Earlier this summer, The Clearing House and SIFMA co-hosted their annual Prudential Regulation Conference, which brings together leading experts from the public and private sectors to discuss the role of regulation in today’s banking system and explore the regulatory developments that are affecting U.S. and European capital markets. Steven Strongin, head of global investment research at Goldman Sachs Group Inc., delivered a keynote address that analyzed a range of post-crisis financial reforms, identifying both core reforms that have made the financial system safer and non-core measures that entail costs that far exceed any marginal benefit. Most notable, Mr. Strongin’s presentation described how new regulations have raised borrowing rates, priced many low-income consumers out of the credit market, and led to small firm employment and wage growth to lag compared to larger firms.
Below is a transcript of Mr. Strongin’s remarks and the question and answer session that followed his presentation.
GREG BAER, president of The Clearing House Association: It’s my pleasure today to introduce our lunch keynote speaker, Steve Strongin. Steve is a partner and head of Global Investment Research at Goldman Sachs, where he also serves on Goldman’s management committee. He also serves on the Client Business Standards Committee and the Technology Risk Committee.
Steve rose to this level by service in a variety of research functions within Goldman over the last almost 20 years now.
Before joining Goldman in 1994, Steve served 12 years at the Federal Reserve Bank of Chicago, rising to the level of Director of Monetary Policy Research. Indicating his fertility of mind, he had previously served as a researcher at a group called Pan Heuristics — there could be no nerdier name — where he looked at energy policy and military strategy.
Lest you think that Steve is now sort of an analytical machine but at one point was sort of living a carefree and fun life, he actually has multiple degrees from the University of Chicago, the place where fun goes to die.
We did ask his staff and others for some fun and interesting anecdotes about Steve. Regrettably, there were none.
Actually, there was one. One of his staff said, well could you say that he’s speaking at a regulatory conference but he’s not a lawyer? So that is the funniest thing that can be said about Steve.
So all kidding aside — I wasn’t kidding — Steve is a joy to have at this conference. We often, within The Clearing House, ask the question, “who is looking at the whole battlefield of regulation, and all of the market and business impacts of that?
And there are regrettably very few people who are willing and able to do that. Steve is certainly in the group of people who can.
The level of his insight, his knowledge and also his willingness to do the hard work to figure things out, I think, is unsurpassed in our industry and, I think, in the regulatory world. So it is, I think, a real treat to have him here. I think you will enjoy very much hearing from him today.
STEVE STRONGIN, head of global investment research at Goldman Sachs Group Inc.: I’ll have to think about that introduction for a while and think of some suitable revenge. Hello, everybody. It’s a pleasure to be here.
Listening to the conference this morning and these discussions over the last few years, I’ve been struck by one particular thing over and over again, which is that about every 90 minutes somebody observes that there’s a tremendous amount of interconnection between these rules so we need to look at them as a whole, and that coherence is a very important priority.
And they usually then mention a regulator who has said ‘it is now time to re-assess the rules now that we have accomplished this much.’
After that sentence, nothing else happens, other than a return to what I think at this point we best could refer to as a knife fight about specific rules. And every six months or so, the number of rules we’re arguing about increases by about 30 percent.
I saw a chart yesterday that compared the Basel III agenda to the Basel number-must-not-be-named agenda. I call it 3.1. Maybe that won’t cause as much trouble. And there were more items on 3.1 than there were on 3.
And then when you listen to discussions today, you realize that you could have subdivided most of those line items considerably.
I think the real reason is that we lack a nomenclature and an intellectual structure to talk about a rule-set that exists across multiple pieces of legislation, multiple continents, multiple aspects of the banking system and multiple aspects of the financial system more broadly.
And a rule-set in which most items are subject not to one or two rules, but to seven or eight rules, where it actually becomes pretty hard to do an analytic analysis.
So the question becomes, how can we reorganize this in a way in which we can have a coherent conversation about the marginal impact of the newest rule and whether it makes any sense or not?
If you listen to the political debate today, it is divided into two camps. On the one hand, we’re tired of this, just kill the banks. On the other hand, we’re tired of this, just repeal the whole thing. Neither of those two is probably going to work.
And so how do we find some structure in between? I think the first essential step in that is to ask what the core reforms are. We’ve been through a financial crisis. Lots of stuff went wrong. We’ve done a lot of things to fix it. What sits at the core of those reforms and what kind of safety did we get from the core rules?
When you ask that question, the answer is reasonably straightforward. And that provides a baseline for asking what all the other rules are supposed to be accomplishing and how we measure their impact. My presentation today is a modest first step in this direction.
Every person I have talked to has three or four separate rules they would like to call the ‘core.’ I’ve chosen fairly abstract and broad language to describe the core reforms and it will become clear why in a moment.
First the equity capital rules. If you go back and look at the behavior of banks and the brokerage houses in 2008 — because remember, in the financial crisis, those were separate classifications with a few people who did both – you see that one of the key differentiating factors was not a lack of capital, but the nature of the capital they had.
