This is a very, very early draft of an explanation of the financial crisis, which will eventually make it into the next edition of my corporate finance textbook. it has not only not yet been reread and rewritten (which I do about five times to each chapter before I usually release it), it has not even been fact-checked yet. I am only posting it now, because it is so topical.
Although not strictly corporate finance, no textbook is complete without a discussion of the financial system that lead to the financial crisis of 2008. After all, the crisis affected not only Wall Street, investment funds, and individual investors, but also every corporation that needed to borrow money.
As we always do, we make up a simple example. There are 5 individuals in our city, Mainstreet-USA, who want to build houses or start business. To do so, they need to borrow money. This simple setup will be enough to explain how finance developed over the decades and what went wrong in 2008:
| Who | Wants to Borrow | |
|---|---|---|
| Wealthy person | A | $700,000 |
| Upper Middle-Class person | B | $300,000 |
| Lower Middle-Class person | C | $200,000 |
| Upper Lower-Class person | D | $100,000 |
| Lower Lower-Class person | E | $50,000 |
Of course, when a bank (or a thrift---they are really the same kind of institution) extends a loan, the borrower usually invests this money, generating jobs, economic development, and so on. If the loan was wisely extended, the borrower will repay the bank, the bank will earn a profit, and all the bank depositors will receive their money bank and with good interest.
In the "good old days"---which were really more like bad old days---local lending was done primarily by the bank in town. Often, this local bank had a de-facto monopoly on lending. In most states, this monopoly was even protected by rules and regulations that state legislatures had enacted (presumably, with some help from bank lobbyists).
How would such a local bank decide who to lend to? It would first look at its retail deposits, which often constituted most of its capital, and then decide which loans it wanted to extend. Let's say our local bank had around $1 million to lend. In this case, it would most likely decide that A and B were the best risks. It would then make loans to them at a high interest rate---remember, the local bank really was the only game in town. If A or B wanted a loan, they had few other choices. Not surprisingly, many local banks were very profitable---and, more often than not, the richest guy in town (A in our example) was also the owner of the bank. If you were not A or B, you were often simply out of luck as far as credit was concerned. Not surprisingly, women, minorities, and immigrants were rarely in the A or B categories.
On the plus side, bank lending was so profitable that many banks were stable and build up good capital cashions. However, even when fat, these banks' situations was still precarious: Local lending is still very risky, because it exposes the local banks heavily to the local economy. Local banks had little diversification in their asset base. When the local economy would take a dive for the worse, many borrowers would not repay at the same time. Thus, despite their high average profitability, many local banks would often go belly up, too. Eventually, many decided that they would rather invest depositors' money in the stock market, not only because equity investments had traditionally yielded higher rates of returns than lending, but also because it spread the risk across more than just local borrowers. While the stock market went up, banks could pay higher interest rates to depositors and comfort themselves that they had made prudent decisions, protecting their depositors from local economy risk.
On the minus side, one law of nature is that what can go wrong ultimately will go wrong---and in the Great Depression, everything went spectacularly wrong. Both the stock market took a deep dive, and many local borrowers defaulted. Moreover, if you had deposited your money in your local bank, and you would see a line in front of your bank, you would wonder whether your bank had lost money. In this case, your smartest choice was to queue up to withdraw your deposits, too, just in case the bank was "underwater." and would have to close shop. Consequently, it was in the interest of other depositors also to run on the bank in order to withdraw all their money while they still could. In turn, this led to more bank collapses, even for banks who had not made any bad loans or invested in the stock market. As a result of such bank runs, the entire financial system effectively collapsed in the 1930s. (Ben Bernanke, now Chairman of the Federal Reserve, made his academic career showing how this financial shutdown had desastrous effects on the real economy, and transformed what might otherwise have been just a recession into a deep depression.)
When FDR came into power, the Federal Government enacted many economic laws in its New Deal. We will just focus on the most important banking and mortgage-related innovations.
