|Financial Services Outlook logo 130|
What caused the financial crisis? Even though the Dodd-Frank Act (DFA) has been signed into law, this is still an important question. If we do not attribute the crisis to the right cause, we could well stumble into another crisis in the future; and if the DFA was directed at the wrong cause, we should consider its repeal. A Brookings Institution paper issued in late 2009 develops an interesting and plausible idea: the "great moderation"--the quiet period of almost continuous growth between 1982 and 2007--caused investors, managers, and regulators to believe that we had come to understand how the economy worked and how to tame the business cycle. This mistaken view in turn caused a decline in the normal aversion to risk, creating a housing bubble and the financial crisis. This is a compelling narrative and accounts for much of the risk taking that was observed in the period leading up to the crisis, but in the end it is no more than an interesting theory. The reality is that, in pursuit of a social policy to increase homeownership, the U.S. government became a willing buyer of an unprecedented number of subprime and other high-risk mortgages. This created a housing bubble of unprecedented size and duration, but only the taxpayers were taking the risks necessary to create this financial disaster.
Key points in this Outlook:
- Martin Neil Baily and Douglas J. Elliott's narrative--that a decline in risk aversion was the ultimate cause of the financial crisis--explains much of the risk taking and mistaken assumptions of government regulators, rating agencies, lenders, investors, and homebuyers before the financial crisis.
- While their argument is compelling and sophisticated, and explains the development of a housing bubble, it does not explain why half the mortgages in the housing bubble were subprime or otherwise low-quality.
- This must be attributed to a U.S. government social policy intended to increase homeownership, which made the government--with its ability to both spend and require others to do so--the willing buyer of an unprecedented number of these subprime and other high-risk mortgages.
- When the housing bubble deflated, the enormous number of delinquencies and defaults among these high-risk loans drove down housing prices, weakened financial institutions, and caused the financial crisis. No special risk aversion was necessary for all this to occur.
Last November, two highly respected Brookings Institution scholars, Martin Neil Baily and Douglas J. Elliott, published a paper entitled "Telling the Narrative of the Financial Crisis: Not Just a Housing Bubble." It is an important paper for two reasons. First, it recognizes that a narrative--a story that explains an event--influences the legislation or other public policy actions that follow. Second, as implied by the title of their paper, Baily and Elliott developed their own narrative for what caused the financial crisis, and they use it to argue that the crisis was not caused by government housing policies. This Outlook considers whether the Baily-Elliott narrative is a better explanation of the financial crisis than the housing policies of the U.S. government.
Writing in 2009, before the enactment of the DFA, Baily and Elliott began their essay by recognizing the importance of narratives. "Major crises," they note, "such as the recent financial crisis, usually end up being understood by the public in terms of some simple narrative, which then heavily influences the choices politicians make. We believe there are three major story lines still vying for acceptance by the public and that whichever one comes to dominate could strongly affect public policy." They observe in particular that "one of the earliest theories of the Great Depression was that it sprang from the crash on Wall Street, which came to be associated with financial manipulation by bankers and rich speculators. This created much of the impetus for the separation of commercial and investment banking and the creation of the Securities and Exchange Commission and associated laws to protect investors." These words are a reminder, not only of the importance of narratives in shaping policy responses, but also that certain impulses and ideas are carried forward over generations and resurface when those who have accepted them find the right opportunity.
Baily and Elliott identify three distinct narratives that have been used to explain the financial crisis. The first is the view that government housing policies caused the crisis "by inflating a housing bubble and mismanaging the resulting risks and problems, especially in regard to Fannie Mae and Freddie Mac. This narrative is popular among conservatives, particularly since it argues for a scaling back of government interventions in the economy and suggests less regulation, not more."
The second narrative is that "Wall Street created the crisis by reckless behavior, greed, and arrogant belief in its own ability to understand and manage excessively complex investments. . . . This narrative is popular with the left, but is accepted much more widely than that, including by a broad populist sentiment that sees large banks and large corporations as at the root of many of the country's economic problems. Many in the media have also adopted this position. . . . The housing part of the crisis is viewed as principally resulting from financiers pushing naïve consumers into taking on mortgages bigger than they could handle and which were structured to hide large fees and interest rates that would jump after a few years." This narrative, one might add, is easily recognizable as a lineal descendant of the ideas that motivated the New Deal.