Those institutions that had a high degree of common equity making up their core equity capital behaved fairly conservatively. Most of them did not have significant troubles across this period. And when the troubles did begin, they all recapitalized or conserved capital because that was what was in their shareholders’ interest.
If you looked more broadly across the banks to those that had a higher percentage of their normal equity capital in more obscure instruments — preferreds, the various flavors of preferreds, sub-debt — the incentives were much more complex.
In a number of those institutions, you began to see situations where after losses, the common equity was actually very light. And in a pure finance sense of the word, you saw situations where the option value of the firm became larger than the intrinsic value of the firm.
And in many cases those firms, instead of taking capital-preserving actions, were actually still buying back shares, still paying dividends straight in to the crisis. And in some cases, the capital that was drained from their own capital actions was larger than the economic losses.
So one of the core changes that the regulators made, very quickly, was to require that equity had to be common equity, and that banks had to raise common equity. That’s a really big change that has kind of gotten lost in the ongoing debate.
The second big change that occurred during the crisis was SCAP, which is now called CCAR. SCAP was necessary because nobody trusted banks’ balance sheets.
It seemed that every day – although it was more like once a month — some large institution would admit that its balance sheet substantially misrepresented its risk in some form or another and this created a kind of information contagion across the system.
It also turned out that for those institutions that had large risks that weren’t on their balance sheets, their capital was probably more questionably calculated.
So what SCAP did was to bring all of the off-balance-sheet risks onto the balance sheet and essentially forced everyone to state their balance sheets in common terms and allowed the market to assess them in a coherent way.
And as we think about CCAR today, we lose that simplicity as CCAR has picked up so many functions — both political and economic, some good, some bad — it’s hard to either be for it or against it.
Today’s announcement that the Federal Reserve is considering including new surcharges in future CCAR exams is another example of another purpose that CCAR is beginning to serve. But we need to remember the basic core original purpose was an informational one. We’re simply going to get common balance sheets that really have everything on them in place.
The next reform that I would refer to as core, which is now going into effect and which serves a very similar purpose to SCAP, is the margining of OTC and counterparty risk.
One of the other concerns during the crisis was that the concern that counterparty exposure might take down even banks with equity capital and transparent balance sheets. This is the AIG risk.
In fact, if a bank’s exposure to a counterparty is appropriately margined, then the bank really only has market risk – it already has today’s risk in-house. The only risk that arises on the failure of the counterparty is the market price change that can occur between that time and closeout. So once you have consistent margining across the board, the information contagion from counterparties goes away. And again, you get greater faith in the balance sheet.
When you look at those first three reforms, you will notice that none of them refers to the level of capital. All these reforms really did was to correct the information problems that occurred before the crisis. Effectively, the message from regulators to banks was: ‘We’re going to make sure that your equity really is equity; we’re going to make sure that your balance sheet really is your balance sheet, and we’re going to make sure that your counterparties aren’t somehow a hidden claim on your balance sheet.’
These are really basic economic principles, non-distortionary, very clear in their intent and very aligned with economic reality. So that if one was going to ask what are the economic costs of those first three core rules, the answer is probably going to be de minimis.
In fact, numerous regulators and others have argued that if an activity was occurring in excess of those three rules, it was probably bad activity. So getting that activity out of the system is probably a good thing.
You really don’t want to defend what AIG was doing during that that period. You don’t really want to defend what was going on in the special purpose vehicles. That probably wasn’t good economic activity.
And in fact, if one wants to blame the banks for the crisis, one could blame the housing finance that helped to inflate housing prices.
So those three core reforms are really about solving information problems. They’re really about allowing markets to do what markets should do and also making sure that regulators have really clear information on the banks.
The calibration question naturally has to be one of cost-benefit analysis. So before I get to that calibration, I want to look a bit into where the possible problems might be on the other side.
I will make an observation before we get there that no G-SIFI — I want to be really specific — or candidate G-SIFI would have failed during the crisis if those three reforms had already been in place. And it’s unlikely that any of them even would have needed government aid. I’ll come back to that in a little bit.
The fourth reform is the one that in some ways I actually consider the most important. This is the requirement that capital instruments become junior to operating liabilities. For anyone who has been through a financial crisis on either side of the wall, the biggest single political and operating problem is the risk that customers run away because they are frightened. That destroys your business.
And if you’re a regulator, this risk is also what stops recapitalization. You’re worried about hurting bank customers. So the fact that that, historically, the line between capital and operating liabilities has not been a distinct line but has been a mix, has always been a real sore spot in periods of stress.
It creates problems for the regulators because it makes capital actions hard. Taking capital actions hurts bank customers and hurts the economy. And from the bank’s side, it’s very hard because banks lose customers when people get worried.
So cleaning up the capital structure, which is what we did through TLAC [total loss absorbing capacity], really smooths the overall process of recapitalization. And depending on your calibration of those numbers, TLAC also provides the core funding that allows all the rest of the rules to work.
