It’s been almost exactly seven years since a sudden and mysterious stampede among financial institutions threated a second Great Depression. The Federal Reserve Board, backed by Congress and joined by other central banks around the world, halted the systemic run, with banks demanding cash from one another, with an equally startling burst of emergency lending to all who might be threatened by withdrawals. Four desperate weeks in the autumn of 2008 have been followed by years of pain and caution.
The ensuing slump was bad enough – peak unemployment of 10 percent, 5.6 million foreclosures over three years. But had the Fed not decisively intervened, former Federal Reserve Chairman Ben Bernanke writes, “historical experience suggests the nation would have experienced an economic collapse far worse” – a lengthy period as grim and dangerous as the Great Depression.
Bernanke’s book, The Courage to Act: A Memoir of the Crisis and Its Aftermath (Norton), appeared last week. It has three parts: an autobiographical prelude; a dozen riveting chapters on the crisis; and a discussion of the Fed’s role in the aftermath, mostly about quantitative easing. I skimmed the first and third parts, but read every word of the second part, 278 pages of a 580-page book.
With those in mind who have been following the serial here since May, I found the relevant portion of Bernanke’s book on pp 398-410, three sections of a chapter in which he describes what happened in those few weeks. He’s summing-up the climax of events that began in June 2007 and quietly ended after the stress tests were administered in early 2009. He writes,
For me, as a student of monetary and banking history, the crisis of 2007-09 was best understood as a descendant of the classic financial panics of the nineteenth and early twentieth centuries. Of course the recent crisis emerged in a global financial system that had become much more complex and integrated, and our regulatory system, for the most part, had not kept up with the changes. That made analogies to history harder to discern and effective responses more difficult to devise. But understanding what was happening in the context of history was invaluable.
Unsurprisingly, Bernanke puts no new construction on events.
New insights will certainly emerge in coming years, just as Milton Friedman and Anna Schwartz, writing in the 1960s, fundamentally changed our understanding of the Great Depression. However, as we battled an extraordinarily complicated crisis, we didn’t have the luxury of waiting for the academic debates to play out. We needed a coherent framework to guide our responses.
The central bankers understood that their playbook had evolved from the guidebook that is Walter Bagehot’s 1873 classic, Lombard Street: A Description of the Money Market. So much has changed in financial markets in 150 years that details of the Victorian world that Bagehot describes are almost as unfathomable for the average reader in the present day as are the details of the alphabet soup of financial instruments with which Bernanke had to deal, but Bagehot’s point is still crystal clear. “[W]hat is necessary to stop a panic is to diffuse the impression, that though money may be dear, still money is to be had.”
Central bankers followed Bagehot’s rule – lend freely, at a penalty rate, against good collateral. The Fed added a few new wrinkles of its own, reducing stigma, lending across national borders,artfully employing the concept of stress tests (quantitative evaluations of vital statistics under worst-case scenarios) to restore confidence in the system. These measures stopped the panic. In the end, most loans were repaid, at a profit. And the oft-bruited specter of inflation never materialized.
The task here at EP, as in the past, has been to be attentive to those “academic debates” about the origins and nature of the crisis, to identify the most interesting ideas among them and trace out their implications for the future. This is the news business, after all, not history, though relating a certain amount of history over the summer seemed desirable, the better to understand what was new.
I zeroed in on a collaboration that began the famous Federal Reserve Bank research conference in Jackson Hole, Wyoming, in 2008 between two very different economists. Gary Gorton, of Yale University’s School of Management, is a historian of banking who, for a dozen years, spent a day a week as a consultant inside AIG-FP, the highly profitable insurance skunkworks that became the very center of the storm. Bengt Holmström, of the Massachusetts Institute of Technology, is an organization theorist who became interested in central banking during the Scandinavian banking crisis of 1991-92. Their fundamental paper, Ignorance, Debt, and Financial Crises, with Tri Vi Dang, of Columbia University, is not yet even a working paper (it has circulated since April 2009 in various seminar versions). Yet in successive iterations and in related papers, the pair has spelled out with increasing precision the economics of the long “hypothecation chains” that were the signal characteristic of a collateral-backed banking system that ballooned after 1983, in response to the demands of globalization, and which threatened to collapse in 2008. The panic, they say, was an information event, originating in the repo market, triggered by recognition of the implications of the recession already in progress.
In the text of his presidential address to the Econometric Society, Understanding the Role of Debt in the Financial System, Holmström, put it this way:
Panics happen when information-insensitive debt (or banks) turns into information-sensitive debt…. A regime shift occurs from a state where no one feels the need to ask detailed questions, to a state where there is enough uncertainty that some of the investors begin to ask questions about the underlying collateral and others get concerned about the possibility [of default] …. This can lead to reduced liquidity and rapid drops in prices.
