The Fed Is Actually Bailing Out Itself
In December 2006, a small, relatively unknown financial firm called First Marblehead completed a $1 billion student loan securitization. At the close of the deal the company expected to receive up-front "structural advisory" fees of about $89.6 million, or 12.4% of the total balance. In addition, First Marblehead modeled and discounted additional advisory fees of $8.8 million and "residual revenue" of $48.7 million. The "blended" yield on the transaction was more than 20% of the total balance of expected student loans (the loans had yet to be acquired).
Because of the gain-on-sale accounting rules as they pertained to securitizations, First Marblehead booked those massive earnings of "expected" cash flows at the deal's inception, right up front. The company was a small operator in the overall picture of loan securitization, but it represents a very good, unfiltered window into the world of modern high finance. What they were doing in student loans, the big banks were doing in mortgages on a much greater scale.
It is easy to see why these deals were so attractive to financial firms. They would book so much revenue and income up front, meaning the larger the deals, the sweeter the internal assumptions in their discounting models, the more profits would flow to the firm up front. Given the credit system's evolution into one based on the accounting notion of equity capital, these booked profits flowed directly into the firm's retained earnings, and therefore greatly increased their ability to create even more credit (and phantom, expected profits). It was a self-feeding, narrowly construed cycle.
None of the risks of these activities were incorporated into the financial statements of financial firms, since the assets had been "sold" to the private investors in the securitized tranches. Instead, from time to time, any "minor" changes in the expectations of cash flows were recorded on the income statement. For example, if First Marblehead saw an increase in default rates or a decrease in expected recovery rates, it would have to adjust those 2006 revenue assumptions, updating its estimate of expected future cash flows, booking only those expected changes as an incremental loss on the income statement.
However, the converse was also true. If default rates fell or recovery rates rose, then the company would actually book an additional phantom gain in income - while patiently waiting to actually collect the cash over the course of the securitization's term. This fine-tuning through earnings was expected over the life of the deal, and any volatility to those expectations was dismissed as trivial (recency bias at work).
None of this was even possible without a few additional assumptions in the form of mathematical modeling. The allure of securitizations was apparent in the 1990's, but there was a practical limitation to their use. There was no established marketplace to offer consistent and continuous pricing of all the necessary financial variables. The most important element of pricing tranches is estimating default correlation among the loans included in the structure. Since historical data that correlated mortgage loan defaults among financial terms did not exist, securitizations remained a boutique part of the overall credit business.
That changed in 2000 with David X. Li's publication of a paper where he essentially extracted correlation from market-traded credit default swaps. His now-infamous (through no fault of his own) Gaussian copula model, to oversimplify here, compared the credit spreads (basically credit default swap pricing) and curves of similarly characterized, and already traded, credit products. If they looked the same, the Gaussian copula inferred a measure of correlation.
Since credit default swap trading had been growing in substance and depth since their first use in the 1980's, this market-based correlation assumption fit nicely into the expectations of structured credit investors and traders, offering real-time pricing of the previously illiquid asset class. The proliferation of securitization came directly from this complex math, which itself was nothing more than a shortcut of circular logic (the market needs estimates of future default correlation to price assets, so the assets get their correlation estimates from the market).
Among the largest assumptions incorporated into the more than ten trillion dollars in credit advanced in the 2000s through these means, was that expectations of future pricing characteristics would follow along the same path as recent history. That, of course, was the primary problem with mortgage bonds (and mathematical models in general) once conditions changed in 2006 and 2007 as default estimates began rising, and recovery assumptions proved entirely too rosy. But more than anything, the feedback loop of the credit crisis was maintained because of this shortcut to extracting vital financial characteristics from the marketplace.
Every firm that used the Gaussian copula, and every investment bank did at the core of their structured finance models, assumed that the market would rationally assert likely probabilities of defaults and recoveries. That is what credit default swap prices were supposed to be, rational expectations of what these variables will look like at some point in the not-too-distant future.