When we talk about resolution, the lawyers will inevitably begin to talk about the resolution process, the living wills and all the other things that generate billable hours in high numbers. And those things are really important. I don’t want to minimize them.
But it turns out to be a lot easier to make the legal stuff work when the numbers add up. When the numbers don’t add up, getting the legal stuff to work requires tremendous creativity and pain.
Guaranteeing that the numbers will add up goes a long way toward getting the rest of the stuff right.
The other reason I want to focus on these four reforms is that they are the ones that most relate to the ability to calculate risk. When you start talking about supervisory standards, when you start talking about resolvability, living wills and that stuff, many of the determinations are highly subject to opinion.
And if you ask an economist who’s trying to do a risk study, or a statistician trying to do a risk study, what has that done to the probability of failure; what has it done to resolvability; they don’t really have an answer.
When I move where my capital line is between common equity and equity, I know what to do with that in a simulation. When I raise the total capital line, I know what to do with that in a simulation. When I get to assume that the balance sheet is a balance sheet, I know what to do with that in a simulation.
So part of the reason I think the core regulatory reforms need to focus on the most numeric stuff is that when we go to do the safety versus economic cost analysis, we want the fundamentals of the core forms to feed into those simulations and cost calculations well.
We want to be able to say, here’s how much safety we’re buying with this rule. Is it a lot? Is it a little? What’s the incremental marginal amount of safety, and what’s the incremental economic cost we’re imposing on the banks, on the economy, on the financial system?
And I think, when you do that, one of the things you’re going to realize is that you’ve rebased the debate rather substantially.
Now what you have here is the capital structures of sort of a selection; all the G-SIFIs, all the large failed institutions and what could be considered the sort of high-risk D-SIFIs of the time.
This chart gives us an idea of what their capital structures look like. The red dots are the cumulative losses over the period and the green dot is the quick short-term hit, or the largest drawdown in statistical sense.
So those would be the two classic questions you’d do for a simulation. Those were the capital structures of the time. And what you have here is the new capital structures that are required.
A couple of simple observations are worth making. First, for the G-SIBs — at current capital structure, no G-SIB would have been in trouble in 2008; period. The only banks that would have had a problem were D-SIBs in 2008. And for those of you who like numerical analysis, I will point out, this would suggest that the G-SIB surcharge has the wrong sign from an economic perspective.
By the way, this is a very typical pattern of crisis. You typically find that diversified institutions have a lower risk profile than specialized institutions. Most of the real losses that have occurred during crisis have occurred in specialized institutions, not in the highly generalized ones.
So then, when we think about the G-SIB surcharge, we really need to think about the social costs of resolution, not the risk of the institution, and we need to offset these costs against how we change the social cost of reorganization.
The basic point is that when you begin to look at that, you begin to realize that the core equity is more than sufficient — under just those four reforms — to have dealt with the financial crisis plus. That’s not allowing for any recapitalization.
Remember, most of the firms that had decent equity capital did recapitalize. Most of them had earnings. They were in a pretty good place.
The other thing to understand, to get a real sense of how much the world has changed, is the role of long-term debt.
Before the crisis, long-term debt was pari-passu to the bank’s operating liabilities, which meant if that if regulators wanted long-term debtholders to absorb losses, they would have had to shut the bank, unwind the bank, do major issues to bank customers, have real economic impact.
Under the new rules, the entire stack of long-term debt is loss-absorbing. This means that the true amount of loss-absorbing capital today is a multiple of what we saw during the 2007-2008 crisis.
So then, in a mathematical sense, when you begin to look at the core reforms, you begin to realize that the right marginal test, when we look at the additional regulations, really can’t have to do with the systemic risk seen in 2007.
The standard argument, which you’ve occasionally seen in some government reports, is that avoiding the $2 trillion loss from the financial crisis is worth any regulation. But we can pretty much get that benefit out of the first four core reforms.
The regulations beyond the core are going to have to stand up to a tougher test. What incremental safety? What market impact? What particular aspect of the system are they making work better? They’re not going to justify avoiding another financial crisis — this cannot be the answer to that question. And so the economic calculations become much more refined at that point.
The other thing is that the CCAR test as currently structured does not allow firms to assume that they can raise capital. Well-capitalized firms under stress raise capital. They did during the financial crisis. It’s the poorly capitalized firms that don’t raise capital.
When you couple that with the new rules and you throw in the fact that in CCAR, a poorly capitalized institution would need regulatory permission to do the capital actions, you realize that we’ve understated capital one more level — that we really are going to get a lot better behavior in the next crisis from the banks. Most of them are going to be very well capitalized.
It’s possible they will even raise capital during the next crisis, on a voluntary basis. But certainly, in no case are we going to see the behavior we saw during last financial crisis, where the highly levered risky institutions actually did buy-backs and dividends to reduce their capital, which generated a lot of the problems. So things are really quite different in that regard.
Now I want to shift to the costs of this. This is a place where I think the industry has done itself a bit of a disservice in that the alarms here, much like the $2 trillion justification for every regulation, is the general case of ‘this will hurt the economy,’ which is a little too general.