The idea that debt is fundamentally different from equity is not new. In comments (pp. 35-41) on Holmström at the research conference of the Bank for International Settlements in 2014, Ernst-Ludwig von Thadden, of the University of Mannheim, dated its formal recognition to a remarkable 1979 paper by Robert Townsend, then of Carnegie Mellon University, subsequently the University of Chicago, today of MIT, “Optimal Contracts and Competitive Markets with Costly State Verification.”
I quote Thadden’s genealogy here because of the glimpse it affords of how modern economics proceeds: how contract theory, seemingly hopelessly abstract, was advanced slowly, trench by trench, by hard-working economists, usually all but unknown to the public, for more than a quarter of a century, until suddenly in 2009, its elaborations provided a foundation to explanation of why the “debt on debt” insurance strategies of collateral banking might have arisen to minimize information asymmetry or information acquisition problems.
They are optimal precisely because they minimize informational requirements along the contracting chain. Therefore debt is opaque for as long as “the music is playing,” as the famous saying by Charles Prince, the CEO of Citigroup, went in 2007.
This argument that debt is the optimal financial security in response to informational problems between borrowers and lenders has been made in several other contexts: debt economizes on information collection costs at the contract execution stage (Townsend [(1979]), debt economizes on liquidation costs of collateral (Hart and Moore (1998)), debt mitigates managerial moral hazard (Innes [1990]), debt is an optimal response to private information at the contracting stage (Nachman and Noe [1994]), debt is least information-sensitive to ex post public information (DeMarzo et al. [2005]) and debt can optimize information collection if lenders screen borrowers at the contracting stage (Inderst and Müller [2006]). All these are, or can be cast in terms of, information-based theories of debt. This list is not exhaustive, and [Dang, Gorton, and Holmström] add an important new element to the list, which is particularly relevant for the recent experience of shadow banking. In the present paper, Holmström builds on the whole list of information-based theories of debt, to emphasize that the informational advantage of secured debt, very generally, comes at a necessary cost: at times when returns are bad, markets that are not well equipped to deal with information processing get overloaded and can become dysfunctional.
Why does it matter? Think back to what you think you know about the crisis: subprime lending, liar loans, the originate-and-distribute model of mortgage lending, securitization, leading to financial instruments so complex and opaque that not even those trading them understood how they were made. What if all of it was more or less intentionally designed to be that way? What if the hastily-improvised system of collateral-backed banking, assembled to soak up the accumulated savings of rapidly growing nations, pension funds, corporations and private savers, actually worked pretty well, until it didn’t?
Money itself is optimally opaque, says Holmström. If the hallmark of money market liquidity really is a state of “no questions asked,” as Holmström puts it, there are good reasons for the lack of transparency. Nobody knows what “the full faith and credit of the government” actually means, but all goes smoothly with those $100 bills because ignorance is symmetrical – until fear of counterfeiting reaches a certain level.
With the crisis of 2007-08, economics acquired a new timeline, quite different from the one with which it has operated for the last 75 years – that is, within the memory of almost anyone living. The Depression that didn’t begin in 2007 joined the one that did begin in August 1929 and which lasted until March 1933. The two events now dominate a list of panics including 1907, the one that brought the Federal Reserve Board into existence, extending back in time to the Mississippi and South Sea bubbles of 1720 and before.
The year in which John Maynard Keynes published The General Theory of Employment, Interest, and Money, 1936, for a while longer will remain year zero for many economists; there is a substantial installed base for whom money and finance are the concerns of somebody else. In a meeting with leading macroeconomist last spring, Bernanke assured them that their human capital was intact, and of course it is. It’s their intellectual capital that is depreciating, rapidly.
Marveling one day at the rough symmetry of the two events – the slowly escalating banking panic of the early 1930s and the suddenly explosive panic of 2007-08 – Yale’s Gorton said quietly some years ago, ”This time there was no Keynes.” Even then the sentiment had become something of a mantra among technical economists seeking to understand the crisis away from the clamor of everyday debate. Economics has grown deeper, richer, more comprehensive and more specialized since Keynes, Hayek, Schumpeter and Irving Fisher competed for the attention of presidents and prime ministers. It has developed very different systems for writing its history. Review articles, invited lectures, anniversaries, and prizes all play important roles. It is true that this time there was no Keynes. This doesn’t mean that no breakthrough occurred. The recognition of the role of purposeful opacity in the design of debt is one of those events that should convince all but the most skeptical that economics does in in fact progress.
. End of serial! Back to business! dw