 Okay. Thank you very much, Bob. Thanks for inviting me. And I should say that it's impressive and really somewhat astounding that the JMCB has survived and flourished for the younger people here. I will point out that financial intermediation was kind of in the ghetto for many years, not to mention financial history. And I wish I had saved many referee reports that told me for 30 years that what I was doing was irrelevant and why didn't I work on something else. And so I did, so I could get tenure. But the JMCB has really held down the fort. And I think lasting 50 years is very impressive. So I'm going to talk about bank regulation. The goal of bank regulation is straightforward. It's to prevent or at least reduce the likelihood of a financial crisis. Of course, there's ancillary objectives, which I don't think are that important. And a financial crisis is essentially a bank run. It's an event where agents in the economy no longer believe that private money, bank debt, is worth par. And so they all want their money. So I'm going to say more about this because I'm going to try to give you, I'm going to try to put some flesh on this. But I can tell that even 10 years after the crisis, the way many people speak about the crisis, they have this sort of nebulous idea that a crisis is just some bad thing that happened. And that's not very helpful for policy reasons. So these crises have occurred in all economies, market economies throughout history. They've occurred in advanced economies, emerging markets. They occur in economies with and without central banks. They occur with and without deposit insurance. And they occur with different forms of bank debt. And now it's true that from 1934 to 2007 was a particularly quiet period in the United States. But it was, however, an exception in the United States. Financial crises are basically very frequent in U.S. history other than that period. And since 1970, in a database that Luke has been collecting or started, there have been 147 systemic events around the world. Now, one of the problems with these modern crises is that they don't look like the old crises. Old crises would start with a bank run. They look like that. And it was very clear when they started and what was happening, but not clear what to do about it. In 2007, 2008, part of the problem is that you didn't see that. You didn't see the bank run. You didn't know there was a bank run. You didn't know what kind of debt was involved. And so you didn't see prices of all structured products fall unless you were on a trading floor. And if you were on a trading floor, most people didn't even know what they were seeing. So the result of that has been that most crises are viewed as idiosyncratic. And the last crisis has been dominated by a narrative which is developed given that you didn't actually see what happened. So if you didn't see what happened and you had to come up with an explanation, you would say that it was greedy immoral bankers who concocted toxic assets they sold to unsuspecting investors with fraudulent ratings, which is the dominant narrative. Even in academia, I should say. So what I'm going to do is try to spell out the mechanism that causes a crisis. Every single crisis has the same mechanism. And it's a feature of a market economy. Market economies have certain features, property rights, resources are allocated through the price system. And they also have the feature that they need short-term bank debt. The short-term bank debt is the output of a bank. So General Motors makes cars, McKinsey makes advice, banks make debt. That's their product. And this debt is subject to runs. And the question is, why is it the case that there are runs? What is it that's inherent about this debt that's a run? And what I'm going to try to convince you is that this debt is produced to be information-sensitive. What does that mean? It means that the debt is going to be created in a way that it's not profitable for any agent to produce private information about what's backing the debt. And everybody knows this. So in other words, the price system doesn't work. We don't want the price system to work. The whole point of this is the price system doesn't work. And if the price system doesn't work, then reallocations are going to happen through quantities. And in the extreme, the quantities are going to be, I don't want any more of this bank debt, and it goes to zero. But before that, the adjustment could be on other margins. We could reduce maturity. We could raise haircuts. But the margins of adjustment are non-price margins. And the reason they're non-price margins is because the debt is constructed to be information-sensitive. So let me try to explain why that is. Let's look at some debt that you're, no doubt, unfamiliar with. It looks like this. This is a private bank note from Bull's Head Bank. It's a $1 bill. These kind of private money bank systems were present in about 70 countries. And this is a liability of Bull's Head Bank. This is from the pre-Civil War U.S. And this is a market that financial economists should love, right? Because it's in an efficient market. So this is the Planters Bank of Tennessee note discount in Philadelphia. So if you arrive in Philadelphia with this Planters Bank note and you go to buy lunch, the shopkeeper's going to look in a little newspaper