The third narrative, and the one Baily and Elliott endorse, is that "the crisis was a very broad-based event with a wide range of people and institutions bearing responsibility, including many outside the United States. . . . Wall Street financial institutions failed to put in place or enforce the sound risk management processes and restraints that were needed . . . [and] government regulators did not adequately oversee these institutions, including, importantly, Fannie Mae and Freddie Mac. . . . Some of the federal government's actions to encourage home ownership also overshot and provided incentives for reckless behavior."
The DFA, adopted after the Baily-Elliott paper was published, demonstrates that they were correct about the importance of narratives. Congress enacted and the president signed legislation that rather faithfully reflected a combination of narratives 2 and 3. Anyone who followed the debate that preceded the enactment of the DFA saw Congress directly responding to elements of both narratives. The Consumer Financial Protection Bureau was certainly written into the act because the Democrats who controlled the process believed--or said they did--that "naïve consumers" were duped into buying homes they could not afford to create fees for the mortgage originators and the "financiers" (this closely follows narrative 2). This is clearly the modern analogue of the New Deal-era view that the Great Depression was caused by financial manipulation and the abuse of investors. Similarly, the stringent regulation and supervision prescribed by the DFA for the largest bank holding companies and other systemically relevant firms were a reaction to the view that regulation had not been tough enough and had allowed too much risk taking to occur. This response can be seen as the result of the general view that private decision making and markets need some government supervision or they will veer into crisis (basically, narrative 3).
There are, of course, elements of truth in both these narratives, but that will not be discussed here. This Outlook simply argues that narrative 1 is a better explanation of the financial crisis than narratives 2 and 3. In other words, government housing policy caused the financial crisis, not too little regulation, not predatory lending, and not excessive risk taking. To be sure, all those things occurred--they always will--but none of them was significant enough to cause a worldwide financial crisis. If we come to believe that these were the causes of the crisis, we will simply be inviting another crisis to creep up on us while we are looking the wrong way. A strikingly clear example of this possibility is the introduction in Congress in late September 2010 of a bill to extend the Community Reinvestment Act (CRA)--which currently applies only to insured banks and savings and loans (S&Ls)--to the rest of the financial system. The CRA requires insured banks and S&Ls to make loans (primarily mortgages) to borrowers in their service areas who are at or below 80 percent of the median income where they live. Banks must show that they are making such loans irrespective of whether the loans meet the bank's usual credit standards. As I will discuss later in this Outlook, that requirement contributed to the large number of subprime and other risky loans that failed in the financial crisis. If the narrative we adopt for what caused the financial crisis does not regard these loans as contributing factors--and narratives 2 and 3 certainly do not--then the conditions that gave rise to the crisis will eventually be repeated.
There is much more to narrative 1 than Baily and Elliott describe, but before introducing those points I will first outline in detail the case that Baily and Elliott make for their narrative 3, which I believe is the most sophisticated and important alternative to the idea that government housing policies created the financial crisis.
Narrative 3: The Decline in Risk Aversion
The principal characteristic of narrative 3 is its comprehensiveness. All the main actors in the financial system are implicated:
- Wall Street did not put in place sound risk-management processes;
- Government regulators did not properly or effectively oversee these processes or the banks, investment banks, and Fannie Mae and Freddie Mac;
- The rating agencies' models were flawed and the agencies themselves had conflicts of interest, allowing complex and ultimately toxic instruments to be released into the financial market;
- Borrowers obtained mortgages under false pretenses, and unregulated mortgage brokers took advantage of unsophisticated buyers; and
- Homebuyers mistakenly believed housing prices would always go up.
This approach--that virtually everyone was responsible for the financial crisis--is susceptible to the objection that it is blaming the crisis on a lot of random errors that all happened to occur at the same time. This is the central conceptual problem with narratives that cite multiple factors as causes: as "perfect storm" descriptions, they do not explain why all the various errors occurred together, except by chance, and what would have happened if they had not. Other perfect-storm explanations are unsatisfactory for this reason. However, Baily and Elliott introduce a new idea that addresses this problem rather neatly and distinguishes narrative 3 from many others. Their idea is this: "the principal underlying cause of the behaviors listed above was a major reduction in the risk premium [the additional interest rate required because of additional risk] resulting from 25 years of strong performance by the financial markets, encouraged by and associated with the ‘great moderation' in the macro-economy, whereby business cycles seemed almost to vanish."
