Lawrence Christiano

Alfred W. Chase Chair in Business Institutions, Northwestern University

Paper: "Leverage Restrictions in a Business Cycle Model" (with Daisuke Ikeda)

Andolfatto

Would you like to give us the rundown on your paper?

The paper's main takeaway,
according to Christiano:

  • Regulating the banks is probably a good idea, and a completely free market when it comes to banking is probably a bad idea.

Christiano

Well, one of the many things that we learned as a result of this recent crisis is that the banks can be a problem in the economy. Before 2008, 2007, most people thought that all banks are rock solid and that if we're looking for vulnerabilities in the economy, the banks are not part of the picture. But that changed with 2008. And to think about this more, you have to think about what banks do. And so what they do is they borrow short term and they lend long term, and there's something inherently dangerous about that because if short-term interest rates go up a lot, then the banks could be caught, actually could go into bankruptcy. Because the banking system is a lot like the air that we breathe, in the economy if the banks go down, we're not going to be breathing any air. The whole economy will go down.

Virtually every relationship in the economy is mediated somehow by a bank. If you're working for a firm, your pay is going to be transferred to you from that firm by a bank. If you buy something, a bank's involved. A bank is involved everywhere in every economic relationship. And so if we see that the banks are getting risky and vulnerable to going out of business, then we really have to be concerned. Now, as I said, up until recently no one was worried about stuff like this. But people have become increasingly concerned about this possible vulnerability.

So that brings me to this issue about leverage. Leverage has to do with how much banks borrow relative to what they have. And obviously the more a bank borrows, just like an ordinary person, the more they're at risk of being caught short. And so there's been a lot of thinking about how it is that maybe banks borrow more than is good for society as a whole and how we should perhaps rein them in. And this reining in is called leverage restriction. So the question is, from a policy point of view, what kind of leverage restriction should we place on banks? And the answer is complicated, because it has to do with what's healthy for the economy as a whole, which means you have to understand exactly how the banks kind of work in the economy as a whole.

For example, in a recession, there may not be enough credit flowing to the economy, and we may want to take that into account when we restrict how much borrowing banks do. We may be willing to tolerate a little bit of riskiness in the banks because there's a benefit to credit flowing in a recession perhaps. And so to think about what the right amount of restriction on banks should be, we have to balance off the riskiness in the banks on the one hand with the desire to provide a lot of credit to the economy. The fact that the answer to the question—how much leverage restrictions we should have on banks&mdashinvolves thinking about the economy as a whole, makes it a very complicated problem and requires that we adopt a global perspective. We have to think about the economy as a whole, and that's what gets us to models of the economy as a whole. And the paper that I'm talking about today is such a model.

Andolfatto

Why in your view can the free market not be relied on to generate the correct amount of leverage? I mean, the title suggests that there's a propensity for the private market to overleverage, overexpose?

Christiano

If anybody is to think about leverage restrictions, they had better do it in a framework where such things are desirable. If I want to think about how many umbrellas we ought to build, it should be in a world where we take into account that there can be rain. And similarly, in the case of the thinking about leverage restrictions, we really need to do it in an environment where we put our fingers on exactly what it is that would make the banking system issue too much leverage if there were no regulation at all.

And so in economics we have many examples of how it is that markets might break down. The obvious example is pollution. If a firm generates pollution when it's producing and no one forces a cost on the firm for that, then we can expect too much of that activity to be happening. So in this paper, for example, we take a very particular position on what it is about the banks that leads them to issue, in the absence of regulation, too much debt, and that's what people call, in fancy words, financial friction. The idea is that the job of banks is to go out and find good lending opportunities, but that involves efforts that are not observable to people, and that's what gives rise to the market not being necessarily able to deliver the best outcomes.

We have a number of examples of this. You know, for example, health care is much in the news these days. As an example, for various interesting reasons that have colorful names like the death spiral, we have various discussions about how wide is it the market might not generate the right amount of health care. And similar examples are used to think about why banks in an unregulated market might issue too much debt.

