It was not just the house price boom and bust but also the mortgage products and practices that went along with the house price movements that were particularly damaging and that were important triggers of the crisis. There were a lot of exotic mortgages, by which I mean nonstandard mortgages (standard mortgages are thirty-year prime fixed-rate mortgages). All different other kinds of mortgages were being offered, often to people with weaker credit. One feature that many of these mortgages shared was that, in order for them to be repaid, house prices had to keep increasing. For example, you might be a borrower who would get an adjustable-rate mortgage (ARM), where the initial interest rate was 1 percent, which meant that you could afford the payment for the first year or two. Now, after two years, the mortgage interest rate might go up to 3 percent, then after four years, 5 percent, and then higher and higher. In order to avoid that, at some point you would have to refinance into a more standard mortgage. As long as house prices were going up, creating equity for homeowners, it was possible to refinance. But once home prices stopped rising—and by 2006 they were already declining quite sharply—rather than having built equity, borrowers found themselves underwater. They could not refinance and found themselves stuck with increasing payments on their mortgages.
Some examples of bad mortgage practices include:
• interest-only (IO) adjustable-rate mortgages (ARMs);
• option ARMs (which permit borrowers to vary the size of monthly payments);
• long amortizations (payment periods greater than thirty years);
• negative amortization ARMs (initial payments do not even cover interest costs); and
• no-documentation loans.
Most of these bad mortgage practices shared the feature that they reduced monthly payments early in the mortgage but allowed mortgage payments to rise over time. Take, for example, an option ARM, which is an adjustable-rate mortgage with the option for borrowers to vary how much they pay. They could pay less than the full amount owed each month and what they did not pay just got rolled back into the mortgage. The other common feature of bad mortgage practices, such as no-doc loans, was that there was very little underwriting, which means very little analysis to make sure that the borrower was creditworthy and was able to make the payments on the mortgage.
Figure 22. The Deterioration of Lending Practices
Figure 22 shows two advertisements from the period that can illustrate some of the issues. I like the one on the right. We removed the name of the company. Let’s look at the features it is offering. A “1 percent low start rate”: the start rate is what you pay the first year; it does not tell you about the next year. “Stated income” means that you tell the company what your income is, and they write it down; that is all the checking they do. “No documentation” is self-explanatory. A “100 percent finance” means that no down payment is required. “Interest-only loans” means that you pay the interest but you do not have to pay any principle back. And “debt consolidation” is an interesting arrangement that allows you to go to the mortgage company and say, “Well, not only do I want to borrow money to buy the house, but also I want to add in all my credit card debt and everything else I owe, and put that into one big mortgage payment, and I’ll pay for that with the 1 percent start rate.” Obviously, these are some very problematic practices.
Mortgage companies, banks, savings and loans, and a variety of other institutions originated these nonprime mortgages, but where did these mortgages wind up? How were they financed? Some of them were kept on the balance sheet of the mortgage originator, but many or even most of these exotic or subprime mortgages were packaged in securities and sold off into the market.
Some mortgage-backed securities were relatively simple. If the mortgages were sold to Fannie and Freddie, they had to meet Fannie and Freddie’s underwriting standards. Fannie and Freddie would combine them into mortgage-backed securities and sell them with a guarantee, as I discussed. Those are relatively simple securities made up of basically just hundreds or thousands of underlying mortgages.
But some of the securities that were being created were very complex and very hard to understand. An example would be a collateralized debt obligation (CDO). This would often be a security that combines mortgages and other types of debt together in one package. And it could be sliced in different ways so that one investor could buy the safest part of the security and another investor could buy the riskiest part. These securities were very complicated and required a lot of analysis.
One reason that many investors were willing to buy these securities was because they had the reassurance that the rating agencies, whose job is to rate the quality of bonds and other securities, gave triple-A ratings to many of these securities—essentially saying that these securities were very safe and one did not have to worry about their credit risk. Many of these securities were sold to investors, including pension funds, insurance companies, foreign banks, and even, in some cases, wealthy individuals. But the financial institutions that created these securities often retained some of them as well. For example, sometimes they would create an accounting fiction, an off-balance-sheet vehicle, which would hold these securities and finance itself using cheap short-term funding such as commercial paper. So, some of the securities went to investors and some of them stayed with the financial institutions themselves.
Figure 23. Subprime Mortgage Securitization
In addition, we had companies, such as AIG, that were selling insurance. They were using various kinds of credit derivatives to say, “Pay us a premium and if the mortgages in your mortgage-backed security go bad, we will make you whole.” That makes the security triple-A rated. Of course, these practices made the underlying securities no better, and basically they created a situation where risks could be spread throughout the system.
