now only looking for a harmonization of the definitions, and standardization of the reporting.

Herring: There is some standardization that, I think, could and should be accomplished pretty easily. We now have massive data sets on G-SIBs that are being used to actually identify who they are. It would be very nice if even those data—they should be more standardized, I suspect, than they are—but if those were publicly displayed, it might be a step forward.

But going back to the question about, does the cycle make a difference? I think it's very much true that where you stand in the Basel Committee depends very much on where you sit. If you look back to the Germans over the cycle, they were of course original participants and they held out in the original negotiations to have nothing but equity (defined as regulatory capital), because they thought letting anything else in—innovative instruments of any sort—was undermining the rigor of German supervision.

Well, flash forward to 2016. What's the position of Germany? Well, they're pretty light on equity, so they're now arguing pretty strenuously in favor of all kinds of other sort of equity things that would make it easier for them to apply.

So I think you've got to expect that countries are going to be arguing their own self-interest in these situations, which also makes it very difficult to standardize, because the U.K., I'm sure, has a very good rationale for the way they want to think about liquidity. We feel we do, too. There's no higher authority to crack heads together and say, "Sort it out and give us an answer." And that's just not in the nature of international coordination among sovereign countries.

Natalucci: Why eliminate OFR [Office of Financial Regulation]? That seems to be a very popular question here, with lots of votes. Isn't that literally the least controversial, but probably the most welcome, regulatory change?

Novick: I'll just give one example of something, a project that's going on right now. The SEC completed their rules on reporting for both mutual funds and advisers. Those new reporting standards include securities lending reporting—very extensive—and yet there's a pilot project by the OFR to redefine the reporting and have a different set of reporting standards, to a different group. I don't really see the point.

So I think we need to decide which of our own regulators—forget international coordination—within the U.S.—what does the SEC do? What does the CFTC do? What does the OFR do? Can they leverage each other rather than replicating or overlapping and duplicating each other?

Saidenberg: I don't have a good answer for why.

Natalucci: Okay. So, what is the best case example you're aware of, of optimal cooperation? What lessons can we draw from that going forward, on a positive note?

Herring: In my view, probably the most valuable thing that has come out of the whole Basel effort has been the personal relationships that the supervisors have with one another. I think it inevitably makes it much easier—it's not perfect, but easier—for information, both hard and soft, to flow across borders, and it really sets up a framework for cooperation when it seems necessary. I think that structure, those relationships, are probably more important than any one particular core practice or minimum standard paper they've released. I think it really is important that these people know each other well enough to be able to at least get a head start on collaboration in a crisis when it's needed.

Saidenberg: So, the example I would give—which, interestingly, is at risk right now—one of the changes that the Basel Committee made to market risk capital is actually that market risk capital related to both trading and counterparty risk should be a function of stress volatility. And it's not exactly an example of cooperation, in the sense of—but it is a case where the committee took a look at the regulatory framework and tried to understand why the same group of people had failed 10 years earlier to devise an effective capital regime. You would say that the market risk rule had a lot of attributes we think of as good: a good regulatory regime, it was very much sensitive both to the current market environment, it was sensitive to position taking on the firm side, it relied on firms' estimates of their own sensitivity of price risk—these are all good aspects. It was a model that was actually—particularly in the U.S.—it was a set of models that were subject to approval by the regulators at a very granular basis. It was a set of models that were tied to the performance, it was regulatory capital that was tied to performance.

The challenge is, all that made sense except for one thing. It was all predicated on the assumption that volatility could change in the environment—or, I should say, banks could change the level of capital they held and adjust their risk taking in line with changes in volatility in the market. I say it now, and I sort of laugh even at my sentence—we all know, and we certainly experienced in 2008, that volatility could change a lot faster than banks could adjust their position taking or, in particular, their level of capital.

So this is a case where I think moving to a world which said, "While it is painful at times, it requires seemingly a lot more capital than the current environment would call for"—it was the effort to actually make that aspect of the capital regime more time consistent.

Novick: So I would love to answer this, because we get so many questions about being a global company and what's going to happen in this global Armageddon when a large asset manager of global stature "fails."

So I think you have to start by unpacking every part of the question, and understand: none of it applies to asset managers. Asset managers are not using their own balance sheet. Asset managers are not the counterparties to any of these transactions—not to trades, not to derivative contracts, not to anything. So there actually is no cross-border impact. In addition, I think fundamentally, asset managers don't fail the way a bank fails. They don't have the exposure in terms of deposits and retail funding. They don't have an inventory that they're funding using repo or other short-term funding sources.

So the whole question actually is in the "does not apply" category. And yet I can't tell you how often we get questions about cross border and cross-border exposures that actually don't even make any sense to the business model.

Natalucci: All right. So I have one last [question] to close the session—we are one minute over time. Looking forward, how can transparency and accountability in international processes, generally speaking, be increased?

So for example, we have an FSB paper out.

