A Brief History of Doom

by Richard Vague

  I would annotate Minsky’s thoughts in the following way: lenders don’t have to believe that they are making this speculative Ponzi type of loan for it to end up as one. They may truly believe they are making a hedge loan if, caught in the euphoria, their projections for a borrower’s future cash flow were too optimistic or if competitive circumstances changed adversely after the loan was made.

  To test, discover, and develop this thesis, my team and I collected comprehensive private debt data for the several years before and after each financial crisis listed above. This was a very different approach, not only because private debt gets overlooked but also because historians, especially those of older crises, have tended to minimize or gloss over the several years leading up to each financial crisis. They focus on the dramatic facets, such as bank failures, stock market declines, commodity price collapses, and the agony and ruin that follows, not the preceding years of accelerated lending, new bank formation, skyrocketing prices, and pervasive euphoria.

  Records on rapidly escalating private debt have especially been neglected.

  One goal of this book is to remedy that, though there are unfortunately still significant gaps in the data for many crises. I have made these assembled data for each crisis, along with other schedules and analysis covered in this book, fully available at www.bankingcrisis.org.5 Though my team and I have read hundreds of books and articles, hoping to find new sources or snippets of data, much data remains elusive, especially data related to private debt. Records on private debt have not been well kept, in large part because economists and historians did not deem the subject of high importance. In some cases these records may never be found or reconstructed.

  The United States sets the standard globally in data collection, though it still has important omissions in sector data and derivatives. Japan has remarkable data from the Meiji restoration forward, although it lacks key sector data for the post–World War II period. The United Kingdom, Germany, and France did not keep records in the 1800s as well as one might have expected. I would also note that I have a lower level of confidence in all aggregate data and statistics, such as GDP, prior to 1945, based not only on incomplete data but also the reliability of the extant data.

  All in all, it was a research adventure. These data are available to anyone who wishes to download, view, add to, or challenge them. I have a comment section where we invite and welcome new data submissions or corrections to the assembled data. We hope these data sets continue to improve through time. I encourage readers to study and pursue their own conclusions.

  In each chapter or introduction to a new financial crisis, you’ll see a chart that I call a “crisis matrix.” It distills into one visual display the dividends of my team’s efforts to find data on private debt for a given country and crisis. It contains data selected from the much more comprehensive set of data found on our website. The crisis matrix chart in each instance shows the growth in private debt, government debt, GDP, and select other data during the most relevant five-year period before a crisis, along with that same data in key years after the crisis. Where a crisis matrix for a given country is not shown in the chapter, in most cases I have provided the broader set of data on the website.

  At the bottom of each matrix, I also express these data as a percentage of GDP, to help with comparisons through time within the same crisis, as well as comparisons between crises. For periods in which we have no private debt data, we have included other key economic data, such as construction spending or capital formation, which serve as a proxy for private debt data. All numbers in the charts and tables throughout this book are nominal unless indicated otherwise.

  The following chapters tell the story of financial crises, first through the major financial crises for which we have the most, and most reliable, data. We start with what many economists think of as the important crisis and the one for which they deem to have the most elusive explanation—the Great Depression. This chapter excavates data on housing, commercial real estate, and other sectors to tell a surprising new story about this iconic financial crisis: at its heart, it was a real estate crisis brought on by rampant private lending and debt. We then move to the two major crises that followed, one in the 1980s United States and the other in 1990s Japan, to show how the same pattern of financial crisis played out.

  After these first three chapters, the grim logic that underlies financial crises will be clear.

  A reader might wonder, however, if the patterns of financial crises as private debt and real estate crises apply to earlier eras. Have we discovered a plot of financial crisis since the industrial era began, in other words, or only a plot of relatively recent crises?

  To settle the matter, I move in the next two chapters back in time to test our hypothesis further on the frequent, yet quantitatively murkier, global financial crises of the 1800s. We find the very same pattern. In Chapter 4, I take a look at the crises of 1819, 1825, and 1837, while Chapter 5 looks cumulatively at the many crises of the “railroad era,” from roughly the 1840s to the early 1900s. I argue in these chapters that the crises confirm our core findings: while commercial real estate and home mortgages could, and did, bring down the U.S. economy in the 1930s and the late 1980s and Japan’s economy in the 1990s, in the 1800s, it was the behemoth railroad sector of the economy along with associated construction that was large enough to inflict that kind of damage. The proliferation of railroads—forged by private debt—is the key to this extraordinary period in U.S. and global economic history.

  With fresh eyes, and this narrative in place, we return to the recent past in our last chapter, on the 2008 crisis. We establish that the crisis of 2008 was inevitable as early as 2005. In fact, it should readily have been predicted.