Figure 23 is a diagram showing how a subprime mortgage securitization might work. On the left, where the box reads “low-quality mortgages,” you might have a mortgage company or a thrift company making loans. This thrift or mortgage company does not care too much about the quality of the loan because the company is going to sell it anyway. So they sell the mortgages to large financial firms, which in turn take those mortgages, and maybe other securities as well, and combine them into a security that is essentially an amalgamation of all the underlying mortgages and other securities.
Now, the financial firm that created the security might negotiate with the credit-rating agency, asking, “What do we have to do to get a triple-A rating?” There would be negotiations and discussion and, in the end, the security would be rated triple-A. The financial firm could then cut up the security in different ways or sell it as is to investors such as pension funds. But again, financial firms kept many of these securities on their own books or in related investment vehicles. And finally, on the right in the figure, you have credit insurers such as AIG and other mortgage insurance companies that for a fee provided insurance in case the underlying mortgages went bad. So this is the basic structure of the securitization process. I have seen complete flowcharts and they are incredibly complex. This is a very simplified version, but the basic idea is there.
As you recall, a financial crisis or panic occurs when you have any kind of financial institution that has illiquid assets (long-term loans, for example) but liquid short-term liabilities (deposits, for example). In a classic bank panic, if bank depositors lose faith in the quality of the assets held by the bank, they run and pull out their money; the bank cannot pay off everybody because it cannot change their loans into cash fast enough; and so the run on the bank is self-fulfilling. The bank will either fail or have to dump all of its long-term assets quickly in the market and take big losses.
The crisis of 2008–2009 was a classic financial panic but in a different institutional setting: not in a bank setting but in a broader financial market setting. As house prices fell in 2006 and 2007, for the reasons I described, people who had subprime mortgages were not able to make the payments. It was increasingly evide
nt that more and more were going to be delinquent or default, and that was going to impose losses on the financial firms, the investment vehicles they had created, and also on credit insurers like AIG. Unfortunately, the securities were extremely complex and financial firms’ monitoring of their own risks was not sufficiently strong. The problem was not just the losses. If you put together all the subprime mortgages in the United States and assumed they were all worthless, the total losses to the financial system would be about equivalent to one bad day in the stock market: they were not very big. The problem was that they were distributed throughout different securities and different places and nobody really knew where they were and who was going to bear the losses. So that created a lot of uncertainty in the financial markets.
As a result, wherever you had short-term funding, whether commercial paper or other types of short-term funding, the lenders refused to lend. We had all kinds of funding that was not deposit insured; it was so-called wholesale funding, which came from investors and other financial firms. Whenever there was doubt about a firm, as in a standard bank run, the investors, the lenders, and the counterparties would all pull back their money quickly for the same reason that depositors would pull their money out of a bank that was thought to be having trouble. So there was a whole series of runs, which generated huge pressures on key financial firms as they lost their funding and were forced to sell their assets quickly, and many important financial markets were badly disrupted. During the Depression, thousands of banks failed, but almost all of them, at least in the United States, were small banks (some larger banks failed in Europe). The difference in 2008 was, in addition to the many small banks that failed, there were also intense pressures on quite a few of the largest financial institutions in the United States.
Let’s look at some of the firms that came under intense pressure. Bear Stearns, a broker-dealer, came under intense pressure in the short-term funding markets in March 2008 and was sold to JPMorgan Chase with Fed assistance on March 16. Things calmed down a bit after that, and over the summer there was some hope that the financial crisis would moderate. But then in the late summer, things really began to pick up.
On September 7, 2008, Fannie and Freddie clearly were insolvent. They did not have enough capital to pay the losses on their mortgage guarantees. The Federal Reserve worked with Fannie and Freddie’s regulator and with the Treasury to determine the size of the shortfall, and over that weekend, the Treasury with the Fed’s assistance placed those firms into a form of limited bankruptcy called a conservatorship. At the same time, the Treasury got authorization from Congress to guarantee all of Fannie and Freddie’s obligations. So, the firms were in a partial bankruptcy but the U.S. government now guaranteed their mortgage-backed securities. So that protected those investors. That had to be done or else it would have been an enormous intensification of the crisis because investors all over the world held literally hundreds of billions of dollars’ worth of those securities.
In the middle of September, Lehman Brothers, a broker-dealer, had severe losses. It came under great pressure and could not find anybody either to buy it or to provide it with capital. And so on September 15th, it filed for bankruptcy. On the same day, Merrill Lynch, another big broker-dealer, was acquired by the Bank of America, basically saving the firm from potential collapse.
