The railroad crisis era in the United States and Europe had ended, but financial crises would continue. China would see financial crises in 1911 and both the 1920s and 1930s as it struggled through wars, revolutions, and economic turmoil to establish a succession to the imperial government overthrown in 1911.197
The United States and Britain would see financial crisis again in July 1914—in the United States by way of overlending and in Britain by having lent too much in countries that were suddenly its enemies. But World War I quickly overshadowed and subsumed those crises.
Next came the Great Depression, the U.S. crisis of the 1980s, and Japan’s crisis of 1990, each of which we have already examined. Having now also seen the pattern of crises through the 1800s and early 1900s, we can turn to one of the most recent and calamitous crises of them all: the Global Crisis of 2008. With the perspective of history, and an understanding of the core private debt–fueled overcapacity in financial crisis, one thing is chillingly clear. The Great Recession might have been foreseen as early as 2005.
CHAPTER 6
The 2008 Global Mortgage and Derivatives Crisis
In the summer of 2007, the unemployment rate was at a five-year low,1 both the stock market and overall household wealth were nearing all-time highs, consumer and business confidence were solidly positive, and banks—the institutions best positioned to know trends in credit—had such confidence in loan quality that they were setting aside only small reserves for future loan losses. Most economists, politicians, businesspeople, investors, and government officials were sanguine about the future.
Yet only a few weeks later, the largest financial crisis to hit the United States since the Great Depression rocked the foundations of the country, and soon the stock market fell nearly 50 percent, banks and corporations began to implode, and unemployment soon doubled to 10 percent.2 Though the summer of 2007 was calm, almost two years earlier, in late 2005, so many high-risk mortgage loans had already been made that a financial crisis was virtually certain. Almost $1 trillion in soon-to-be-problematic loans—many to borrowers with no jobs or income—had already been made in a lending industry with less than $2 trillion in capital.
The Global Crisis of 2008, which led to what is now called the Great Recession, was inevitable before it was obvious, although few had noticed. Few, that is, except for a small number of the savviest contrarian investors, who began making huge bets in 2006 on what they were certain was an extraordinary and epic problem.3
The 2008 economic crisis in the United States was the direct result of the runaway growth in private debt from 2001 to 2007, especially growth in mortgages. Mortgages doubled from $5.3 trillion to an astonishing $10.6 trillion in a mere six years.4 Subprime loans—the general designation for loans approved under loosened credit criteria—accounted for $1.7 trillion of these loans.5 The bulk of this subprime lending happened in a short four-year span between 2002 and 2006. But, in sheer volume, the bigger problem was the group of ostensibly better loans, the “prime” loans, where loan terms and credit standards had also been compromised. Indeed, “cash-out” refinancings for prime borrowers—where a new mortgage was made at a higher valuation and the owner pocketed the difference—were among the worst-performing loans of all. Far too many loans were made to improperly qualified buyers than should have been, and far more houses were built to take advantage of this lending spree than were needed.
Figure 6.1. United States: Private Debt as a Percentage of GDP, 1998–2015
From 2002 to 2007, private debt to GDP grew 18 percent and reached an absolute level of nearly 170 percent.
The other category of runaway debt growth was commercial real estate, which shot up by 90 percent to $3.4 trillion, and would cause less but still extensive damage. In addition, and typical of a boom, margin loans for stocks skyrocketed. They grew 120 percent in five years, helping power stocks to record highs.
Together, the story of this runaway mortgage and CRE debt, along with a stock market fueled in part by margin debt, is the story of the Global Crisis and Great Recession. It is the indelible pattern of rampant lending that yields prodigious excess and then widespread failure. And the story of the Great Recession is also the story of things that might have been foretold but were not. Economists, examiners, and policymakers at the Federal Reserve Bank missed predicting the crisis. They missed it because they largely ignored the most relevant and telling data required for predicting financial crises: trends in the nation’s loan volumes. The Federal Reserve’s model for economic forecasting gave little importance to loans as a factor. Fed thinking had long been that for every borrower there was an offsetting lender, and it did not properly consider that if the volume of ill-advised loans became too large, borrowers would not be able to repay those lenders, which could cripple the banking system.