Novick: Okay, so I'll give some examples of transparency. The FSB—I have this great org chart that I've put together of the plenary, and then there's the next group—you could tell me all the names of all the right groups.

Natalucci: Steering committee.

Novick: Steering committee, thank you. Then there's the SRC [Standing Committee on Supervisory and Regulatory Cooperation], then there's the SCAV [Standing Committee on Assessment of Vulnerabilities], then there's another committee—you remember the name of that one, too. Then under the SRC, there's six work streams—and I forget how many others—but you get the point, right? You have this incredibly complex organization which was started postcrisis, and the assumption is, everyone should know how it works. Most people didn't even know it existed or what it meant in asset management. OK? Most people didn't even have the vaguest idea that they were there, and the insurance world wasn't much better. There were people in the insurance world—when they got their designation by fax, it was the first time they knew that this thing called the FSB was doing designations of insurance companies.

So the lack of transparency, at even the organizational level: who's on which committee? Keep it current. What are the work plans of the different committees? Why not tell people in advance? Why not have the meeting schedules? These meetings are all over the world. We get notices on a regular basis, couple of weeks before an important meeting— "We'd like you to come in. Send your best person." Now, how many of you, on two weeks' notice, can fly to another continent? Because, it's really important and you want to go. A lot of times, we drop something else to do it, but it's not sensible. They know about these meetings in advance. They have the meeting schedule. It's part of the work plan. I've seen the work plans. OK?

So why not just be more transparent, be more inclusive? Why not have meeting minutes? Why not post the subjects, the agenda, and the minutes? And not in an American "Boy-we-need-the-Sunshine-Act-and-we-need-the-APA-we-need-everything-under-the-sun" [approach], but just basic governance—good governance, good transparency, be more inclusive, include people upfront who are practitioners. Don't think of it as a supervisory event where you want to keep the bad party in the dark until the end. This should be a cooperative, work-together [situation]. Our basic interest as an asset manager is financial stability. We don't want bombs going off. We really want better market...

When people say, "Securities regulators and banking regulators don't see eye to eye," I laugh. I say, "What do you think 'investor protection' is?" It's said by banking regulators, like, "Somehow, securities regulators, they only care about investor protection." Well, I think the most basic thing of investor protection is stable markets. So, they're using different language but they actually come to the same thing that's important to them. And as an investment manager, we care a lot about investor protection, and the implications for a fair and effective market.

So I think there just needs to be more engagement early in the process—not when a conclusion is done, but when people are really thinking about hypotheses and trying to test them early.

Natalucci: Just for the record, because we spent a lot of time explaining this, the FSB doesn't designate, it identifies. Only FSOC designates in the United States.

Novick: Right; and they also have no enforcement power, just moral suasion and "name and shame."

Natalucci: Marc.

Saidenberg: So, I would actually pick up on something Barbara said. I'd go a little further than you went on one point—I think the four principle standard-setting bodies—the FSB, IAS [International Accounting Standards], Basel Committee, and IOSCO—their meetings should be minuted, released with a lag—but I can tell you that, having sat in one of those committees for five years, there are people who said stuff in those committees where you can only say, "If only people outside this room knew what you just said." And the only way to combat that is, similar to what the FOMC does—you know, lag is fine, but if people knew that the arguments they were making in that setting, on behalf of their agencies or otherwise, were going to be public, I think you would have the accountability. You would get that sort of corporate governance discipline you're looking for, that people would know that the arguments they're making are going to be recorded and they're going, with a lag—they're going to be made public.

Natalucci: Last word?

Herring: There's one very important decision the Basel Committee made that was not at all transparent, and I think it had a huge impact; and that was the decision, which seemed like a technical issue, about how much you could change tier 1 capital between the equity portion and other innovative instruments. At one point in the late '90s, they lowered the amount of equity capital to half of that, or 2 percent. If you then look at the leverage, as it usually computed—and bear in mind that risk-weighted assets are, say, 50 percent of total assets—they were implicitly authorizing a leverage ratio of 100:1.

Now, if they'd reported it—if they'd had some transparency—we might well have had some ability to criticize and ask, "What in the world are you thinking?" It really doesn't make any sense in any kind of context, and I don't think they knew what they were doing, and I think it was something they could do as a matter of technical adjustment, and simply not have to worry about it.

Saidenberg: Dick, just one thing. The only thing I'll say is that rule in the U.S. regulations was actually put out for notice and comment, and made completely transparent in U.S. rules. And you'll be shocked to learn that not a single commenter asked for a higher level of capital.


Herring: That, I think, is a problem, because the complexity of regulation has gotten to be such that no disinterested party will take the time to pull it apart. So the only people who have any real incentive to look at that are people who would benefit from lower capital requirements.

Saidenberg: Yes, but just because—this was in the simple days. This was way before Basel II.

Herring: Yes. Well, this was in the '90s, and the global standard, and it's just not in the record anywhere.

Natalucci: Thank you very much to the panel participants. Thank you for your questions, and I guess we are adjourned now.

Thank you.