The Glass-Steagall Act, which was in force until the late 1990s, created a distinction between commercial investment banks and ordinary consumer banks. Ordinary banks became subjects to regulatory oversight by the Federal Reserve System, and were not allowed to invest in many risky assets, such as the stock market. (In 2008, the last commercial banks converted into ordinary banks in order to qualify for Federal Reserve assistance.)
The Feds also created deposit insurance. This meant that the government forced consumer banks to maintain certain capital requirements, but in return the Federal Deposit Insurance Corporation (FDIC), guaranteed the deposits of retail clients of these banks. (You can find the sign on the doors of almost any bank these days.) FDIC insurance means that, as an ordinary depositor, you no longer have to run on the bank even if you fear that your bank is in trouble. In turn, with fewer runs by their retail depositors, this meant that fewer banks ever got into trouble to begin with. In fact, the FDIC practically eliminated runs on retail banks for many decades to come.
Finally, the Federal Government revolutionized the mortgage market. It created a government agency called Federal National Mortgage Association (FNMA, more commonly known by its moniker Fannie Mae). Fannie purchases mortgage loans made by banks if these mortgages follow a set of standard rules. Qualifying loans are called conforming. The interest rate on such loans is usually fairly similar across the entire United States and not too different across different types of borrowers. The mortgage loans are not allowed to be too large (e.g., $417,000 in 2008 for an ordinary, single-family home; 20% downpayment). This means that Fannie loans are more useful to the middle-class than the wealthy. However, the loans require fairly stringent qualifications by the borrower, which also meant that few lower class individuals could qualify.
All in all, the legislations of the 1930s were a huge step forward from the old days. Your local bank was no longer as heavily risk-exposed to the local economy. It could now extend mortgages to B, C and sometimes even to D, because it could immediately resell them to Fannie. Moreover, with its original capital freed up, it could now make more loans to other parties (such as companies or poorer clients) that previously did not have access to credit. In sum, Fannie allowed banks to extend more loans and at lower interest rates. In fact, it was Fannie that made it possible for most Americans to own their own homes.
But these advantages did not come without drawbacks. First, the New Deal regulations introduced decades of government regulations, redtape, compliance costs, etc. More importantly, the disadvantage of having Fannie was that local banks mostly lost their incentives to determine in great detail whether individual borrowers would likely repay their loans. As long as the loans were conforming, the local banks would make money by issuing the mortgages and then selling them on. Fannie relied on the qualification criteria of borrowers, based on honest disclosure by home buyers, and on the law of large numbers, to keep the mortgage default rates within reasonable levels. Occasionally, it also meant that if you just failed to qualify for a conforming mortgage, you could find yourself in a situation in which it was hard to find anyone to pay attention to you. If you could get a loan at all, it would came with a much higher interest rate than that charged for conforming mortgages.
Not much changed until 1968, when most of Fannie was privatized. (The remaining public entity is now called GNMAGinnie Mae] and it underwrites mortgages primarily to Veterans.) To avoid a monopoly, a second mortgage company, the Federal Home Loan Mortgage Corporate (also called FHLMC or Freddie Mac) was created in 1970. Because Fannie and Freddie are really quite similar, we shall consider them as the same kind of institution.
Fannie's business model as a private corporation was to purchase conforming mortgages from local banks and thrifts, and fund these purchases by issuing new securities, mostly bonds. Just like ordinary firms, when fewer mortgages defaulted then expected, Fannie would earn more of a profit. If more mortgages defaulted, it would lose money. For many decades, although it lost money on some mortgages, Fannie was profitable year after year. As real estate values continued to rise, mortgage borrowers paid more than what was generally anticipated. Fannie paid a good share of its profits to its shareholders and some to its executives. (It is a matter of debate what fraction of its social surplus went to homeowners in the form of lower interest rates.) As of 2008, Fannie and Freddie together held almost $1.7 trillion in assets and liabilities---a staggering number.