Thus, Baily and Elliott are identifying a deeper cause of the financial crisis than has generally been outlined in conventional perfect-storm explanations, an element that in effect underlies the more superficial actions that have been cited as causes. That cause, explained in narrative 3, was a general relaxation in the usual fear of risk; this was induced by the "great moderation"--a period of general prosperity and growth that prevailed with few interruptions for the quarter century from 1982 to 2007.
During this period, they argue, "[m]any people, both experts and nonexperts, believed that central banks and governments around the world had learned the secrets necessary to tame the business cycle. . . . People learned to take risks. Not only would one expect on average to be rewarded, as the textbooks tell us, but actual experience showed it almost always paid off much more handsomely and with less pain than the theories said. Individuals learned a similar thing with housing. . . . Since homeowners were usually highly levered through mortgage debt [that is, they made low downpayments when they bought their homes], a fairly steady and decent return [on a home investment] became a very attractive levered return. . . . This increased willingness to take risks worked in dangerous combinations with the ‘easy money' conditions of the mid-2000s."
So the heart of the argument for narrative 3 is that the long period of growth in the economy, without any serious financial crises, taught investors, corporate managers, and consumers that the risks of excessive leverage were limited; they could acquire assets and make investments with very little money of their own and profit a lot when the assets appreciated in value. This unhealthy process was exacerbated by Federal Reserve policies that made large amounts of credit available to support risk taking. These elements account for most if not all of the errors made by banks, rating agencies, investors, and consumers in narrative 3.
A major implication of this argument is that it turns the housing bubble into just one of a number of bubbles that could have, and probably would have, developed if the housing bubble had not deflated first. "A key policy question," Baily and Elliott write, "is whether the housing bubble was the unique driver of the crisis or whether we might have had a damaging crisis even if housing had played a more minor role. . . . Few would dispute that other bubbles existed and could have burst in a painful way--the argument is whether those bubbles would have been more like the stock market crash of 1987 or the Long-Term Capital [Management] failure, both of which hurt, but neither of which dramatically damaged the economy. . . . However, we believe there would have been a painful recession due to the bursting of the more comprehensive financial bubble, even if housing had played a significantly more minor role."
In other words, the housing bubble was not unique; it was caused by the same underlying problem--the failure of nearly everyone to perceive the risks of high leverage. Rather than blaming the government for creating a huge housing bubble, Baily and Elliott blame investors, corporate managers, regulators, and homebuyers for failing to understand the risks involved when investors and consumers used high leverage to carry assets or buy homes with low downpayments or adjustable-rate mortgages.
After laying this foundation, the Baily-Elliott paper provides a number of examples that are intended to demonstrate that the general perception of risk had fallen to unprecedented lows by 2006 and 2007, before the financial crisis. They note the many credit losses that occurred after the housing bubble burst, including commercial real estate, ordinary business loans, and credit cards. "This does not rule out the possibility," they write, "that housing started the problem and the blow was so strong that it took down the other sectors. However, it is strongly suggestive that the imbalances in the other sectors were also very large prior to the housing bubble bursting, reinforcing the notion of a more comprehensive bubble made up of many sectoral bubbles. . . . [T]he problem did not have to begin with housing."
Moreover, if housing was the precipitating cause of the crisis, Baily and Elliott argue, sectors of the economy unrelated to housing would not have been among the first to fall. One example is the leveraged loan market, which had no apparent exposure to housing. If the housing bubble had not burst when it did, they argue, other bubbles would have grown so large that they would have been equally destructive when they ultimately burst.
Finally, according to Baily and Elliott, it is not clear that government policy was a necessary element of the housing bubble; conditions in the credit markets were sufficiently lax that the housing bubble could have grown to the size it did without any government help. Using market data, Baily and Elliott show, among other things, that price-to-earnings ratios of U.S. stocks were very high by historical standards in 2006-2007, that spreads over treasury rates of the sovereign bonds of emerging-market countries and U.S. high-yield corporate bonds were at historic lows, and that the International Monetary Fund's calculation of the average volatility forecast built into option pricing for a wide range of assets was at a low point in the 2003-2007 period. All of these data support the key point of the Baily-Elliott argument that the housing bubble--which appears to have been at least a trigger for the crash that became the financial crisis--was only one aspect of a much broader economy- and society-wide increase in leverage and risk taking. Accordingly, the government's role in promoting homeownership, whatever it was, was not the central cause of the financial crisis.