Figure 6.2. A Century of Real Estate Debt as a Percentage of GDP, 1916–2016
Not only had residential mortgage loans doubled, but the overall ratio of private loans to GDP was increasing by as much as 4 percent per year, had increased by 18 percent in the previous five years, and now exceeded 167 percent of GDP. Thresholds of 15 percent to 20 percent for the increase and 150 percent for the overall ratio level signal a high likelihood of a financial crisis.
Soon enough, this runaway lending brought the Great Recession, which took its place alongside the Great Depression as bookends to a real-estate crisis era (Figure 6.2).
After a stock market crash in 2000 and the terrorist attack of September 11, 2001, the economy was looking for good news and welcomed the renewed mortgage loan growth of 2002. Some have attributed the lending boom to a decline in interest rates, and rates were certainly a part of the story. In early 2001, thirty-year mortgage rates were 7 percent and dropped by roughly one percentage point in the second half of 2002. They then decreased by another half point in mid-2003—even though they quickly went back up to between 6.5 percent and 7 percent by mid-2006.6 “Real” rates (rates minus inflation) declined slightly more.7 But there have been many interest rate declines not followed by lending accelerations—or by busts. Furthermore, rapid loan growth has also happened in moderate, stable interest-rate environments. This mortgage lending growth came more from the same ferocity of ambition that we have seen in every crisis.
Table 6.1. U.S. Crisis Matrix: 2000s
In the five years leading to 2007, U.S. private debt grew by 55 percent. Sectors with the greatest concentration of overlending were mortgage and commercial real estate, which together comprised 70 percent of the increase.
Figure 6.3. United States: Total Homes Sold, 2000–2007
Aggressive lenders dramatically increased the volume of mortgage loans in this period. Home sales jumped. This accelerated lending brought more buyers and higher home prices. House prices rose from an index of 117.2 in 2001 to 204.8 in 2006.8 Home sales were 6.4 million in 2002, picked up further to 7.7 million in 2003, and then rampaged to 8.5 million by 2005.9 The higher home prices went, the more incentive there was to build more and lend more. Loan growth itself was driving demand.
Through the key period of this drama, from 2002 to mid-2006, lenders increasingly dispensed with the task of checking on jobs or income, instead making more of the now-famous “no income, no job, no assets,” or NINJA, loans. This dramatically enlarged the universe that qualified for a loan and helped enable the huge lending increase. The subprime industry, led by a new, more aggressive breed of lenders, made loans packaged into $1.78 trillion in residential mortgage backed securities, which pushed up home sales and prices.10
While it was happening, this burst of lending brought enhanced economic growth. It created jobs and wealth, lifting home prices to dizzying—unsustainable—levels. Unnoticed by many, the national ratio of home price–to–income, which had been 3.5 to 1 in 2000, had reached a troubling 4.7 to 1 by 2005. Miami and Las Vegas, sites of the most aggressive lending, had ratios considerably higher than that.11
As in the Roaring Twenties, a CRE expansion followed close on
the heels of the residential mortgage expansion, as commercial builders watched the housing boom and escalated their activity accordingly. CRE loans grew by $1 trillion, or 42 percent, during the peak CRE lending growth period from 2004 to 2007, thus both starting and ending later than the housing boom.12 This emerging CRE lending problem was a third the size of the residential mortgage problem but still huge by any measure.
Like the mortgage lenders before them, CRE lenders began to loosen terms and reduce loan pricing. By 2007, 53 percent were interest-only loans; the allowable loan-to-value ratios had increased, with many made on a ten-year, nonrecourse basis; and underwriting covenants became less stringent.13 Traditional lenders in this space—banks, life insurance companies, and pension funds—found themselves competing with a growing commercial mortgage-backed security market for deals, which further pressured the industry toward leniency in loan terms.