Although Fannie and Freddie were legally private companies, it was always believed that if either failed, the government would have no choice but to rescue them (a situation which indeed came true in 2008). Thus, they were called Government sponsored entitiesGSE). Because they were considered quasi-governmental, the two agencies enjoyed a credit rating of AAA and low borrowing costs---despite enormous leverage. The market for mortgages issued by Fannie and Freddie eventually became the largest financial market in the world, exceeding even that for the debt of the U.S. Treasury.
In 1970, another financial innovation occurred. Ginnie Mae created the first pass-through mortgage backed security (MBS). Fannie and Freddie soon entered this business, and in 1977, the major investment banks jumped in, too.
Let me illustrate this business with a simple, though exaggerated, example of how mortgage bundling worked. We assume that there are only three repayment dates (rather than 360 monthly payments). Ginnie would buy three conforming loans with the following payment patterns from a local bank:
| Who | Wants to Borrow | Rate | First Repayment | Second Repayment | Third Repayment | |
|---|---|---|---|---|---|---|
| Upper Middle-Class | B | $300,000 | 4.9% | $110,000 | $110,000 | $110,000 |
| Lower Middle-Class | C | $200,000 | 6.1% | $75,000 | $75,000 | $75,000 |
| Lower-Class | D | $100,000 | 9.7% | $40,000 | $40,000 | $40,000 |
| Total | $600,000 | 6.1% | $225,000 | $225,000 | $225,000 | |
Ginnie would then turn around and create the equivalent of a "mini-firm," more formally called a special interest vehicle check. Let's call this one "Mainstreet-USA." Its assets are the three bundled mortgages it owns. It also has some expected costs (such as mortgage administration costs and some allowance for non-payments of some mortgages). Let's presume this allowance is $10,000 per period in our example:
| Financial Security | First Payment | Second Repayment | Third Repayment |
|---|---|---|---|
| Main-Street-USA | $215,000 | $215,000 | $215,000 |
Value of Main-Street-USA = $215,000/1.03 + $215,000/1.03^2 + $215,000/1.03^3 ∼ $608,151
Thus, Ginnie would earn $608,151-$605,000= $3,151 in immediate profits.
The most interesting aspect of a MBS is the analysis of where the risk lies and who ultimately carries it.
First, there is the default risk: When the issuer of these pass-throughs is Ginnie, Fannie, or Freddie, they guaranteed the performance of the mortgages to the buyers of Main-Street-USA. With the government standing firmly behind Ginnie (and de-facto also behind Fannie and Freddie), Main-Street-USA buyers were therefore protected from default risk. When the issuer was an investment bank, they often purchased credit risk insurance from an insurance giant (like AIG) that would guarantee the payments in exchange for an upfront fee. (This is called credit enhancement.)
In a perfect capital market, an issuer should be able to sell the bundled mortgage securities with credit enhancement for about the same amount of money that buyers would pay. (The fact that issuing these mortgage-backed securities was a very profitable business should thus have set off alarm bells left and right---it should not have been easy to manufacture money out of nothing. But I am getting ahead of myself.).
Second, there was an interest rate risk, which was increased by a peculiar feature of consumer mortgages: Homeowners can refinance at will and/or to pay down principal. If interest rates fall, many homeowners will refinance in droves, and the security buyers will no longer receive the fat interest receipts they were hoping for.) In contrast, if interest rates rise, few homeowners will prepay their mortgage. This is called prepayment risk. It is illustrated in Figure ?:
| Figure: Payment Patterns and Prepayment Risk | |
|---|---|
Statutory 30-Year Mortgage Payments | |
Payments if Interest Rates Rise  |
Payments if Interest Rates Fall  |
For the first few decades after inception, few mortgages defaulted (thus leaving Fannie holding the bag [of risk]), but many mortgages were repaid early (thus leaving the securities buyer holding the bag).