"Put another way," the authors write, returning the question to its public policy focus, "if we find ourselves in the future in another situation of widespread, excessive risk-taking, preventing problems in the housing sector is unlikely to be enough to prevent severe financial and economic problems when the bubble bursts."
In the next section, I will outline why the housing bubble was not just any bubble, why it was a sine qua non--a "but for" cause--of the financial crisis, and why it was sufficient to cause the crisis without any excessive risk taking by market participants.
FSO 2010-10-11 Figure 1
Narrative 1: The Role of Government Housing Policy
Since much of the Baily-Elliott paper is about bubbles, and whether the recent housing bubble was any more significant than other bubbles that might have been created by a market that had lost its fear of risk, any response should begin by examining the dimension of the housing bubble that began to deflate in 2007. Figure 1 is a chart prepared by the New York Times and based on the work of Robert J. Shiller. It shows the extraordinary growth of the 1997-2007 bubble, especially when compared to previous bubbles.
Two things stand out about the most recent bubble: it was much larger in real terms than any previous housing bubble, and it lasted more than twice as long as any previous bubble, especially the two that occurred before 1980 and 1990. The most recent bubble involved increases in real (not nominal) home prices of 80 percent over ten years, while the earlier ones involved increases of about 10 percent before they deflated. Asset bubbles in other sectors might be comparable, but none is likely to involve an asset that is one-sixth of the U.S. economy, and none--as I will show--was composed to such a large degree of weak and high-risk assets.
Why did this bubble last so long? This is an important question; bubbles get more destructive the longer they last because people who believe that prices will continue to rise stretch further and take more risks (use more leverage) to acquire assets that they believe will increase in value. This is exactly what happened in housing, as more and more people took out adjustable-rate mortgages with low "teaser" rates, or mortgages with low or no downpayments, so they could afford the monthly payment on homes that they thought were going to rise in value. A bubble, as Baily and Elliott point out in their paper, feeds on itself because as long as it lasts it disguises the risks that are being taken by those buying the inflating asset. In the housing bubble, for example, there were few losses from this risk taking until the bubble burst because those who could not afford to pay for their homes could always refinance by using the higher appraised value of the home in a rising market. When the music stopped, these people were left without a method of refinancing, and the delinquencies and defaults began. As Warren Buffett said, "when the tide goes out, you can see who's swimming naked."
FSO 2010-10-11 Table 1
Baily and Elliott describe how a bubble finally collapses as follows: "The work on bubbles by Shiller and others makes clear that a bubble can grow for a very long time before it hits a natural limit. Such a limit will always be reached, because a key part of the dynamic of a bubble, like a Ponzi scheme, is that it needs to suck in increasing amounts of money in order to continue . . . leaving the bubble vulnerable to any piece of bad news that causes questions about the excessive valuation." This is an accurate portrayal, and it emphasizes the fact that bubbles naturally end when those who are adding funds become nervous about excessive valuations.
In the case of the 1997-2007 bubble, however, one of the major contributors was not motivated by profit or concerned about risk. It was the U.S. government, following a social policy--using government's financial and regulatory power to boost homeownership by increasing the credit available to low-income borrowers. This was done, first, by requiring Fannie Mae and Freddie Mac to acquire increasing numbers of "affordable"-housing loans. An affordable-housing loan was one made to a borrower at or below the median income in the area where the borrower lived. This was required in legislation that Congress adopted in the early 1990s, the Housing and Community Development Act of 1992. Initially, the act required that 30 percent of all loans Fannie and Freddie acquired had to be affordable, but the Department of Housing and Urban Development (HUD) was given authority under the act to increase these requirements. It did so repeatedly from 1995 until 2007, so that by 2007, 55 percent of all loans Fannie and Freddie acquired had to be affordable, with a subgoal of 25 percent for low-low income borrowers who were at or below 60 percent of the area median income (see table 1).