and say, oh no, that's not a dollar in Philadelphia. That's only 85 cents. So the discount there on the left can vary from about 25% down to roughly 5% or below 5%. So this is an example of a market, an information-sensitive market. And this bank note is a perpetual, non-interest-bearing obligation of a bank with an embedded put option that allows you to go back to the bank and ask for a part at any time. And if we calculate how long it would take you to get back in 1847, for example, we can price that option and we can back out the implied volatility from that option. And that implied volatility correlates with all the characteristics you think it should correlate with. Does the state have insurance? Does the state regulate the banks more toughly and so on? So from the point of view of financial economics, this is an efficient market. Now remember, efficient in financial economics, that says nothing to do with economic efficiency. It's a kind of marketing label, right? It means, I remember when I was first starting in the finance department, I asked one of my colleagues, James Dow, I said, what is this stuff with efficient markets? What is efficient markets? I said, I don't get it. He said, it's conditional expectation. Anyway, so this is an efficient market. Now the problem with this efficient market is that if you have this kind of money, everybody's a trader. So when you go to trade, you're obliged to make use of this money, even if you're in no way qualified to produce information. And consequently, laborers and mechanics of all descriptions suffer all sorts of losses. So this market's efficient. The price system works, but we don't want it to work. And we don't want it to work because when you go to buy lunch, you don't want to have all these problems. So what I want to do is give you the intuition, broadly speaking, from some work with Tree v. Dang and Bank Holmstrom. So it's something to convey everything very well, but here's the idea. So here's the contractual payoff on bank debt is the dotted line. Okay. And on the left axis, the y-axis, you can think of the payoff at maturity, or later in a minute, we're going to think of this as the price before maturity. And to the left of the kink is when you're bankrupt. Now, the idea of banks is that they want to convince you that their backing for their debt is sufficiently far to the right of the kink that you never have to produce information about how far we are away from the kink. So all bank debt, as we'll see in a minute, is going to be backed by debt. And that debt is going to have to convince people that they shouldn't bother producing information, that it's not profitable to produce information. Okay. So let me give you an example. Suppose you thought the distribution of collateral values at the end, at maturity, was going to look like this, the normal distribution. And if we, if you thought that, and we say, okay, let's look at the debt at some date prior to maturity, then it would look like that. So here's the value of the debt on the left-hand side before maturity. And the issue here is, what would you, what benefit would you get if you paid the higher people, buy data and produce information? What's the benefit of that? And the answer is it's the integral of that little triangle. Right? Now, if that little triangle is small, and we're going to make it really small in a minute, then it just doesn't pay you to produce information. Now, of course, you have to, hey, you have to be convinced that this is the correct distribution. And for that, we may need regulation, for example. We require that you back it with certain kind of securities. But so just to contrast this, because I'm going to do this more dramatically in a minute, here's, here's, here's what it looks like if there's a straight, a slight shift in the macroeconomy. So the macroeconomy worsens means this distribution shifts slightly to the left. And if it shifts slightly to the left, that triangle is now the red triangle. It's bigger. And the problem with that is that if you think now somebody is producing information, then you are going to be the sucker and you worry about that. And that's when we can get this, this quick adjustment of the quantity says zero. Okay. Now contrast that with equity, equity, as long as the firm is solvent, we're to the right of the kink, equity is always information sensitive. Right? It's always information sensitive. The value of equity depends on the value of equity on the left on the y axis depends on where we are on the x axis. Right? So equity, equity is the sort of canonical security in financial economics. But the point is that when we cut the cash flows, we create, we cut the information, and that's the point of debt. And that means that there's two systems, two systems. Let me go over this kind of slowly. If you look down a column, especially on the left, you see a consistent set of attributes. In stock markets is about risk sharing. There's price discovery. It trades in a centralized location. There's many analysts. The prices are transparent in this market efficient sense. Information sensitive means you can produce information and trade on it. The volume is very volatile. But you could always go trade. You could always go buy or sell your stock. There may be a bid ask spread. So there's a kind of liquidity. It's