In 1980, the next advance in financial innovation occurred. There were many funds, such as pension funds, who were unwilling to purchase pass-through mortgage backed securities because they disliked the prepayment risk. Some other investors wanted to purchase riskier securities. Yet other investors wanted securities that were much shorter-term than mortgages. And then there were the Japanese investors, who were willing to pay more for securities which paid only principal but not interest (for Japanese tax reasons).
In response to heterogeneous investor demand, Salomon Brothers invented the first Collateralized Mortgage Obligation (CMO). Despite its fear-inspiring name, the idea behind it is really rather simple. One or more mortgage bundles, often themselves the kind of pass-through mortgage backed security that we just discussed, are purchased up by an investment bank and then resold in a different format that buyers find more apetizing. You can think of the purchased MBS itself as the assets of a newly created firm. This new firm's assets are the future payments from the mortgages and/or Fannie. The investment bank can now decide on a better capital structure for this firm, in the sense that the price for this new firm's capital structure exceeds the prices at which the investment bank bought the assets. It is simply Modigliani and Miller in action.
Return to our example. Let's presume that our investment bank has purchased the Main-Street-USA security for $608,151. The payments of Main-Street-USA are what they are. Ultimately, taking our Ginnie MBS and placing it into a CMO vehicle makes sense only if the sum-total that bond buyers were willing to pay would be more than $608,151 if they could just purchase what they really wanted. For example, let's say that some bond buyers really wanted securities that pay off sooner. Take an extreme example, let's assume that there was one buyer who was willing to pay $220,000 for receiving $225,000 for the first payment; and another buyer who was willing to pay $390,000 for the remaining two payments. Together, the two buyers are willing to pay $610,000 for Main-Street-USA.
| Mortgagee | Wants to Borrow | Rate | First Repayment | Second Repayment | Third Repayment | |
|---|---|---|---|---|---|---|
| Upper Middle-Class | B | $300,000 | 4.9% | $110,000 | $110,000 | $110,000 |
| Lower Middle-Class | C | $200,000 | 6.1% | $75,000 | $75,000 | $75,000 |
| Lower-Class | D | $100,000 | 9.7% | $40,000 | $40,000 | $40,000 |
| Total | $600,000 | 6.1% | $225,000 | $225,000 | $225,000 | |
| Fannie Mae Financial Security | Buyer Pays | Rate | First Payment | Second Repayment | Third Repayment | |
| Main-Street-USA | $608,151 | 3.0% | $215,000 | $215,000 | $215,000 | |
| Collateralized Mortgage Obligations | Buyer Pays | Rate | First Payment | Second Repayment | Third Repayment | |
| Tranche A | $210,000 | 2.4% | $215,000 | $0 | $0 | |
| Tranche B | $400,000 | 3.5% | $0 | $215,000 | $215,000 | |
The two securities are called CMOtranches. In our case, Tranche B is riskier than Tranche A: over the long run, interest rates are more variable, and homeowners may prepay mortgages in droves. Thus, Tranche B is offering a higher interest rate than Tranche A. Be this as it may, at this point, the investment bank has just earned for itself $610,000 — $608,151= $1,849.
The CMO business model was very successful. There were apparently many buyers quite eager to purchase tranched securities instead of the original MBS securities---and willing to pay for it. Thus, investment banks were bidding up the prices that they were willing to pay for MBS's, which in turn increased the prices that the GSEs were willing to pay local banks. In turn, banks became more eager to originate conforming loans. In turn, this increased the access to credit and lowered the interest rate for borrowers on main street USA.
At some point, the investment banks realized that they did not have to limit these services to securities issued by Fannie. In fact, they could bundle many different types of loans together, whether issued by Fannie or not. For example, banks could first bundle many many non-conforming jumbo mortgage loans together, and then sell many different tranche claims on their payments. Or they could bundle mortgages that do not qualify for conforming status because the borrower does not qualify on credit rating score or sufficient downpayment ("subprime loans"). Or they could bundle student loans or credit card loans together. As long as these loans could be purchased for cheaper from local banks than what they could be sold for to financial investors, this was a good business model for the investment banks---and for the borrowers, who received unprecedented access to low-cost credit.