If you want to think about taxation — that’s what this is. This is a tax on banks and it’s a tax on banking. And when you think about the general issues of taxes — who pays taxes? Not – in general – the person who they were levied on, but rather the individuals who don’t have good substitutes.
If I put a tax on Burger King, that’s going to help McDonald’s a lot. In that case, Burger King is going to pay the tax. But if I put it on both firms, then their customers are going to pay the tax.
And that’s what we see generally: if you think about how we have seen the effects, it is not probably on general growth. Although I’ll come back to an argument that says there may be some general growth implications.
The impact is really distributional. This hurts those individuals who are dependent on banking services and those companies that are dependent on banking services and –- on a relative basis – helps those that are not. And so when we want to understand the consequences, we need to understand the distributional consequences, not so much the total macro effect.
And you see this in this graph. All we’re really doing is showing you what’s happened to the cost of financing for varying activities before and after. Here’s what it looked like before the crisis, here is what it looks like now on average.
What you discover is that those activities that the rules have favored, large corporate issuance and conforming mortgages that are run through the government are actually cheaper today.
What’s become more expensive is the activities that people really need a bank for. So the midsized corporate market has gotten a lot more expensive. Small business finance, incredibly more expensive. Low-income consumer finance, a lot more expensive. High-income consumer finance, not so much more expensive.
In fact, low-income consumer finance has gotten so much more expensive that we can’t put it on the graph. Roughly 14 million people simply lost access to credit cards. Those that did maintain access did so at a few hundred extra basis points.
But the vast majority of that group moved into payday lenders and consumer finance companies, which would be somewhere up near the ceiling of the chart, and which are now subject to a new group of regulations, or simply a lack of the ability to finance. And calculating the cost of that is somewhat more difficult.
And so, as we think about where we’re going to look for the cost, it really has to be where banks are performing critical functions and it has to be distributional. And we see that really clearly.
Here is where you see the numbers in consumer finance where, if you look at the distribution by FICO score, which is one close proxy, all the cuts in credit have been to low FICO score individuals.
Now it could easily be argued that this is a good outcome, not a bad outcome. There’s a standard argument that this is a debt trap situation. But it does mean, when we look at the situation in consumer finance, that we really need to think about the overall ability of the low end of the market to access financial services and whether access to loans is appropriate. That’s going to be the key question.
From an economic standpoint, this is probably one of the two big cutting edges. Start-up small businesses. This number is off-the-charts awful.
This is a group of individuals who clearly need banks to get started. If you don’t have a business, it turns out to be really hard to go to the market for finance.
These businesses were historically funded by credit cards, second mortgages, personal signature loans. I’m trying to avoid picking on individual regulations here, but there are at least four or five rules that make sure that none of those things happen anymore.
And so you’re about eight hundred thousand small businesses short this cycle, relative to history, based on our estimates for the current period.
That’s an enormous change in the dynamic of the economy.
Almost the entire shortfall in small business is in businesses that are less than five years old. This is a clean change in the dynamism of the economy.
And it’s worth noting, really importantly worth noting, that the market has tried to innovate to solve this problem. It has largely failed. The typical interest cost is between 50 and 150 percentage points.
By the way, that’s not usury. That figure is about right on a probability-adjusted basis. Most small business startups fail. Do you know the line in Dodd-Frank that says you should have a reasonable expectation of repayment in order to make a loan? Small business lending, new startups would never make it past that line.
So that dynamic is really intrinsic to the notion that banks should not take risks and individuals should only take reasonable risk on someone else’s assessment. That’s a deep social question, but it goes to the heart of what bank regulation has been doing.
And it has secondary social costs that are worth noting. This shows you what has happened in employment.
It’s worth noting that the employment shortfall in this economy, up until now, has been entirely in those small businesses. And a lot of the slowdown in wages is also in that same set of businesses.
It turns out that small businesses hire slightly different demographics than large corporations do. This has segmented the labor market. A lot of the weak wages and a lot of the weakness in employment are in small business employment and in those categories. So again, a lot of the social costs fall on business startups, dynamism and in the employment market.
One of the questions we’ve been talking about a lot is the impact of reduced market liquidity .
I’ve been on a number of panels in which the Fed has gotten up and said to a bunch of market participants that there is no problem in liquidity, to a bunch of market participants who spent half their day complaining about the lack of liquidity. It’s an interesting interchange.
The real question is what’s the gap in this conversation. Is it that the Federal Reserve is running its numbers incorrectly? It’s not that the market participants are all deluded about the liquidity of their portfolios; they clearly have to be talking about different things. So what is the difference in what they’re talking about?
Well, what the Fed is talking about is that when you look at the most liquid highly traded instruments, there are more of them and they’re trading more liquidly and more often today.
When a market participant is saying that they have been forced to trade those things because they can’t trade the little stuff.