a risky liquidity. But that's that system is the polar opposite of the system that's in the right column. The right column is a system where the whole purpose of this is liquidity provision in the sense that when I go to use this short term debt, I can be fairly certain that it's retained its value overnight repo, for example. It's a system that's opaque. We don't want people poking around into this stuff. Right? We want them to believe it's credible. It's information insensitive. It doesn't pay anyone to produce information because that little triangle is too small. Trading is bilateral. And there's a pretty much of a constant volume. Now, on the right, the volume and activity in these markets makes the stock market look like a pimple on an elephant. But on the right, academics and regulators don't really know very much about these markets. We don't collect data on repo, for example. We don't collect data on bills of exchange. So this is sort of the uncharted territory on the right. And it's important to emphasize that the intuition from the left shouldn't apply to the right. The intuition on the left shouldn't apply to the right. So it's true, stock markets are transparent. There's no result in economics which says, oh, then everything on the right should be transparent. There's no result that says that. In fact, this whole system on the right is designed exactly to not be transparent, is designed exactly so that the price system doesn't work. So it's a bit heretical to say we don't want the price system to work. Of course, it shouldn't be too much of a surprise. Inside firms, price systems don't work either. So that's another sort of undiscovered area. So these two systems are at sort of a contrast that is, I think, confused a bit. When we look at policy advice, a lot of it flows from the left system and is claimed to be something about the right. Now, so let's go back to the earlier figures. Here's the loss distribution for debt. So this is a bond, an asset-backed security, if you want. The x-axis has the loss, x is the loss, f is the face value, and the y-axis is the frequency of the loss. So what I'm going to do is I'm going to reflect this around, like that, and I'm going to put it together with the figure that shows the contract for debt. There it is. So you can see there that the little triangle now is tiny. It's tiny. And so that's one of the results in this paper with Holmstrom and Dang is that the maximal information insensitivity is debt on debt, debt backed by debt. So in that paper, there's a compound contracting problem in which an agent who's going to trade with somebody has to decide what the collateral is going to be. And the answer is the collateral should be debt and his contract should be debt. So it's debt on debt, which has the feature that it makes that little triangle very small. So if you think about repo, it's backed by a bond almost exclusively. Demand deposits are backed by loans, mostly to small enterprises, mortgages, things that are pretty opaque. And so that's the way that the system is constructed to maximize the feature that nobody should produce any information about it. Now you can immediately see the problem here. The problem is if we don't have the price system working and we have the quantity system replacing it, then you can see how we can have a crisis. The crisis is going to be a manifestation of the fact that we suddenly want to produce information or we suddenly think somebody is producing information and suddenly quantities adjust, nobody wants the short term debt, and nobody knows how to price the collateral. The collateral is things like triple A credit cards. Those are things that nobody could do a loan by loan Monte Carlo to actually value mortgage and asset backed securities before the crisis. No one. So and that's sort of the way it's constructed. The system is, you know, we need for other reasons, we're going to produce this debt that's safe mortgage back and asset backed securities. So any securitization, 85% of the securitization is triple A, legitimately so. But if you start to question the backing of your short term debt and it turns out to be securitized bonds, nobody knows what the price of those bonds is. And so what happened in the market was the prices of those bonds plummeted. They plummeted and there was no bottom because nobody knew exactly how to even go about producing information about those bonds. And then of course the accountants toddled along and told you to mark to market where for them the idea that there was no market was kind of inconceivable because they live on that stock market column. Okay. So this immediately suggests that there's banks are surrounded by secrecy and they're surrounded by secrecy by design, but they're also surrounded by secrecy in central banks. We don't want discount window borrowing to be known. We don't want the identity of banks borrowing at emergency lending facilities to be known. We don't want weak banks to be revealed. So we ban short sales in the crisis. This was, you know, you could see many financial economists tearing their hair out. You're banning short sales. We don't want weak banks revealed. So it also immediately says well there's two approaches to regulation. One thing is you can require very high quality collateral which says don't worry about the