With Fannie out of the picture, the credit risk became much more important, however: mortgage borrowers could default. Someone other than the GSEs had to absorb the risk. For example, let's consider a credit card issuer which had extended 10,000 loans for $1,700 each, and with $2,000 due each. Let's say that the common estimate was that 500 borrowers would default completely (and no one else would), leaving $19 million as expected payment. If your local bank credit card issuer had to carry this risk, after it had extended credit to these 10,000 clients, it may not have wanted to qualify on any more credit card purchasers. If you needed money then, you might have been out of luck---in fact, in the extreme, your suddenly limited access to credit might even have caused you in the past to have to close your business.
However, the investment banks came to the rescue. In their new business model, they would purchase the loans from the credit card issuer for, say, $18 million. With $17 million credit extended to its clients, the card issuer had made $1 million in immediate profit. It could now immediately lend more money to other clients---a win for the bank and the consumers. The investment banks would now own claims to $20 million, with an expected payout of $19 million, that it had purchased for $18 million. Like the credit card issuer, it would want to sell these loans to third parties for as high a price as possible---but as quickly as possible, so that it could do the same transaction again and so that it would not be exposed to the risk of unexpectedly high default rates.
One way to sell these 10,000 credit card loans to potential buyers is to sell, say, 100 securities, each for $185,000. When a credit card borrower does not pay, each security is hit the same way. For example, if only 1,500 borrowers pay, each security ends up with only $150,000. If all borrowers pay up, each security ends up with $200,000.
However, as with CMO'ed Fannie bonds, there are some investment funds that want to buy only safe bonds, even if they earn a lot less in interest. Other investment funds are ok with purchasing riskier securities, because they promise higher interest rates. Let's use an extreme example, in which either 9,000 card holders repay, 9,500 repay, or 10,000 repay. The most important difference is that, because these are not mortgages, we have to change the name now to Collateralized Debt Obligation (CDO).
| Number of Repayers | 9,000 repay | 9,500 repay | 10,000 repay |
|---|---|---|---|
| Structure 1: Simple Security on repayment | $18 million | $19 million | $20 million |
| Structure 2: CDO Tranche A | $18 million | $18 million | $18 million |
| Structure 2: CDO Tranche B | $0 million | $1 million | $2 million |
If this reminds you of how firms are financed by debt and equity, you are exactly right. The firm here is the underlying loans of the credit card customers, and they are financed by a senior claim and a junior claim. If the investment bank can sell structure 1 for, say, $18.25 million (an expected interest rate of 4.1%); and CDO Tranche A for $17.5 million (a 2.9% interest rate) and CDO Tranche B for $950,000, then CDO'ing into tranches earns the investment bank money. Ultimately, this built the CDO market into one that worked with trillions of dollars (no, this is not a typo).
I have skipped over some details in my explanation: for example, someone had to make sure that the credit card customer payments would end up in the hands of the CDO Tranche owners. The CDO Tranche owners would have to be assured by lawyers, accountants, and rating agencies that they were actually purchasing what they were thinking they were purchasing. Some of the covenants and contracts that went into these CDOs were literally many hundred pages long. To further assure the Tranche A buyers that the securities were safe, investment banks would purchase additional insurance from large insurance companies, such as AIG.
On the plus side, this financial innovation allowed consumers to receive credit as unprecedently high levels and at relatively low interest rates. Students, credit card issuers, and poor housebuyers could afford purchases that they could never have afforded otherwise. (When borrowers use credit for choices that are not wise for themselves, it is hard to fault finance. After all, we are not a nanny state, in which the state makes buyers' decisions for them.) Many poor and moderately wealthy family gained tremendously, especially in the 1990s and the beginning of the 2000s, both from purchases of house and from college education that this additional credit allowed them. Of course, the credit card issuers, mortgage lenders, student lenderss, investment banks, lawyers, accountants, and rating agencies also did very well for themselves. Everybody won. This was easy in an era in which everything went well---the economy mostly boomed, education was a great investment, and house prices steadily increased.