  One final disclaimer: Because we spend so much time in this book on the pernicious effects of private debt, with such debt almost serving as the economic villain of financial crisis, it is crucial to note that private debt is a necessary and positive element of an economy and is one of the fundamental ingredients of growth, trade, profit, and investment. Take away private debt, and commerce as we know it would slow to a crawl. The world suffered crisis after crisis in the 1800s, but per-capita GDP increased thirtyfold in that century, and private debt was integral to the growth. Many of those nineteenth-century crises were rooted in the overexpansion of railroads, but they left behind an impressively extensive network of rails from which countries still benefit. China’s unprecedented and extraordinary rise since 1980 has been largely built using private (nongovernment) debt. And though there has been calamity along the way, and more likely yet to come, the rapidity and breadth of the country’s rise would have been impossible without it.

  Private debt is indispensable but becomes a problem when it grows too fast or gets too large in relation to the economy. That is the paradox of debt. And as many studies have shown, if an economy’s private debt–to–GDP ratio is high but not rapidly growing, then that private debt dampens GDP growth.6

  Life goes on after a crisis. An economy that had been growing will sometimes simply resume its course, as was the case for the United States in the nineteenth century. Yet understanding the anatomy of financial crises is a critical obligation we have to the governments and political systems that are disrupted, and even more to those millions who do not cause the crisis but are badly damaged by it. Predicting financial crises should be our duty. My hope is that if we better understand financial crises by focusing on the neglected yet vital matter of lending and private debt, as the following chapters amply illustrate, then much of the damage that comes from them can be avoided.

  CHAPTER 1

  A Jazz Age Real Estate Crisis

  The Great Depression

  From the distance of more than eighty years and in the cold light of economic analysis, it is easy to forget the death and despair of the Great Depression. Historian Adam Hochschild describes a country that “simmered in misery,” with thirty-four million Americans living in households without a breadwinner. H
e writes of unemployed steelworkers living inside idled coke ovens with their families and rummaging through the garbage bins for food, of riots, and of men in cloth caps waiting outside churches and charities for quickly depleted food supplies.1

  The Great Depression brought a level of misery rarely seen in American history. Oceans of ink have been spilled to explain why the bust phase of the Great Depression, with the collapse of GDP, was so protracted, painful, and deep. For economists, explaining its cause has sometimes been referred to as the holy grail of macroeconomics. But largely absent from the most widely read books on the Depression is its one central cause: runaway private debt.

  Those who do mention private lending, such as Milton Friedman and Anna Jacobson Schwartz in their iconic Monetary History, misrepresent the phenomenon. Though Friedman and Schwartz had a number of pivotal insights about the collapse in the 1930s, they said little about the overlending of the 1920s. “If there was any deterioration at all in the [pre-Depression] quality of loans . . . it must have been minor,” they wrote, and “any [pre-Depression] deterioration in the quality of loans . . . was a minor factor in subsequent bank failures.”2 By dismissing private debt, Friedman and Schwartz didn’t just miss its contribution. More fundamentally, they missed the human factors involved in a crisis: the competitive intensity, the drive for wealth, the pervasive hubris, and the self-delusion.

  Figure 1.1. United States: Private Debt, 1920–1935

  In the United States, private debt increased 52.7 percent from 1920 to 1929.

  John Kenneth Galbraith’s Great Crash of 1929 glossed over the national real estate boom beyond Florida. In his final chapter on “causes,” he only briefly mentioned lending, noting mostly that in the late 1920s “money was tight,” even though loans were then growing at a robust 6 percent per year.3

  The misery that unfolded from 1929 to 1933 was a story of rampant and unsound lending followed by widespread loan contraction that contributed the lion’s share of the cataclysmic GDP contraction. The Great Depression was a massive residential and commercial real estate crisis. The financial records of the 1920s, which have largely been overlooked, indelibly show this. During the 1920s, annual housing and commercial real estate construction almost tripled—and nearly all of it was financed by debt (Figure 1.2).

  This explosion in residential and commercial construction lending, augmented by lending for utilities and stock purchases, created the euphoria of the Roaring Twenties, the jazz age of robust spending and celebration. Companies used the new money from loans to expand and employ more people.

  The acceleration in construction resulted in such extensive overbuilding that by the final years of the decade, before the stock market crash, thousands of newly erected office buildings, houses, and apartments sat empty. Office vacancy rates rose, and residential mortgage foreclosures nearly doubled in the final years of the decade.4 As in other cases, this crisis was inevitable before it was obvious. The only question, and the only area where the president and the Federal Reserve could still have a discretionary impact, was the length and severity of that correction.

  Figure 1.2. United States: Construction Spending, 1915–1940

  The Great Depression, like most financial crises, can be thought of as occurring in three distinct phases. The first, from 1923 to 1928, was the runaway lending boom that led to the avalanche of terrible loans that brought so many banks to failure. The second—the great contraction—lasted from 1929 to 1933, when GDP fell by an astonishing $48 billion, or 46 percent, and brought 30 percent unemployment. The third phase began in 1934, when the economy ceased contracting and slowly struggled forward. (There was a notable new recession outside of the scope of this book that began in late 1937.)