short-term debt. And the other is we ensure the short-term debt which says don't worry about the collateral. The government is the collateral. And both of these both of these approaches have been seen throughout history. Just some examples. Peele's Act in 1844 required that Bank of England notes be backed with gold. U.S. Free Bank notes had to be backed by state bonds. U.S. National Bank notes had to be backed by treasuries. And the BIS has adopted this logic which I should point out never worked. It never worked, right? So Peele's Act was quickly repealed. U.S. Free Bank notes didn't, it wasn't a solution because the state bonds were risky. The National Bank Act which is exactly like liquidity coverage ratio said for every national bank note you issued, you're a bank, you had to have a dollar of treasuries. Well there weren't enough treasuries. They couldn't make enough bank notes. And so the shadow banking system arose which was demand deposits which grew seven times faster in the U.S. than in any other country. So one of the one of the solutions, here's the here's the insurance, one of the solutions has been to create charter value, right? So charter value is basically comes mostly from entry restrictions which creates an incentive for banks to follow the rules, right? So we're basically giving them monopoly rents. And if you have monopoly rents, you don't want to jeopardize your charter, right? That's the logic of this. And there's an empirical literature about this as well which I'll come to. But it's not something that you know policy makers can sort of say, you know, out to the Congress or the parliament that we need to create monopoly rents for banks. But the reality is that banks are private firms and regulators can only do one thing. They can only decide where the banking system is. That's all they can decide. They can have all sorts of regulations and you can beat up banks with a big stick but banks can move. And I'll show you in a minute that that's what's happening. So this charter value is not trivial. Now these sorts of things have always been confusing and demand deposits were particularly confusing. This is Bray Hammond, the last bank book that won the Pulitzer Prize. Probably the only one that will ever win the Pulitzer Prize. And he points out the importance of deposits was not realized by most American economists until after 1900. So you think, what was wrong with these people? I don't understand deposits were debt. It's not so different now, right? I mean now is, you know, all these other forms of short-term debt that banks produce, banks meaning firms that produce a short-term debt, you know, are not understood in the same way that there was this intellectual confusion. So let's talk about this system that resulted in the financial crisis. It's this system. So this is a figure that shows the following. This is total privately produced short-term debt, the different components divided by total privately produced short-term debt, right? So if you look at the largest component there, it goes at 80% in 1952. So that's 80% of all privately produced short-term debt in the United States was demand deposits, okay? 80% of all privately produced short-term debt was demand deposits. And you can see it goes straight down. You can see the red arrow there. It goes down starting in the late 70s, it goes down, okay? And the component that goes up, the next component here, this component, is money market instruments. So this is money market mutual funds, bilateral repo, commercial paper, asset-backed commercial paper. And this is private label and agency AAA securitization, okay? So these guys are making loans. We're taking the loans and we're turning it into this so it can be used as collateral, some of which is used to back this, okay? Now there's a couple of points about this. One thing is you can see that this system, this system, the red bracket, which is shadow banking, has been in gestation for decades. It's not something that just happened in 2004 when bankers got greedy and they invented this. It's the whole system has been transforming, it's been morphing, because lots of other things in the world change. So this system is a fundamental transformation, and there's other aspects of this too, right? Like the gross flows in the capital account. It's all sorts of things which are associated with these fundamental changes. So it's not, in a way, it's not so surprising that, you know, the world changes. It's just that in banking, people didn't think it was ever going to change, right? I mean, we now have, I don't know, iPhones. So why wouldn't banks, you know, the world of banking change where most of the depositors turn out to be large institutions, for example. So this is just a snapshot of the evolution. Now, one of the things that keep in mind is that if your perspective is short, for example, if you don't look past prior to the Great Depression, and you can't understand the Great Depression, then you think there's never going to be a crisis, this requires that you track this for 30 or 40 years and see this evolution. So this is flow of funds data, right? Federal Reserve data, you can download it. So why didn't the Fed see this? Answer? They had no concept, no concept of safe debt, no theory of what banks do, right? They studied inflation for 30 