Omits many other innovations, such as structured investment vehicles.
The principles that made CDOs work also led to its failure. Perhaps this was because they were too successful for too long, making everybody too confident.
The advantage to your local bank extending you credit is that the bank really wants to get its lending right. If it extends credit to you, and you fail to repay the loan, the local bank loses. Your local bank will err on the side of not giving you credit, even if you deserve it.
The advantage of the CDO market is that is much less risk-averse and capital-constrained than your local bank. It is also its disadvantage:
In some ways, the best line of defense should ultimately have been the smarts of the CDO buyers---as it is in any market. However, despite fancy degrees, titles, and control of billions of dollars, most CDO buyers were either remarkably unsophisticated, and pampered by decades of profitable CDO purchases, or remarkably conflicted. They often cared more about looking good this year in terms of the yields they were supposedly earning than in assessing the real risk. More often than not, the CDO buyers relied on diversification and the rating agencies to assure their own investors that they held only high-quality securities.
Both the rating agencies and the ultimate buyers of these CDOs should have been a lot more suspicious than they were. It was clear that CDOs in the 2000s were structured in a way that made it intentionally difficult to value them. These securities had not only books full of contractual provisions, but contractual features that are hard to explain otherwise. (Some CDOs were themselves on other CDOs, called CDOs-squared.)
The rating agencies could just have declared these securities as "suspect," not because a technical model should have valued them better, but because it should have been obvious to them that the design of the securities was such that they (the rating agencies) could not easily assess the risk. In fact, as you already know (from Chapter...), absent major market problems, the appropriate risk of the package should be equal to the weighted risk of its components. The complex repackaging magic violated the basic law of nature: it made it appear as if an underlying C rated security had been split into one C-rated security, one B-rated security, and one A-rated security. Nature demands that if one component becomes an A, then the rest should be rated lower than C. However, by the time this became obvious, about two-thirds of the business of the rating agencies had become rating these CDOs. It was not in their interest to dig deeper.
The buyers should have known this themselves, too. However, most of the ultimate owners of these CDOs (which know often sell for as little as 10 cents on the dollar) had little knowledge or influence either: they were the shareholders of the insurance companies (which wrote insurance for some of these CDOs; AIG went bankrupt over it), the shareholders of the investment banks (which ultimately kept many CDOs on their own books at inflated values), and the participants in many institutional funds which purchased many of these CDOs. In all of these cases, the management appeared to have made money for years and years, leading to large bonus payments to themselves.
To make matters worse, in the mid-2000s, Washington had pressured Fannie and Freddie to make more mortgages to poorer people. This led to even more lending to house buyers who should have never purchased houses to begin with. In the beginning of the crisis, this even accelerated, as politicians believed that easier access to credit could prop up stagnating housing values.
Of course, risk exposure was fairly unimportant while the economy and housing values went up---and they did for decades. It would have even been ok if only a few uncorrelated assets in the economy had lost value (see Chapter ). However, when house prices tanked economy-wide in 2007 and 2008, everything went wrong:
Did the investors get what they deserved? Yes and no. They did partake in the profits in the good years. But they also took all the losses. Ultimately, it was the executives and fund managers that made the decisions, not the shareholders. (See Corporate Governance chapter.) They gambled with the shareholders' money, won for a long time, and only finally lost all.
A good analogy for the crisis would be that it was like an airplane crash. It is not that the executives wanted to bring down the plane. But they neglected to carry out all but the most essential airplane maintenance, just enough to keep the airplane flying, but not enough to keep it safe. This strategy made modest amounts of money most of the time, which the executives of the investment banks, of the rating agencies, of the investment funds, of the insurance companies, of the GSAs promptly paid out in equity dividends and bonuses. It worked for decades. But in 2008, an airplane crashed. It was inevitable.
Describe how bank runs came back.