Lenders enticed borrowers through artificially low initial interest rates that automatically ratcheted to much higher rates after one or two years. They also enticed borrowers by reducing down payments to little or nothing—and by 2006, the number of borrowers making zero down payment had increased markedly to 43 percent of all new loans.14 Practically, this allowed borrowers with good incomes but no savings and borrowers of limited means to get a mortgage. They would face much higher payments after one or two years, but homes were appreciating in value by 10 percent to 15 percent annually, so borrowers weren’t worried. They reasoned that in a worst-case scenario, if cash-squeezed, they could simply refinance and gain cash access to the additional equity or sell the house at a profit—or so the theory went.
New and newly aggressive mortgage companies spearheaded the explosion in mortgages. Bethany McLean and Joe Nocera characterize these men and women as “worlds apart from the local businessmen who ran the nation’s S&Ls and banks. They were hard-charging, entrepreneurial, and intensely ambitious. . . . Some of them may have genuinely cared about putting people in homes. All of them cared about getting rich.”15
Angelo Mozilo, the CEO of the mortgage lender Countrywide Financial, had a relentlessness drive to “be No. 1” that rocketed his company to the top of the business. The priority at Countrywide was to push “affordable products,” the euphemism for artificially low introductory rates and low or zero down payments. In fact, adjustable rate mortgages (ARMs) with artificially low introductory interest rates were 49 percent of its business in 2004, up from 18 percent in 2003, and Mozilo was advocating a policy of ending down payments entirely. Countrywide grew from $363 billion in originations in 2004 to $495 billion in 2005, the largest amount in the business.16
Lenders made these loans on more lenient terms because they were getting rich doing it. Until credit problems overwhelmed the lenders, they were engaged in the most lucrative game in town. They charged big fees, sold off (most) of the loans, and reported huge profits. Their employees were getting rich, too. Thousands of newly recruited loan originators and salespeople, who had been earning more conventional salaries, began making hundreds of thousands and even millions of dollars annually in commissions for selling new mortgages, often through telemarketing to the most vulnerable borrowers.
Like Charles Keating before him, Mozilo courted friends in Washington, in a program internally referred to as “FOAs”—“Friends of Angelo.” He extended preferential mortgages to elected officials, including the chair of the Senate Banking Committee, Christopher Dodd (D-CT), and the chair of the Senate Budget Committee, Kent Conrad (D-ND).17
Subprime loans grew from 7.51 percent of total mortgage loans in 2001 to 17.34 percent of total mortgage loans in 2006.18 By 2005, 30 percent of the mortgage loans made by New Century, another leading subprime mortgage lender, were interest only.19 They were among a number of lenders who created ARMs with interest rates that would adjust to a much higher rate in mid-2007, making that date one of particular interest and concern to those few who were paying attention. Stories abound of the unemployed and underemployed getting mortgages during this period, sometimes for more than one home. Home sales were so hot that in some markets frantic buyers would enter bids moments after listings appeared.
In postmortems of the 2008 crisis, much was made of the use of “credit scores,” especially an industry standard known as the FICO score. Generally, a mortgage borrower with a FICO score of 670 or above was referred to as a “prime” borrower and those with lower scores as “subprime.” Many lenders mistakenly interpreted the FICO score as an absolute assessment of borrower risk. It was not and is not. Actual risk varies based on a number of factors.
As just one example, mortgage borrowers with a FICO score of 700 who make no down payment differ in risk from those with a score of 700 who make a 30 percent down payment, a difference not noticed or understood by most lenders. If a high down payment is required and only a low debt-to-income ratio is tolerated, then the lender can feel comfortable that it has two sources of repayment: first, the borrower’s income should be adequate to make monthly payments; second, the sale of the house will more than cover the amount of the loan. The FICO score further reassures that this borrower has been reliable in the past.