years, right? So there was never a reason to compute this figure. So as you know, in the face of this, you know, there have been a lot of changes in bank regulation, a lot of changes to put it mildly in bank regulation. So, you know, it might be worthwhile checking to see how we're doing in terms of charter value. So charter value in the literature has been measured by Tobin's Q, market to book ratio, and a lot of people have criticized this for all the usual reasons. Nevertheless, here's a picture of Q ratios. So this is three, three crises. So the left panel shows all U.S. banks, which is that kind of P green line. Blue is all Euro banks. These banks are total asset-weighted, and in the case of Euro, also GDP-weighted. And then there's this Rogoff-Reinhardt five big crises thing. The left is all U.S. banks. The right are the big banks, the investment banks, and the big banks that have big investment divisions. Now, for Europe and the United States, you can see before the crisis, which is lined up at zero, charter value as measured by Tobin's Q was high. It was above one, which is what you expect from banks. You expect banks to be enticed to follow the rules by having a high charter value which appears in a Tobin's Q above one. One would be the equilibrium. If you're above one, people should invest, Q should go down. If you're below one, people should disinvest and Q should go up. In banking, it should be above one. And you can see what's happened over here, or in either panel, once after the crisis, Q is below one forever, except for the orange line. So the fact that Q is below one for 10 years is pretty worrisome. It basically says banks don't have a business model. In the United States, banks' business model is get 240 basis points on your excess reserves. That's your business model. The market says they have no business model. And if they have no business model, then people aren't going to invest in banking or disinvest in this case. And so you should see the migration of activity out of the banking system. And there's a lot of evidence on this. I'm just going to give you a few pieces. One is that today, in the United States, 50% of mortgages are originated outside the banking system. Before the crisis, it was 7%. So mortgage origination has gone outside the banking system. Here's a list of the top lenders to medium and small-term enterprises. On the left is 2011. If I had this chart for, say, 2007, there would be no non-banks on this chart. Green is a non-bank. If you look in 2016, it's dominated by non-banks. Lending is happening outside the commercial banking system. And so is this a big cause for concern? No, it's not a cause for concern. As long as they're not financing themselves with short-term debt, are they financing themselves with short-term debt? Who knows? We don't collect any data on that. Anecdotally, only a small number of them are. But it tells you that the banking system is moving. And the banking system is moving in response to the fact that they have no business model. So it's not a good situation. And it's not consistent with the view that regulators can only determine where the banking system is. If you just beat up banks all the time, they're going to move. It's free enterprise. This is privately owned capital. And so it's kind of disheartening. So let me conclude. History suggests that financial crises are inevitable because the system is constantly transforming. Are they inevitable? Well, they're only inevitable insofar as there's an intellectual failure. If there wasn't an intellectual failure, we would see the system morphing and we would be proactive. But that would require a perspective that's longer than most people's professional career. I mean, many economists think that economic history is the stuff that happened before they went to graduate school. That's sort of their perspective. So if you have a very short perspective, you're not going to think, well, gee, it's been 40 years since we had a crisis. Maybe things are changing in a way that we don't know. The answer is probably that that was occurring. And finally, these two systems that I outlined is kind of a fundamental typology of the financial system has huge implications for how to think about regulation. So policies derived from stock markets don't work. Let me give you a couple of examples. One, the money market reform in the United States just destroyed money market funds. Oh, they're all government funds now. You can't have money market funds anymore. They all died. And why did they die? They lost their moniness. They were forced to be transparent. They can no longer be money. No one wants them. The ECB has a program that requires securitizations as they purchase and supply all sorts of information, very fine information on a monthly basis. And what has happened there? Well, what's happened there is that the junior tranches of securitization, which were information sensitive, information sensitive anyway, became more liquid. The more senior tranches, which were information insensitive, have now become information sensitive. So he succeeded in destroying that stuff as collateral. So that's all policies that come from this view that somehow, because stock markets are efficient in this information sense, that everything in the world should be efficient. And that's just not the case.