From an economic perspective, when you look at the life cycle of a company, they’re hard to start. If you can get them started, the sort of bank finance, early stage stuff really hasn’t changed that much. In fact, the senior loan officers and so forth will tell you that banks are reasonably friendly.
At about $100 million, in order for those companies to go that next level, they start going to the markets for public finance, which is cheaper than bank finance. And historically, a lot of the growth and a lot of job creation have come from those $100 million companies becoming $2 billion companies.
What this graph is showing you is that the markets no longer work for those companies. This graph explains two things. It explains where those companies have gone; it also explains why M&A has been so strong.
Because when you are a banker with a client with $100 million company and you run your spreadsheets, instead of going to market in an illiquid market at a high selling commission, you will instead sell the company to a large buyer who is funding at the other end of that spectrum.
So we’ve changed the lifespan of a company in a fundamental way through lack of liquidity. Because down here is where you need a broker-dealer who’s willing to inventory risk. Up there, you don’t need a broker-dealer.
And so again, we see the pattern. It’s the people who actually need finance where we’re seeing the real damage in the economy. The people who have substitutes find a way around.
And so when you hear the argument that markets will innovate around these problems, it’s true. But you have to ask the question, are we going to like those innovations? Do we like an economy in which the role of small startup businesses is greatly diminished? Do we like an economy where mid-sized companies just feed into large companies rather than growing on their own?
Is that an important dynamism for our economy or is that something that’s easily given up because large companies are just fine and we don’t mind that process? And that’s another gap between the two conversations.
Because if all you care about is big companies, markets will find a way around it. If you care about the little companies — what is happening on the regulatory side is having real impacts. It’s changing the structure of the economy. It’s making it harder to be small. It’s making it harder to innovate. It’s making it harder to succeed. It is creating advantages for incumbents.
Short-term, that probably has very minor impacts. Long-term, it could have very large ones. It could take 10 or 15 years to figure that out. How many years of no business startups do you need before you’re missing the Googles and the Facebooks and the Intels? Because it’s the next generation of startups that’s going away. The current generation is doing just fine.
And so that’s how I think we have to think about this. The way I think we need to organize that argument is as follows:
We’ve got a bunch of rules. First, we take the core rules and we really begin to do the calculations relative to this core. And then we’ve got to ask the following questions; is this getting us much safety and what’s the economic cost?
If it’s really getting safety, the way I would argue that TLAC rules are, as an example, and it doesn’t have much economic cost — and I’d probably put TLAC roughly in that category — it’s a keeper.
When the rules have high economic costs but don’t generate much safety — I’ll throw one candidate out there, the SLR [Supplementary Leverage Ratio] — we really probably ought to think about whether that really is a good idea or not and begin to segment the rules in a rational way and then begin to think about the rules that are expensive but yield limited benefits and see that we really need to do something there.
These are areas where everyone would agree that change is appropriate. I think that’s how we need to organize the conversation and recognize there are real benefits and real costs. But they don’t necessarily come from the same rules. Thank you very much.
BAER: The first question is just picking up where you left off. You gave one example of an underachieving rule. Are there others or is it actually more a question of rules that make sense individually but where there are multiple rules doing the same thing?
STRONGIN: There was a comment made by one of the regulators earlier which I thought was central to the problem, which was to say we like having lots of different ways of measuring capital because that way we’re not too bound to any one of them.
There is a sense in which I agree with that statement and there’s a sense in which I’m deeply frightened by it.
If you have a primary measure of capital and then you have nine or 10 ways of checking whether somebody’s playing games with that calculation, that’s probably a really good idea.
If you, on the other hand, have nine separate sets of rules, some of which are risk-based, some of which aren’t risk-based, you’re going to get really complicated calculations, really strange incentives. And it’s going to probably be counterproductive.
So when you begin to look at the rules, I think we need to get better about segmenting them into primary and binding, and advisory and informative. The capital ratio is probably a simple example.
I think it’s fine that everyone report 17 capital ratios. I think they’d probably only be optimizing two, at most. And the others should generate supervisory looks over the shoulder instead of something more than that.
And then if I go deeper into those, the capital ratios that I think are counterproductive, it’s all the non-risk-based ones. It’s all the ones that say secured lending using incredibly good collateral at an over-collateralized basis on a 99 percent, two-week basis still has risk. Those are the rules that you have to wonder about.
The notion that we have a hard leverage ratio bound on the entire system. Hard leverage ratio bounds make no sense under stress.
I can still remember. I was in the Fed in October of 1987. In October of ’87 we spent a great deal of time calling up banks telling them that they were going to lend to every broker-dealer who wanted it and every customer who needed it, that we weren’t going to have the system dissolve because people couldn’t make their margin calls.
Under today’s rules, if you made those calls, I suspect you’d reach outside counsel. And if you managed to get somebody in the bank to answer the call, they would tell you it’s hardwired into the systems and even if they wanted to change it, they couldn’t. Because we’ve built lots of hard lines in the system and you really have to wonder about the value of those.
You’ll also notice that I didn’t have any liquidity rules as a member of the core reforms.
