The Federal Reserve and the Financial Crisis
Page 5
This situation could not go on forever. Figure 14 shows the debt-service ratio. As house prices went up and up and up, the share of borrowers’ incomes being spent on their monthly mortgage payments went up. As you can see, eventually mortgage payments became quite a large share of personal disposable income, finally reaching the point that the cost of homeownership was high enough that it began finally to dampen the demand for new houses. The debt-service ratio collapsed after that, basically because interest rates came down. But the main point here is that high payments on mortgages finally meant that there were no longer new home purchasers, and so the bubble burst and house prices fell. Figure 15 shows home prices. You can see the sharp increase from the late 1990s up until about 2006. But from 2006 until today, house prices have fallen more than 30 percent. So there has been a very sharp decline in home prices across the country.
Figure 13a. Nonprime Mortgage Originations (as a Share of Total Originations), 1995-2009
Source: Federal Reserve staff estimates, based on data from Inside Mortgage Finance
Figure 13b. Percentage of Nonprime Loans with Low or No Documentation, 2003-2007
Source: Derived from data in Christopher Mayer, Karen Pence, and Shane M. Sherlund,“The Rise in Mortgage Defaults,”Journal of Economic Perspectives 23 (winter 2009): 27–50, table 1 and table 2, panel C.
Figure 14. Mortgage Debt-Service Ratio, 1995-2011
Source: Federal Reserve Board
Figure 15. Prices of Existing Single-Family Houses, 1980–2010
Note: Includes purchase transactions only.
Source: CoreLogic
One comment about Figure 15: if you look at this graph you might say to yourself, “Oh my gosh, we have a long way to go,” because house prices today are still significantly above where they were fifteen years ago. But remember, these prices are in dollar or nominal terms; there is no adjustment for inflation. So even if there was just 2 percent inflation per year, over a period of fifteen years that would raise prices by 30 or 40 percent. So if you adjust for inflation, you find that house prices now are coming much closer to where they were before the beginning of the bubble.
Figure 16. Mortgages with Negative Equity, 2005-2012
Note: Negative equity number likely is understated because of incomplete data on junior liens. Source: Federal Reserve staff calculations, based on data from CoreLogic and LPS Applied Analytics
The house price collapse had some significant consequences. One consequence is that many people who had felt rich because their home values had gone up and they had a lot of equity suddenly found themselves underwater, which means that the amount of money they owed on their mortgages was greater than the value of their homes. This is an upside-down situation where the borrower in fact has negative equity in the home. In figure 16, you can see that starting in 2007, the number of mortgages that were in negative equity grew very sharply. Currently, about twelve or thirteen million mortgages out of a total of about fifty-five million or so in the United States—roughly 20 to 25 percent of all mortgages—are currently underwater.
Figure 17. Mortgage Delinquencies, 2003-2012
Note: Loans ninety days or more past due or in foreclosure.
Source: Federal Reserve Board estimates based on data from the MBA National Delinquency survey
At the same time, given the fact that a lot of people borrowed more than they could afford, the decline in house prices also led to a big increase in mortgage delinquencies, people not paying on time, and ultimately the bank taking over the property—that is called a foreclosure—and then reselling the property to somebody else. Mortgage delinquencies are graphed in figure 17, and you can see that in 2009, there were more than five million mortgages in delinquency, which is almost 10 percent of all mortgages. That is a very, very high rate of delinquencies.
We just looked at the effects of the house price bust on borrowers and homeowners, and those are quite serious. But there is another side to this, which is the effect on lenders. With approximately 10 percent of mortgages in delinquency, banks and other holders of mortgage-related securities suffered sizable losses and that proved to be an important trigger of the crisis. There is an interesting question here. In 1999, 2000, and 2001, we had a big increase in stock prices, including, but not confined to, dot-com or tech bubble prices. Those prices fell very sharply in 2000 and 2001, and a lot of paper wealth was destroyed. In fact, the amount of paper wealth destroyed by the decline in dot-com and other stock prices was not radically different from the amount of wealth destroyed by the bursting of the housing bubble. And yet, the dot-com bust led only to a mild recession. The 2001 recession lasted from March to November 2001; it was only an eight-month recession. Unemployment rose, but not nearly so dramatically as in the 1980s or more recently. And so, here we had a big boom and bust in stock prices, but without causing too much serious or lasting damage to the financial system or the economy. In the recent case, we had a housing boom and bust. If we were looking back at 2001, we would think that would cause a slowdown in the economy, but probably it would not be very serious. That was one of the views we were discussing in the Fed in 2006, as we saw house prices decline. Yet, the decline of house prices had a much bigger impact on the financial system and the economy than the decline of stock prices did. To understand that, it is important to make a distinction between triggers and vulnerabilities. The decline in house prices and the mortgage losses were a trigger. They were a match thrown on kindling. There would not have been a conflagration if there had not been a lot of dry tinder around. In this case, there were vulnerabilities in the economy and in the financial system that the housing bust in some sense set afire. In other words, there were weaknesses in the financial system that transformed what might otherwise have been a modest recession into a much more severe crisis.