In effect, Fannie and Freddie were put into competition with the Federal Housing Administration (FHA), which also was required to insure mortgages for borrowers at or below the median income, and with insured banks, which were required under CRA to make loans to borrowers who were at or below 80 percent of the median income in the banks' service areas. It should be obvious that when Fannie and Freddie, FHA, and all insured banks are trying to find the same borrowers--those who are at or below the median income (or 80 percent of the median income in the case of CRA) where they live--borrowers who could meet the standards for prime loans (a substantial downpayment, unblemished credit, a steady job, and an income that would support a mortgage) might be difficult to find. But all these lenders were required by law or regulation to make the loans, so they had to settle for lower-quality loans than they would prefer or had customarily required. Moreover, in competing with one another, they paid more for these loans than they were worth on a risk-adjusted basis and thus underpriced the risks. The result by 2008, according to research by my AEI colleague Edward Pinto, was a telling distribution of subprime and other high-risk loans in the U.S. financial system.
FSO 2010-10-11 Table 2
As shown in table 2, two-thirds of the subprime and other high-risk loans were held or guaranteed by government entities or entities required by the government to acquire, guarantee, or insure the loans. This makes it very clear that the great bubble of 1997-2007 did not develop naturally as an ordinary bubble; it was driven by a government social policy intended to increase homeownership in the United States. For this reason, that bubble cannot be classified as just another bubble among many that the Baily-Elliott paper describes. Not only was it larger than any other known bubble in its dollar amount, but because of its provenance as an artifact of government policy it lasted well beyond the time when other bubbles would naturally have collapsed. For this reason alone, it was more destructive than any other bubble in history when it finally burst and, because of the low quality of the mortgages it contained, resulted in an unprecedented number of delinquencies and defaults. Table 3 on page 8 shows the delinquency rates on the 27 million subprime and Alt-A mortgages that were in the bubble before the financial crisis.
The bubble was more destructive than others for yet another reason. While it is certainly true that as bubbles grow older their asset quality declines, this phenomenon is seldom forced and exacerbated by government policy. Without the government's active involvement, it is doubtful that the 1997-2007 bubble would have contained as many as 27 million subprime and Alt-A loans. If the bubble had developed naturally, in the ordinary way, it could conceivably have contained the 7.8 million loans that were eventually securitized by Countrywide (the largest originator and issuer of securities backed by subprime mortgages, or PMBS), as well as other originators and the Wall Street banks. But even then the quality of the mortgages in these PMBS issuances would have been higher than those that Countrywide and the others ultimately contributed to the recent bubble. This is because the government agencies or government-regulated entities like Fannie and Freddie could afford to pay more than their private-sector competitors for the subprime mortgages they needed; they acquired the cream of a bad crop, and that drove Countrywide and other securitizers further out on the risk curve to find mortgages they could securitize. Without competition from Fannie and Freddie and the banks under CRA, the PMBS ultimately securitized would have been far less "toxic" when the bubble collapsed.
FSO 2010-10-11 Table 3
In any event, these PMBS were less than one-third of the high-risk mortgages outstanding. If the FHA is included, because its role was to make subprime loans, we can increase the total to almost half of all the high-risk loans--still only 25 percent of all mortgages outstanding in the United States in 2008. There is clearly a relationship, as one would expect, between the quality of the mortgages in the bubble and the bubble's destructiveness when it deflates. Without Fannie and Freddie and the CRA obligations of insured banks, the number of subprime and other high-risk mortgages in the financial system would have been, at most, only half the 27 million it eventually became.
It is impossible to know what would have happened if the bubble had burst when only 25 percent of the mortgages in the United States were subprime or otherwise high-risk, but clearly the damage to the financial system and the economy would have been much lower. For example, in the housing bubble that ended in 1979, when almost all loans were traditional mortgages, foreclosure starts in the ensuing slump peaked at just 0.87 percent in 1983. In the next bubble, which ended in 1989 and primarily involved traditional mortgages, foreclosure starts reached 1.32 percent in 1994. But in 2009, when half of all outstanding mortgages were subprime or otherwise high-risk, foreclosure starts jumped to a record of almost 5 percent by the middle of 2010. And this was at a time when the banks were going slow on foreclosures because government programs were encouraging the renegotiation of delinquent mortgages, and banks did not want either the capital writedowns or the bad publicity that comes with large numbers of foreclosures.
What we know is that almost 50 percent of all mortgages outstanding in the United States in 2008 were subprime or otherwise deficient and high-risk loans. The fact that two-thirds of these mortgages were on the balance sheets of government agencies, or firms required to buy them by government regulations, is irrefutable evidence that the government's housing policies were responsible for most of the weak mortgages that became delinquent and defaulted in unprecedented numbers when the housing bubble collapsed. Under these circumstances, one does not have to reach for reduced fear of risk as the explanation for the financial crisis. It is right there in the housing data.