BAER: That was going to be my next question.
STRONGIN: The reason is fairly straightforward. In a well-capitalized institution where an Opco is senior to the HoldCo, it’s really hard to come up with a coherent situation where the Opco can’t borrow.
One of the problems that you’re beginning to get in the living will process is the fact that coming up with a coherent scenario under which most of these banks can fail is requiring greater and greater creative writing skills.
And liquidity is one of the places where this challenge is really hard. The rules keep changing in ways that make sure the Opcos need less funding and are easier to fund.
A couple of layers of liquidity rules basically require that the derivatives books now unwind to positive liquidity. So if a bank manages this under stress, it’s going to find itself awash in extra cash and unable to do anything with this cash.
That set of rules is essentially solving a problem that the other rule solved a couple of times before you even got to liquidity rules.
BAER: So in addition to solving problems that have already been solved, do you see out of the liquidity rules unintended consequences akin to what you’ve been highlighting for capital?
I saw recently an analyst report that described NSFR as “not suitable for repo,” another one talking about how, as the Fed’s balance sheet normalizes, the NSFR could become a cap on loan growth. Would we see the same sort of consequences for liquidity rules as we have for capital?
STRONGIN: Kind of. Let me separate that question into two parts. I don’t think the liquidity rules represent a substantially different issue. It’s one more tax on banks providing funding.
The simplest way to think of it is that loan costs are going to be the cost of your deposit plus a fraction. If you also have to hold expensive capital, that cost has to be added on to the price of the loan. Therefore, lending is going to be more expensive. Therefore, bank lending is now more expensive. I don’t think that’s exceptionally different. Then each successive round of CCAR raises everyone’s capital one more layer or two.
I think the real question is going to be the way the rules interact with stress. Typically, in stress periods, the banks are the provider of liquidity to the system. It’s not clear with those ratios whether the banks are going to be in a position to provide liquidity to the system. And what’s the perverse response to that?
A number of central banks, basically, have now set up facilities to be the broker-dealer of last resort as well as a lender of last resort.
The Fed has now set up, basically, a repo facility of last resort. The Bank of England has been broader than that and the ECB is now a very large dealer in corporate debt. And so, you’re essentially forcing the central banks directly into the markets.
I don’t know whether to call that intended or untended. I don’t even know whether to call it good or bad. It definitely doesn’t look like what people had intended all of this to result in.
I think if you asked any of the reforming politicians whether they think it’s a good thing for central banks to become the broker-dealer of last resort, I’m pretty sure the answer would have been no. I’m less sure that they would have objected to each of the steps that got us here, and that’s part of the reason I think the reassessment is really critical.
We moved a lot of risk out of the banking system. Most of that risk has just been moved to other places and it’s time to assess whether we like where it’s ended up.
BAER: That was a question I was going to ask. If you think of the banks and their affiliates as the Burger King for whom costs have increased, and McDonald’s as the non-banks, not affiliated with banks, for whom costs really haven’t increased very much, given the primary — one might even say monomaniacal – focus on banks and their affiliates, why wouldn’t we see prices actually not go up for consumers over time because McDonald’s is fine?
STRONGIN: For individuals and companies where the market is a perfectly good solution, that’s what we’re saying. Apple doesn’t pay the higher bank lending costs – this is not a problem for them. It’s a problem for institutions whose debt doesn’t trade liquidly enough or doesn’t trade at all.
BAER: Is that simply because the alternative is the McDonald’s, which simply can’t fund as cheaply as banks fund, and therefore customers must pay more, or is it a lack of innovation by non-banks filling those needs? Why isn’t that need being filled?
STRONGIN: The answer is different market by market. In the case of the securities market, there are two solutions.
There’s essentially the equity solution and the equity-like solution, which is that you’ve gone to the equivalent of a dark pool, direct matching of securities, and people trade more slowly. That answer works for large securities that trade on a fairly regular basis. There are other consequences but it does work.
As you go down the liquidity spectrum, those bonds don’t trade very often. Somebody has to warehouse them for trading to occur. Warehousing requires balance sheet. And while non-banks have capital and hold risk, they cannot create balance sheet.
And so, in general, what you have seen broadly across the financial system is that we have successfully moved risk. What we have not been able to do is to replicate the market-disciplining balance-sheet-intensive relative-value mechanisms.
I’ll give you a simple example, where the arithmetic is compelling. Let’s suppose the world has a 10 percent return threshold on capital. Then all I need to get people to hold an asset is to give it a 10 percent return. If it’s a low-risk asset, I need six percent. The world is fine.
Market-disciplining trades are different. I hold a market-disciplining trade because there’s a two percent mispricing or a one percent mispricing or a 15 basis point mispricing.
In order to be willing to allocate capital to those trades, I need to be levered. I still need to get my 10 percent. The only way I can make that work is to do five transactions if of them if the mispricing is two percent, and to do 10 of them if the mispricing is one percent and to do 20 if it’s a half percent.