We've been trying to think what kind of a framework would be useful for thinking about this. And, in particular, we're interested in answering the question: What kind of framework has the property that you would have the banks issuing too much leverage? And then we're asking more related questions like: Does this framework that we have in some loose sense look like the data? But in terms of really digging in yet, we're very hopeful that we can use this model to do that.

Andolfatto

When you say look at the data and when you use the term banks, do you mean literally banks? Or do you think more broadly to include those activities that occurred in the so-called shadow banking system?

Christiano

Not all of them, but a lot of the problems that we talk about are not so big in the context of the commercial banking system. So we really do have in mind the shadow banking system when we're talking about that stuff. And that's exactly the place that was thought to be the source of the problem in 2008. In actual fact, if you look at the commercial banks in 2008, they look beautiful. It was actually the rest of the banking system that was performing very badly and doing damage to the economy. But it was hard to see that because, since they're not heavily regulated, there's not much data on them, and so you didn't see them very much.

Andolfatto

Well, the commercial banks had the federal deposit insurance as well to protect the downside.

Christiano

And the deposit insurance to a large extent gets rid of the problems that we're talking about, although, not entirely. Because there are stockholders who look like ordinary lenders in the real world, namely people who hold preferred stock. People who hold preferred stock, from the point of view of our model, look like ordinary lenders.

Andolfatto

Does your model have anything to say then in terms of the most recent crisis? If we had had more severe leverage restrictions in place, would much of the worst impact of the most recent recession have been avoided or mitigated in some manner?

Christiano

One of the problems that the model highlights is that if banks have too little net worth, if the assets that belong to the bank get too small, then that interferes with the ability of the bank to do its job, which is to send funds from savers to investors. And in the case of the 2008 crisis, what it looks like is that what made 2008 actually different from, say, 2001 is precisely that the wealth of the banks themselves went down with the collapse in housing prices. This is different from 2001, for example, where we had the dot-com bust, and banks were not heavily invested in that so they did not suffer a big reduction in their own resources. But 2008 is very different because they were heavily invested in houses. That is, they had bought a lot of mortgage-backed securities, those dropped in value, and that interfered with their ability to conduct their business: to move the funds from savers to investors.

Andolfatto

What would have been the costs, if any, of having these leverage restrictions in place? Can you quantify it in terms of would it have reduced growth over the period?

Christiano

In our analysis, the leverage restrictions actually make the economy stronger. It's like pollution. So from a point of view of the individual bank, they don't internalize all the costs of issuing all this debt.

And actually for an individual bank there are not big costs. But if they all do that simultaneously, it does damage to the economy. That's why the market doesn't work because the banks are not getting all the right signals themselves, just like the polluter is not. There's no market mechanism for a polluter to get a signal about what costs they're imposing on other people. Similarly, in an unregulated market, banks are not getting all the signals about the consequences of all the borrowing that they're doing. There's no market mechanism for warning them that they're issuing too much debt or something like that.

Andolfatto

Does your model speak to the role that these leverage restrictions might have had, say, in terms of the evolution of real estate prices or the real estate price bubble?

Christiano

Well, this is stretching it. But in the model what happens is that if the economy does go into a collapse, it can be self-reinforcing. As the economy goes down, the value of assets go down, that makes the banks worse, that makes the economy go down more, and so on. So, I mean, this is an example of why banks are so important because they're right at the heart of the economy. If something goes wrong with those things, bad stuff can happen. And the model does capture that idea.

Andolfatto

Apart from policies that determine leverage restrictions, does your model have anything to say about what the fiscal authority or the monetary authority might do in an emergency situation?

Christiano

Well, so far the way the model works is, in some respects, it falls into a general pattern of models that says that recessions can be excessive. The economy in a recession may go down too much. The 2008/2009 episode in the United States, I think there's general agreement that things went down too much. And so what the model says is that anything you can do to stop that from happening is a good idea. So for example, in terms of monetary policy, what it would say is that under circumstances where there's a recession developing, it's a good idea to cut interest rates. And that's roughly what the policy actually is. Right now that policy, of course, has been short circuited by the lower bound on the interest rate. But my guess is that all that this will do is add to the reasons why monetary authorities ought to be stabilizing the economy.