Baily and Elliott make a strong case for explaining the financial crisis as the result of a general decline in risk aversion because of the effect of the great moderation--the period from 1982 to 2007 when it seemed that we understood the causes of financial crises and had found a way to avoid or mitigate them. The evidence for a general weakening in risk aversion coming out of this period is plausible. But the Baily-Elliott narrative assumes that the 1997-2007 housing bubble was also caused by this factor, and that seems implausible. The extraordinary lengths to which the government went to force private-sector lending that would not otherwise have occurred--through affordable-housing requirements for Fannie and Freddie as well as demands on FHA and on the banks under CRA--shows that the housing bubble that ended in 2007 was not a natural occurrence or the result of mere risk aversion. If it had been, there would have been no need for these government programs.
The housing bubble that finally burst in 2007 was driven by a U.S. government social policy that was intended to increase homeownership in the United States and was thus not subject to the usual limits on the length and size of asset bubbles. As such, it was far larger and lasted far longer than any other bubble in modern times, and, when it deflated, the vast number of poor-quality mortgages it contained defaulted at unprecedented rates. This drove down U.S. housing values and caused the weakening of financial institutions around the world that we know as the financial crisis.
Market participants were certainly taking risks as the bubble grew, and it may well be, as Baily and Elliott posit, that this private risk taking was greater than in the past. But the facts show that the bubble was inflated by a government social policy that created a vast number of subprime and Alt-A mortgages that would not otherwise have existed. And the risks associated with this policy--which could produce losses of more than $400 billion at Fannie and Freddie alone--were being taken by only one unwitting group: the taxpayers.
Peter J. Wallison (firstname.lastname@example.org) is the Arthur F. Burns Fellow in Financial Policy Studies at AEI.
1. Douglas J. Elliott and Martin Neil Baily, "Telling the Narrative of the Financial Crisis: Not Just a Housing Bubble," Brookings Institution, November 23, 2009, available at www.brookings.edu/papers/2009/1123_narrative_elliott_baily.aspx (accessed October 27, 2010).
2. See, for example, Peter J. Wallison, "Deregulation and the Financial Crisis: Another Urban Myth," AEI Financial Services Outlook (October 2009), available at www.aei.org/outlook/100089; Peter J. Wallison, "The Dodd-Frank Act: Creative Destruction, Destroyed," AEI Financial Services Outlook (July-August 2010), available at www.aei.org/outlook/100983; and Peter J. Wallison, "Going Cold Turkey: Three Ways to End Fannie and Freddie without Slicing Up the Taxpayers," AEI Financial Services Outlook (September 2010), available at www.aei.org/outlook/100993.
3. Douglas J. Elliott and Martin Neil Baily, "Telling the Narrative of the Financial Crisis: Not Just a Housing Bubble," 1.
4. Ibid., 2.
5. Ibid., 3.
6. Ibid., 4.
7. To Amend the Community Reinvestment Act of 1977 to Improve the Assessments of Regulated Financial Institutions, and for Other Purposes, 111th Cong., 2d sess. (September 29, 2010).
8. Douglas J. Elliott and Martin Neil Baily, "Telling the Narrative of the Financial Crisis: Not Just a Housing Bubble," 4.
9. Ibid., 5. Baily and Elliott are not the only ones to make this argument. Raghuram G. Rajan, a professor of finance at the Chicago Booth School of Business, makes the same general argument in his book Fault Lines: How Hidden Fractures Still Threaten the World Economy (Princeton, NJ: Princeton University Press, 2010).
10. Douglas J. Elliott and Martin Neil Baily, "Telling the Narrative of the Financial Crisis: Not Just a Housing Bubble," 12.
11. Ibid., 7.
12. Ibid., 8.
13. Ibid., 8-11.
14. Ibid., 15.
15. Housing and Community Development Act, Public Law 102-550, 102d Cong., 2d sess. (October 28, 1992).
16. Mortgage Bankers Association, National Delinquency Survey (Washington, DC, 2010), available through www.mbaa.org/ResearchandForecasts/ProductsandSurveys/NationalDelinquencySurvey.htm (accessed November 1, 2010).