However, if no down payment is required, then the sale of the house might actually result in a loss since there are significant costs to administer the foreclosure and sale, and if no documentation of job or income is required, then the lender has no reassurance that the borrower’s income will allow for repayment either. A further complication was that younger, newer borrowers with scant activity recorded in the credit bureau files—the “thin files”—were assigned a fairly high score to start, which then changed apace with their actual credit behavior. The initial thin-file score therefore gave false assurance of good credit behavior. For these reasons, any number of loans labeled “prime” could instead reasonably be characterized riskwise as less than prime. By relying so blindly on credit scores, the mortgage market had made a fundamental mistake—a mistake that eventually would lead to billions of dollars in losses.
Traditional banks and S&Ls get funds by getting customers to open checking accounts and buy CDs. They can increase funding by raising the rates they are willing to pay depositors. Smaller depositors don’t have to worry about the risk those banks or S&Ls are taking because they’re protected by FDIC insurance. In 2008 deposits were insured up to $100,000.20 However, mortgage lenders such as Countrywide that led this lending boom needed different sources of funds. And that source was largely mortgage-backed securities (MBSs), a technique pioneered in the 1980s.21 Through this process, known as securitization, mortgage banks bundled mortgages into marketable securities and sold them to investors, such as life insurance companies, pension funds, mutual funds, hedge funds, and banks.
The key sources for purchasing the securitizations of these mortgage banks were the Federal National Mortgage Association (FNMA, or Fannie Mae) and the Federal Home Loan Mortgage Corporation (FHLMC, or Freddie Mac), which together accounted for 55 percent of activity in this area.22 Both institutions were quasigovernmental, authorized by government but not explicitly guaranteed by government, though many assumed that they implicitly were. They were formed to provide institutional buyers for mortgages made by banks, savings institutions, and mortgage bankers. In turn, they were funded by issuing bonds backed by the mortgages they held. The mortgage market became heavily reliant on these institutions. In 2004 alone, they bought $169.4 billion in subprime bonds.23
Investment banks, such as Goldman Sachs and Morgan Stanley, were also a source of securitization for mortgage banks. Fannie Mae, Freddie Mac, or the investment banks would take a group of mortgages originated by Countrywide, bundle them into a bond, and either sell them to investors or hold them on their own balance sheets. Fully 80 percent of mortgage loans made in the era were packaged and sold to Fannie, Freddie, Wall Street firms, and other large institutions rather than kept on the originating lenders’ books. An astonishing $3.6 trillion in MBSs were sold from 2002 to 2007.24
Even tho
ugh lenders were using loose credit criteria, investors bought the MBSs backed by those mortgages, largely because debt rating agencies—primarily Moody’s and Standard & Poor’s—assigned most of them their highest ratings—Aaa for Moody’s and AAA for Standard & Poor’s. These two private companies were filled with credit analysts whose business it was to evaluate debt securities such as these securitizations and assign ratings from the highest down to their default and near-default ratings of C and D.
Although these debt securities included some portion of substandard mortgages, they got the highest ratings because Fannie, Freddie, and investment banks structured them in a way that rating agencies believed mitigated or solved any substandard risk problem. In a growing number of cases, this was accomplished by slicing the securities into a stack of “tranches,” or portions, with the top tranche being the most senior, with a priority claim on the mortgage collateral in the event of default.25 The last or most junior tranche had the last claim on the collateral. In other words, the last tranche took the brunt of any credit problems, and the top tranche took the least. The tranches were all rated and priced accordingly and then packaged and sold separately.
This meant that the bottom tranches in the bonds, referred to as “mezzanine” tranches, would have terrible ratings and therefore could not be sold—right? Wrong. Many of them received higher ratings, too, because of a clever trick. Bankers would pool a number of these mezzanine tranches into a new security called a collateralized debt obligation (CDO).26 They argued that grouping different mortgages from different originators and different regions provided sufficient additional diversification to merit high ratings. There were even more obtuse CDO variations for securities with even worse credit.
A Brief History of Doom Page 19