The next day, on September 16th, AIG, the largest multidimensional insurance company in the world, which had been selling credit insurance, came under enormous attack from people demanding cash either through margin requirements or through short-term funding. The Fed provided emergency liquidity assistance for AIG and prevented the firm from failing.
Washington Mutual was one of the biggest thrift companies, a big provider of subprime mortgages. It was closed by regulators on September 25th. After parts of the company were split off, JPMorgan Chase acquired this company as well. On October 3rd, Wachovia, one of the five biggest banks in the United States, came under serious pressure and was acquired by Wells Fargo, another large mortgage provider.
All the firms I am talking about were among the top ten or fifteen financial firms in the United States, and similar things were happening in Europe. So, this was not a situation where only small banks were affected. Here we had the largest, most complex international financial institutions on the brink of failure.
Let’s review the lessons from the Great Depression, which I discussed in the first lecture. First, the Fed did not do enough to stabilize the banking system in the 1930s, and so the lesson there is that in a financial panic, the central bank has to lend freely, according to Bagehot’s principle, to halt runs and to try to stabilize the financial system. Second, the Fed did not do enough to prevent deflation and contraction of the money supply in the 1930s, so the second lesson from the Great Depression is that you need to have accommodative monetary policy to help the economy avoid a deep recession. So, heeding those lessons, the Federal Reserve and the federal government took vigorous actions to stop the financial panic, working domestically with other agencies and internationally with foreign central banks and governments.
Now, one aspect of the crisis that, I think, does not receive enough attention is that it was global. In particular, Europe as well as the United States was suffering very severely from the crisis. But it was also a very impressive example of international cooperation.
On October 10, 2008, as it happened, there was a previously scheduled meeting in Washington of the G7 industrial countries, the seven largest industrial countries, and the central bank governors and the finance ministers of those seven countries met in Washington. Now, I will tell you a deep, dark secret: these big, high-profile international meetings are usually a terrible bore because much of the work is done in advance by the staff. We have a discussion, but the communiqué has already been written by the staff and in most cases what happens at the meeting is fairly routine. This was not one of those boring meetings. We essentially tore up the agenda and discussed what we should do. How should we work together to stop this crisis that was threatening the global financial system? In the end we came up with a statement of principles that was written from scratch, based on some Fed proposals. Among those principles were that we would work together to prevent the failure of any more systemically important financial institutions. This was after Lehman Brothers had failed. We would make sure that banks and other financial institutions had access to funding from central banks and capital from governments. We would work to restore depositor confidence and investor confidence, and then we would cooperate as much as possible to normalize credit markets. This was a global agreement, and the following week, the United Kingdom was the first to announce a comprehensive program to stabilize its banking system. The United States announced major steps to put capital into our banks, and so on. So a lot happened in the next couple of days after this meeting.
To show you how effectively this worked, Figure 24 graphs the interest rate charged on overnight loans between banks, the interbank interest rate. Normally, the overnight interest rate between banks is extremely low, far less than 1 percent, because banks need some place to park their money overnight and they have a lot of confidence that it is safe to lend to another large bank overnight. As you can see, starting in 2007 banks lost confidence in one another, which is shown by the increase in the rates they charged one another to make loans. For example, in 2007, you begin to see the pressures as house prices began to fall and there were increasing concerns about the quality of mortgage securities and the financial soundness of firms. In March 2008, you can see another little peak around the time Bear Stearns was forced to sell itself. That does not look like much, in comparison to what happened later, but that was a very difficult period. It was a period of quite sharp movements in financial markets and in funding markets. Now, look what happened in response to Bear Stearns’s liquidity problems. There was an enormous spike in interbank market rates, and probably not much lending was taking place even at those high rates. This indicated that suddenly th
ere was no trust whatsoever even between the largest financial institutions because nobody knew who was going to be next, who was going to fail, who was going to come under funding pressure.
Figure 24. Cost of Interbank Lending, 2005-2009
Note: London Interbank Offered Rate (LIBOR) minus overnight index swap rate Source: Bloomberg
Look what happened after the international announcements. Within a few days we began to see a reduction in funding pressure, and by early January there was an enormous improvement in the funding pressures in the banking system. This is a great example of international cooperation and it illustrates the point that this was not just a U.S. phenomenon, it was not just U.S. policy, it was not just the Federal Reserve. It really was a global cooperative effort, particularly between the United States and Europe.
The Federal Reserve and the Financial Crisis Page 7