What were those vulnerabilities? What was it about the financial system of the United States and of other countries as well that transformed the housing boom and bust into a much more serious crisis? There were vulnerabilities both in the private sector of our financial system and also in the public sector. In the private sector, many borrowers and lenders took on too much debt, too much leverage. And one reason they did that may have been the Great Moderation. With twenty years of relatively calm economic and financial conditions, people became more confident, willing to take on more debt. The problem with taking on too much debt is that if you do not have much margin, if the value of your asset goes down, then pretty soon you will find that you have an asset that is worth less than the amount of money you borrowed.
A second, very important problem was that during this period, financial transactions were becoming more and more complex but the ability of banks and other financial institutions to monitor and measure and manage those risks was not keeping up. That is, their IT systems and the resources they devoted to risk management were insufficient for them to understand fully what risks they were actually taking and how big the risks were. So if in 2006 you asked a bank about the effect if house prices fell 20 percent, it probably would have greatly underestimated the impact on its balance sheet because it did not have the capacity to measure accurately or completely the risks that it was facing.
A third problem is that financial firms in a variety of contexts relied very heavily on short-term funding such as commercial paper, which can have a duration as short as one day and most of it is less than ninety days. So, like the banks in the nineteenth century that were relying on deposits and making loans, on the liability side of their balance sheets, they had a very short-term, liquid form of liability, which was subject to runs in the same way that deposits were subject to runs in the nineteenth century.
A final private-sector vulnerability was the use of exotic financial instruments, complex derivatives, and so on. An example of this was the credit default swaps (CDSs) employed by the AIG Financial Products Corporation. AIG used CDSs essentially to sell insurance to investors on the complex financial instruments that the investors held. So
basically, AIG was promising that if the investor lost any money on collateralized debt obligations or whatever, AIG would make good. As long as the economy and the financial system were doing well, then they were just collecting the premiums on this insurance, essentially, and there was no problem. But once things went bad, their being on one side of all these bets meant that they were exposed to enormous losses, which had, as we will see, very serious consequences. So those are some of the problems that occurred in the private sector.
There were serious problems in the public sector as well. First, the financial regulatory structure was basically the same structure that had been created in the 1930s during the Depression. And in particular, it did not keep up with changes in the structure of the financial system. One aspect of that was that there were many important financial firms that did not really have any serious, comprehensive supervision by any financial regulator. An example was AIG, which was an insurance company. The insurance regulators looked primarily at the insurance products AIG sold. The Office of Thrift Supervision looked primarily at the small banks that AIG owned. But nobody was really looking carefully at this CDS problem that I was just describing. Another category of firms that did not have much oversight was investment banks such as Lehman Brothers and Bear Stearns and Merrill Lynch. There was no statutory oversight of those firms. They had a voluntary agreement with the SEC for oversight, but there really was not comprehensive oversight of those firms. Another group of firms was the government-sponsored enterprises (GSEs) Fannie Mae and Freddie Mac, which did have a regulator but, for reasons I will explain, the regulation was very inadequate. The regulatory structure had lots of holes in it, and there were many firms that proved important during the crisis that did not have good oversight. Even where the law provided for regulation and supervision, it often was not done as well as it should have been.
Although this was true across the whole range of agencies and parts of the government, since I am the Fed chairman, let me talk about the Fed. The Fed made mistakes in supervision and regulation. I would point out two. One would be in its supervision of banks and bank holding companies, it did not press hard enough on this issue of measuring risks. I mentioned earlier that a lot of banks simply did not have the capacity to thoroughly understand the risks they were taking. The supervisor should have pressed them harder to develop that capacity and, if they did not develop that capacity, should have restricted their ability to take risky positions. The Fed and other bank supervisors did not press hard enough on this, and that turned out to be a serious problem. A second area where the Fed performed poorly was in consumer protection. The Fed had authority to provide some protections to mortgage borrowers that, if used effectively, would have reduced at least some of the bad lending that occurred during the latter part of the housing bubble. But for a variety of reasons that was not done to the extent it should have been. In 2007, when I became chairman, we did undertake some of these protections but it was too late to avoid the crisis. So, where there were authorities and powers, they were not always effectively used, and that led to some weaknesses.