So when you need someone to step in and hold the relative-value hedge trade, which is what I need in the midcap corporate market, I need balance sheet to be at a reasonable price and reasonably available.
But regulation has made balance sheet more and more expensive. What we’ve essentially done is eliminated the market-disciplining arbitrage activities. We’ve eliminated intermediation.
And while a hedge fund can provide capital, it can’t create balance sheet. The only entity that is allowed to create balance sheet is a bank. And so most of what we’re seeing is the balance sheet constraints, not the capital constraints.
In some other cases of small business lending, it’s a different issue. The reason banks are able to do small business lending much better than the internet platforms is, one, they can see their daily cash flow where the internet platform can’t. So they have a tremendous information advantage. Two, most of that lending is done on a personal co-sign basis.
BAER: What do you mean by they can see their daily cash flow?
STRONGIN: If I’m going to lend money to Agatha’s Bakery, I’m probably also its bank, which means I see its receipts every day as their banker. If I’m an internet platform, I don’t see its receipts every day.
So as a bank, I’m in a much better position to manage my risk on that startup loan and cut it off at a very different point than I can if I’m an internet startup. Second aspect is if I’m a bank, I probably didn’t actually lend to Agatha’s Bakery. I lent the money to Agatha and I co-signed it against her house or I did a second mortgage, which means I have a secured asset. So it’s a different type of lending, as well.
As Fintech has tried to innovate into that space, they’ve had to basically do it as VC lending, which is incredibly expensive. They haven’t been able to do it as sort of amplified personal lending, which is what a bank can do.
So it’s a different kind of loan at a very different price. That lending has not been cut off by the rules on capital. That lending has been cut off by the rules on change of circumstance, disparate impact or credit card rules. That change has been in rules that have nothing to do with the prudential rules.
BAER: Actually, speaking of disparate impact, there was a story this morning. I think it was a Wall Street Journal analysis of bank data, basically pointing out that banks may be caught in something of a trap and that the regulations have pushed them, for example, to jumbo mortgages in the mortgage market.
And that if you look in terms of disparate impact, that is not a very good story. And so they may face a very difficult choice between violating — well I guess you can’t really violate capital laws — but conforming to regulatory requirements, but resulting in a business mix that will subject them to considerable legal risk under the fair lending/fair housing laws.
STRONGIN: That is quite conceivable. But part of the problem with those rules is that they’re high subjective in the way they can be applied. Those rules are basically a question of what is legitimate versus non-legitimate discrimination.
BAER: Not exactly, but I hear you’re not a lawyer.
STRONGIN: Yes, I’m not a lawyer. And I would like to exit on the same note. I’m not a lawyer and I’m done with this conversation.
But let me give you an equivalent category. One of the three ways in which small businesses used to be financed was personal loans. Small business lending below $1 million was basically a personal loan category.
It’s essentially a category that no longer exists because, it probably won’t surprise you, the people to whom banks will loan $1 million on their signature typically are not representative of the overall community.
And so nobody does that lending anymore. That whole field of finance is just gone.
BAER: Getting back to something you referenced earlier: your focus here has mostly been on how regulation is kind of barnacles on the ship of economic growth. But how is the ship going to perform in crisis, in battle?
I think you gave one example around liquidity rules. But I sometimes use the analogy of, you know, large banks as the Maginot Line in the next crisis. Obviously a very complicated question, but how do you think it plays out?
STRONGIN: At a Fed conference, I once raised the following question. Post the TLAC rules, one of the safest places in the world to put money is an operating company of a G-SIFI.
Much like in the last crisis, the top tier of banks actually had money running toward them. This time you will probably have money running toward all the G-SIFIs, certainly all the U.S. ones.
Deposits, even if you take them and deposit them at the Fed, count against the SLR. So there is a fairly large probability that if the next crisis looks anything like the prior crisis, somewhere around a third of the way through the crisis, the G-SIFIs will actually run out of capacity to take deposits.
Jamie Dimon has talked about closing the gates or — raising up the drawbridge I think was the phrase he used and they’ve talked about not wanting deposits.
So the question I raised to the Fed was, what were they going to do when the major banks announced they can’t take deposits? And the response I got back at that particular conference was that they’d suspend the rules, which left me somewhat confused.
So if Goldman Sachs, J.P. Morgan, the rest of the G-SIFIs all announced that we’re not taking deposits, the Federal Reserve’s response is that they’re going to suspend the capital and leverage rules. And the next day is going to trade how?
It didn’t sound like a good day to me. I’m pretty sure I’m going to be in the office early and I’m pretty sure I’m not going home.
And by the way, that goes more broadly. It’s not even clear you’ll be able to process trades under those rules. Right now there is an ongoing argument about the way in which clearing balances on due to’s, due from’s on an ongoing trade will count against the leverage ratio.
If we decide that clearing is going to count against the SLR, you’re not only going to not take deposits, you’re not going to take trade orders. And that’s going to generate some really perverse behavior.
And so I think that the brittleness of the system under these rules is something that’s really going to have to be looked at.