Once the far-too-optimistic assumptions of the entire structured finance industry were challenged by the explicit realities of the housing market as it transitioned into 2007, the trickle of negative securitization revenue adjustments began in earnest, along with a reset in perceptions of the riskiness of those illiquid assets. Modern finance being what it is, completely captured by the notions of hedging only static and mathematical perceptions of risk, saw an increase in demand for ways to offset risks that were now clearly rising. Since structured finance had grown so large, abetted by regulators and central authorities basking in the genius of their "managed" and "stimulated" economy, that meant a huge increase in demand for credit default swaps as hedges.
A rush of demand for default swaps pushed many idiosyncratic instruments in the same direction at the same time. Since the Gaussian copulas interpreted similar moves as correlation, this meant that the mathematical indication of correlation "measurements" rose with these new fears. And of course, as interpretations of rising correlations made their way into the math of pricing models, tranche pricing became even more problematic. That forced incrementally more demand for credit default swaps, feeding back into estimates of correlations rising even higher, further heightening fear and the need to hedge, and so on.
By the time Bear Stearns had failed in March 2008, correlation estimates had gotten out of hand. Because of some quirks of financial products and their mathematical modeling, characteristics like negative convexity and correlation smiles, it was not uncommon for traders to quote various mortgage bond tranches in correlations of greater than 100%. Of course that makes no logical sense within the confines of what correlation is supposed to confer or what mathematics actually defines, but the market realities of the period introduced by David Li's shortcut undercut the ability of the marketplace to make sense of itself. The math just could not handle the emotion.
What was really going on was nothing more than supply and demand driven by fear. Demand for credit default swaps continued to grow while supply was shrinking as financial firms retreated from writing them (the monoline insurers, a large segment of default swap issuers, were in serious trouble at the time, in fact only AIG really continued to write swaps in volume). When demand far outstrips supply, prices rise uniformly fast, further reducing the values of the very instruments banks were trying to hedge in the first place. In a morphed sense of Fisher's paradox, the more banks tried to hedge and offset their risks, the greater the risks became. The impartial mathematics of the day locked the system into this loop of destruction.
It may be easy for many to conclude that we have learned the lessons of this sorry episode in financial and economic history, the proper authorities have swooped in to fix the flaws, and all is well. Regulators and politicians do not understand the nature of the crisis they faced, nor do they see that it is the modern system itself, down to its mathematical core, that is the problem. The failure of MF Global in November maddeningly followed this same exact path, right down to the gain-on-sale accounting and incremental charges to net income as fear in the marketplace forced a change in assumptions. The banking system is not currently constituted to function outside of a narrow window of assumptions.
So far, all the monetary solutions offered have been to change the world to fit into that narrow window - rather than change the banking system to reflect the world. The Federal Reserve, in particular, offers endless money to keep the banking system afloat, a life raft to the expected better days ahead. To ensure those better days actually appear, central banks all over the world alter the interest rate structure, directly hurting the responsible savers and investors that might actually lead a real recovery, while further manipulating prices everywhere and causing widespread consternation and uncertainty.
In a big picture sense, what is the Fed (and its central banking cousins) trying to save? Is the credit situation in the world so advanced (perverted) that we need the mathematical shortcuts like the Gaussian copula just to make sure enough individuals have access to mortgage markets?
The conventional wisdom of modern economics, and the cursory understanding of it all by the public, sees Wall Street, its Gaussian copula and the economy as one inseparable whole. Rising credit equals rising economic activity, so the advancement of the banking system necessarily and uniformly leads to advancement in the real economy. This is a pervasive belief that is accepted in too many places without critical questioning, especially in the political arena.
As I (and many others) have said numerous times, it is a deliberate prevarication. The Fed and central banks around the world coordinate dollar swap lines to save the banking system from its umpteenth moment of illiquidity simply because the banking system, through credit creation, equals control over the economies those central banks are supposed to serve.