A final, and perhaps more subtle, point is that the way our regulatory system is set up, individual agencies, such as the Fed or the Office of the Comptroller of the Currency or the Office of Thrift Supervision, typically had as their responsibility just a specific set of firms. So the Office of Thrift Supervision was responsible only for thrifts and similar institutions. Unfortunately, the problems that arose during the crisis were much broader based than that. They transcended any single firm or small group of firms; they encompassed the whole system. And so essentially what was missing here was enough attention being paid to things that could affect the system as a whole, as opposed to just individual firms. Nobody was in charge of looking to see whether there were problems related to the overall financial system or the relationships among different markets and different firms that could create stress or even a crisis. So those were some of the vulnerabilities in the public sector.
Let me conclude by talking about a controversial topic, the role of monetary policy. Many people have argued that another contributor to the housing bubble was the fact that the Fed kept interest rates low in the early part of the 2000s following the recession of 2001. When the economy got very weak and there was very slow job growth in 2001 and subsequently, and when inflation fell very low, the Fed cut interest rates. In 2003, the federal funds rate got down to 1 percent. There are people who argue that this was one of the reasons that house prices went up as much as they did. And it is true that one of the purposes of low interest rates that the monetary policy achieves is to increase the demand for housing and thereby to strengthen the economy. As I say, this is very controversial. But it is also very important, not only because we want to understand the crisis but also because we want to think about what we should take into account when we formulate monetary policy in the future. To what extent should we be thinking about things like housing bubbles when we make monetary policy?
We have looked at this in great detail inside the Fed, and there has been a lot of research outside the Fed, and (there is no consensus on this and you will probably hear different points of view) the evidence that I have seen suggests that monetary policy did not play an important role in raising house prices during the upswing. Let me talk a little bit about some of the evidence on this question.
One piece of evidence is the international comparison. People do not appreciate that the boom and bust in the United States was not unique. Many countries around the world had booms and busts in house prices, and those booms and busts were not very closely related to the monetary policies of those particular countries. For example, the United Kingdom had a house price boom that was as big or bigger than that in the United States. But monetary policy was much tighter in the United Kingdom than it was in the United States. So there is a puzzle for the monetary theory of the house boom. Another example: Germany and Spain both share the euro, so they have the same central bank, the European Central Bank, and the same monetary policy. Germany’s house prices remained absolutely flat throughout the entire crisis, whereas Spain had an enormous house price increase, considerably larger than that in the United States. So the cross-national evidence raises at least some doubts that monetary policy played a large role in the housing bubble.
The second issue is the size of the bubble. It is true that changes in interest rates and mortgage rates should affect house prices and demand for homes. And there is a lot of evidence to look at that over a long period of time. But when you look at how much interest rates changed, including mortgage rates, and how much house prices moved, based on historical relationships, you can explain only a very small part of the increase in house prices. In other words, the increase in house prices was much too large to be explained by the relatively small change in interest rates associated with monetary policy in the early part of the 2000s.
The final piece of evidence I would cite is the timing of the bubble. Robert Shiller, an economist who was well known for his work on bubbles, including the housing bubble, argued that the housing bubble began in 1998, which of course is well before the 2001 recession and before the cut in Federal Reserve interest rates. Moreover, house prices rose very sharply after the tightening began in 2004. So the timing does not line up particularly well. Now, the timing does suggest a couple of other possible explanations. One is that 1998 was right in the middle of the tech boom. And it could be that the same psychological optimism, the same mentality that was feeding stock prices, may have been feeding house prices as well. Another possibility that has been pointed out by a number of economists is that in the late 1990s, there was a very serious financial crisis that hit a number of Asian countries and other emerging-market economies as well. After that crisis was tamed, one response was that many emerging-market countries began to accumulate large amounts of reserves, which meant they had to acquire safe dollar assets. So there was a big increase in the demand for assets, including mortgages. It came from abr
oad as countries decided they needed to acquire more dollar assets to serve as reserves. Interestingly, probably the strongest correlation across countries that you can find to house price increases is capital inflows, the amount of money coming in to buy mortgages and other assets that were perceived to be safe. That timing would also fit with the beginning in 1998 or so.
So, those are some arguments against the view that monetary policy was an important source of the housing bubble. But I emphasize, economists continue to debate this issue, which is a very important one because, going forward, we have to think about the implications of low interest rates for the economy and the financial system. And in particular, currently, just out of caution, the Fed is doing a lot of financial and regulatory oversight to do the best it can to ensure that nothing is getting unbalanced in the financial system.4