BAER: So are there other places for that cash to go in your first example? Do they go to money market funds?
STRONGIN: They go to money market funds, and then they will go to that Fed repo facility because they can’t come back.
BAER: If it’s still there, yes.
STRONGIN: If it’s still there. And so that works for a while. And that’s why I said, the system basically has been setup so the Fed becomes —
BAER: So the Fed becomes the borrower of last resort in that case.
STRONGIN: Yes. And the broker-dealer of last resort and the repo market of last resort.
On the clearing of trades, I think that one is more troublesome. Essentially, you’re going to have to stand in line to do your trade. Basically, you’re going to go from a modern electronic market to a swap meeting. You’re going to have to find someone who wants to trade and you’re going to trade security-for-security on a matched basis, and it’s going to slow down trading.
Now I can hear some people saying, well slowing down trading in a crisis is exactly what you want to do. I think that creates a wall of selling that’s of undetermined size and kind of worrisome.
BAER: Sort of lining up around the block to withdraw money is good in a crisis.
STRONGIN: Yes. I’m reminded of that scene from It’s a Wonderful Life where the line just got longer and then nobody knew what it was. We could have that in trading.
And so I think those rules need reexamination. I don’t think it’s hard to fix. Largely, the safety trades in the system, highly secured securities lending, the repo on very good collateral, deposits to the Fed need to be excluded from the ratios.
One of the things running through the rulemaking is the notion that no hedge, no collateral is perfect. And so therefore, no matter what it is, we’re going to charge a nickel. And that charge under stress becomes a problem. So I think you have to go back and decide in a couple of places where we’re really comfortable, that we’re not going to charge a nickel, so that the system does have a place to hide in times of crisis.
MR. WATER: Jessie Water (phonetic), SEC. It’s just a brief comment in support of your criticism of liquidity, regulation in general.
My understanding is, having worked at the FDIC, Fannie Mae and Freddie Mac and a couple of other things, that if you look through this history, you will find precious few bankruptcies or government-run resolutions where money was returned to equity holders. Which means that the worry of a liquidity failure where the entity is essentially solvent but just runs out of funding is nearly non-existent, and that’s it.
STRONGIN: You have to be a little careful because bankruptcies cost money, so you’d take a sudden drop in asset value and there’s a gap there. But the point is generally true.
And certainly, no one’s ever gotten in trouble who is well capitalized. This is one of the reasons in the core reforms I have the balance sheet rules and really want to give SCAP a lot of credit.
SCAP really did, in many ways, stop the banking crisis during the crisis by making sure everybody understood what the balance sheets were. That’s an incredibly powerful tool that was not there before the crisis.
But the Fed is in a position, at any time it wants to, effectively, to recertify the balance sheets. That in itself is a deterrent that is amazing in its power, that has not in any way been incorporated into our understanding of the rest of the rules.
MR. PAUL: So in which quadrant would you put the Volcker Rule?
STRONGIN: I’m pretty sure it didn’t buy you any safety. It’s not that clear what its costs have been. Again, the Volcker Rule is really complicated and its enforcement is not yet clear and it’s another backward-looking process.
I’ll give you a classic example of why these rules are hard to assess today. One of the most malfunctioning markets that probably matters most for economic activity is the midcap corporate debt market. Now, because of the balance-sheet constraints and because of the capital constraints and because of how long I would have to hold those positions, that’s a market in which very few people today are willing to make markets.
A lot of large broker-dealers simply exited. And even the ones that remained have substantially curtailed that part of their activities. If they were still willing to do it, Volcker would be a real problem.
A similar example is that if you were trying to run an oil hedge for a country with heavy oil or a large corporation, a Canadian company with heavy oil, that requires similar warehousing. If it wasn’t for the other three rules stopping you, Volcker would be a real problem.
BAER: Which is another way of saying, if you look at the other rules as we like to do now — or at least some of us do — then you could argue, even if you believed in everything that the Volcker Rule was aimed to do, that it’s entirely superfluous.
STRONGIN: I think that could easily be argued. You’d have to get rid of some of the other rules to experiment.
The fun part of Volcker does have real economic consequences and has zero safety, so I think that’s even easier to argue against. Why do the funds structures pose risk? If you do a securitization, you’re required to own a certain percentage. But if it’s structured as a fund, you’re not allowed to own those same securities because the fund falls under Volcker. So literally, if I incorporate it one way, I’m required to hold a percentage of risk. If I incorporate it another way, I’m required not to hold a percentage or my ownership is capped. That set of rules is difficult to understand and to justify on simple logic.
I think there are a few activities where Volcker has been directly binding. I think if we cleaned up the balance sheet rules, Volcker would then begin to matter a lot, potentially for those midcap markets and commodity markets, in which case I would then put it in the bad quadrant. But we’re going to have to do some reform to get it into the bad quadrant.
Disclaimer: The views expressed in this post are those of the authors and do not necessarily reflect the position of The Clearing House or its membership.