  The state of the nation’s housing had become a national issue in 1921, when Secretary of Commerce Herbert Hoover began to advocate for increased home ownership as “the foundation of a sound economy and social system.” In 1923 he wrote, “Maintaining a high percentage of individual home owners is one of the searching tests that now challenge the people of the United States.”5 This argument—that the homeowner was a more committed and responsible citizen—would recur in the 2000s. Hoover urged homebuilders to become more efficient and lenders to become more generous.

  New tax incentives supported home construction. Historian Donald Miller writes that no legislation boosted New York City’s construction industry more than the 1921 exemption from real estate taxes for ten years of residential construction. “In the 1920s,” he writes, “New York City would account for fully 20 percent of all new residential construction in the country. Builders of large apartment houses were the pacesetters. In 1926, 77 percent of all new residential construction was given over to apartment dwellings.”6

  Banks, building and loans, bond houses, and other lenders responded to Hoover’s call. As Robert M. Fogelson reports, Albert E. Kleinert, Brooklyn’s superintendent of buildings, reflected the frenzied pace of the era when he said, “They [the speculators] are now selling property between twelve and one o’clock and then at two o’clock they notify their tenants that the rent will be raised, and then when they show an income gain on paper, they sell again. More often than not, the new owners repeated the process.”7

  In a financial crisis, it is typical for lenders other than conventional commercial banks (those that are less regulated or unregulated) to play a leading role in the preceding lending upturn, and so it was here. The number of building-and-loan associations—savings institutions whose primary purpose was home lending—grew from 8,000 to 13,000 during the 1920s.8 Builders and developers chartered many of them to finance their own projects in a brazen conflict of interest, especially by using low down-payment, long-maturity loans. As one example, to help fund their expanding real estate business, two Florida banker-developers succeeded through political connections in getting bank regulators to create sixty-one new national and state banks in which they took a stake. Others gained charters by extending loans to regulators.9

  In addition to bank loans, real estate bonds were a major source of building finance. Many were sold to the general public, some in denominations as low as a hundred dollars. Billions of dollars in bond funding were raised—for apartments, hotels, office buildings, and more. Unlike bank loans, institutions that sold these bonds did not have risk of loss if the underlying real estate didn’t perform; they therefore embraced looser credit standards.

  Much of this bond activity was unaffiliated with banks, but a number of banks sidestepped limitations on underwriting and trading in bonds by launching securities affiliates, which grew in number from 10 in 1922 to 114 by 1931.10 On a per-capita basis, the U.S. housing boom of the 1920s was every bit as large the housing boom of the 2000s.

  Overbuilding in commercial real estate, while smaller in dollars than housing, was still vast and more visible than overbuilding in houses. Between 1925 and 1931, office space increased by 92 percent in Manhattan, 96 percent in San Diego, 89 percent in Minneapolis, and 74 percent in Chicago.11 In New York alone, approximately 235 new buildings were constructed in this critical period, almost all debt financed. As Daniel Okrent maintains—and as our data indicate—more “buildings taller than 70 meters were constructed in New York between 1922 and 1931 than in any other ten-year period before or since.”12 Historian Robert Fitch estimates that between 1921 and 1929, “developers had added 30 million square feet of office space to the Manhattan inventory—an amount and a rate of increase that approaches the eighties office expansion. . . . They kept increasing the flood of overcapacity.”13

  Table 1.1. U.S. Crisis Matrix: Portrait of the Great Depression

  In the five years leading to 1928, private debt grew by $40 billion or 34 percent, far greater than the GDP growth of $12 billion. Sectors with the greatest concentration of overlending were household mortgages, commercial real estate, utility debt, and broker loans. Together they comprised 68 percent of the increase during this period. Government debt actually declined.

  Figu
re 1.3. New York Skyscrapers (Taller than 70 Meters) Built, 1910–2000

  The iconic structures of American skylines form the silhouette of the Great Depression: New York’s Chrysler Building, Empire State Building, and RCA Building; Chicago’s Merchandise Mart, Wrigley Building, and Tribune Tower; Philadelphia’s PSFS Building; Los Angeles’s City Hall; Dallas’s Cotton Exchange Building; Detroit’s Fischer Building; and Houston’s Gulf Building. These are enduring architectural feats of the 1920s, vestiges of the real estate eruption that came before the fall. Many were speculative projects, unsupported by actual real estate demand; begun toward the end of the 1920s, when loans were still available; and finished after the crash, when lenders had little choice but to make funds available to complete construction or else see their entire loan go bad. None was financially successful for its original investors. They remained partly or largely empty for a decade or more after completion, as would hundreds of others.

  Lenders were caught in the euphoria, too. Real estate was built when—and because—a loan was available to finance it, even in the absence of hard-nosed analysis of demand. The construction boom meant more job creation and employment, but since so much construction was based on the availability of financing rather than actual underlying need, more workers were hired than dictated by organic demand. Florida had a famous and fraud-riddled real estate surge that came to an ugly end in the hurricane of 1926, but its equivalent played out in any number of major U.S. cities, especially New York and Chicago.14

  The lending boom of the 1920s also brought significant growth in business loans for the acquisition of utility companies by larger utility companies. These came to be known as “utility pyramids.”