This does not mean that a bunch of evil geniuses conspire in the wood paneled offices of the big banks to sink the world into despair for their own gain. Rather, I think there is enough evidence and established history to form the narrative that today's central bankers see themselves as the evolutionary equivalent of Plato's philosopher kings - the benevolent rulers of a just and ordered society, where the elites make the economic rules for the rubes to harmoniously follow.
Wall Street is the primary tool, the hammer, of economic control for our own good. Saving Wall Street (and its global brethren) is simply the struggle to preserve control over the levers of the economy. But, as we have seen so many times in the past few years, the "tool" of choice for the ordered, harmonious economy of the modern economic utopia is not well suited for any real economy. To borrow a phrase from Thomas Jefferson, Wall Street, for the Fed, is like holding a wolf by the ears - they can neither sufficiently hold it nor safely let it go.
The Fed, the economics profession and the financial media spread the idea that this unfettered credit creation paradigm is part and parcel to the basic economic philosophy of capitalism. It is not. Capitalism represents the free expressions of a free society, so leeching onto it achieves another shortcut to allow free people to accept a degree of economic central control.
The usefulness of First Marblehead and the short cut of the Gaussian copula is how they illuminate the stark difference between Wall Street and capitalism. Central banks cannot fairly refer to themselves advocates of capitalism when they so narrowly construe their largesse. The central control of modern economics seeks to control credit independent of actual demand; indeed, it seeks to create demand from nothing.
If a housing bubble achieves the philosophical aims of "stimulating" the economy to some predetermined target or range, then the political aims of the central bank are fulfilled no matter how shortsighted that may be. This, of course, requires the less-than-ideal means to achieve these ends, such as gain-on-sale accounting to attract enough capabilities and capacity to the systemic effort. True intermediation really does not need to be so complicated.
The detachment of credit money from actual money demand to engage in productive transactions is both the glaring difference between capitalism and monetarism, and the ultimate weakness of superimposing the latter on the former. Over-stimulated demand for money, as mass produced securitizations amply showed in this instance, has always led to asset bubbles. But today, conventional wisdom holds that those previous, historical instances of asset bubbles were emotional frenzies of free markets, now cured by rational monetary science (as long as we disregard the credit element of the last few credit bubbles).
The fact that monetary economics today refers to itself as a science means that it views itself as an unimpassioned, rational arbiter of all things economic. Because science is closely tied, in the human consciousness, to evolution and progress, it is, regrettably, unquestioned as a superior option to the passion, desires and emotion of truly free markets. Free markets, this conventional wisdom goes, are inherently messy, unstable and dangerous.
The precision of modern math combined with this rational and supposedly impartial monetary philosophy takes the emotional mess out of the equation, leading to a golden age of economic satisfaction - now inappropriately named the "Great Moderation". As much as central banks want to forgive and forget, there is no separation between the Great Moderation and the Great Recession. The two are intricately linked through causation. To keep the illusion of the Great Moderation going necessarily required First Marblehead and the Gaussian copula, which necessarily led to the Panic of 2008.
It all worked so well, until it didn't. The Gaussian copula was nothing more than an elegant, but inadequate shortcut that was all too easily drowned in the tide of fear that deficient math created in the first place. The First Marbleheads of the world, more commonly embodied within the financial giants of Wall Street and Europe, meant that intermediation was not as "clean" and rational as the central authorities actually believed or wanted us to know. It was not the Platonic utopia of their dreams (Alan Greenspan was once believed to have "conquered" the business cycle), it was a house of cards, a monetaristic façade of the once free market capitalist system that has rotted from neglect.
As the façade plummets to earth in the messy aftermath of what it, not capitalism, has wrought, the central authorities cling desperately to their system. It matters little if bailing out the eurodollar market for the fifth time actually advances the real economy. All that matters is that the tools for maintaining the elitist utopia are preserved for future use. They just want us to accept that they know better, having already crowned themselves Lords of the global economy.