Anniversary of a Market Event That Changed Everything
Tomorrow is an auspicious anniversary largely unknown to even confident observers of the panic period, with events of that date changing the very character of the financial system irredeemably. To dryly observe what happened understates the severity of the event, as London trading for "dollars" saw an inordinate discrepancy that cannot be fully comprehended by mere factual recollection. On August 9, 2007, 3-month LIBOR rose from 5.38% to 5.50%; a trivial number at first glance. But in the context of global dollar flow, it was an immense tremor that kicked off the spate of "emergency" monetary measures that ultimately followed.
Prior to that point, 3-month LIBOR traded in such a narrow range that any movement greater than a single basis point was counted universally as abnormal. Indeed, from May 10 of that year all the way through July 26 the rate was pegged at exactly 5.3600%. To most, that was very reassuring since the initial eruptions of subprime and Bernanke-drained promises ("contained") were up for debate. In June 2007, two Bear Stearns "high-grade" structured credit hedge funds had to receive a "bailout" from their sponsor, Bear itself. The implications of the bailout were not just pricing, but liquidity - in other words there were no "dollars" to be had for the funds to continue to run its positions.
By July 17, Bear had informed investors in one fund that its value had been pared 90%, while the other had virtually no "value" left at all. Two weeks later, on July 31, the funds petitioned for Chapter 11 to be unwound in an "orderly" fashion. But as long as broader indications of "normalcy" were readily apparent, market fears could remain in the shadows, at least for those not closely observing the full scene (including every regulator and policymaker).
Something changed heading into August, however, and the system never recovered. From July 27 until August 7, 3-month LIBOR actually "traded" a few pips below that 5.3600% level in what seemed like stoic calm in funding markets. On August 8, suddenly the rate jumped two basis points, again, in what can only be described as utter incongruence of appearance to function, as it was that day becoming more apparent that trouble lurked beyond just a few hedge funds. The next day, August 9, the eurodollar market essentially "froze", rising a further 12 bps and upsetting funding markets downstream and globally.
There had been an extremely durable relationship between LIBOR, "dollars" on trade in the eurodollar market in London, and federal funds, dollars on trade domestically. In fact, there was a very close connection between the effective federal funds rate and LIBOR to the point that US monetary policy fairly covered, as much as it intended anyway, the global dollar market. That was always disavowed, but the careful relation of LIBOR and federal funds contradicted that policy stance of a "closed system." Initially, the effective federal funds rate jumped a similar amount, by 14 bps, on August 9 to 5.41% from its near-target the previous day of 5.27% (FOMC policy had targeted 5.25% at that time), but it quickly settled significantly under the target for the rest of the month and all the way into September. In fact, federal funds would not return up to its policy target until two days before the Fed's first "emergency" 50 bps rate cut on September 18, 2007.
That meant that for more than a month, eurodollars were in very short supply, as LIBOR persisted above its previous structurally apparent relationship, at the same exact time federal funds were "overly" plentiful - and thus the geographic fragmentation that eventually led to panic was opened, never to be resolved by a policymaking framework incapable of understanding exactly that. For "dollars" to be at odds in such dramatic fashion on both sides of the Atlantic, ostensibly delinking banking institutions from themselves (their subsidiaries) was the end of the road for the financial system as it existed from the 1980's to that point. It was, from there, a fully violent paradigm shift.
It was not, as I may be making it sound, as if the Fed and its twin policymaking bodies were unaware of the problem, they just seemingly had little interest in it. They were captured instead by their statistical regressions that had, as it turned out, arbitrarily assigned causation and correlation in other directions. If you read through the transcripts of that time, and I cannot urge this strongly enough, though it is extremely tedious and, more often than not, insanely infuriating, the FOMC's attention was directed by the Open Market Desk toward indications and spreads like LIBOR/OIS or commercial paper/OIS. Bill Dudley, at the January 8, 2008, intermeeting (read: emergency) conference call was quite assuring on that front:
"Another positive development has been the improvement in the asset-backed commercial paper market. The volume of ABCP outstanding has stabilized, and the spread between the thirty-day ABCP rate and the one-month OIS rate has narrowed sharply. The spread relative to one-month LIBOR is about back to what it was before the financial market turbulence began in August. Bank sponsors have generally stepped forward to take problem SIV assets back on their balance sheets, and this has reduced the risk of asset fire sales."
Again, this is not to say that they were ignoring the potential downside, but they based their further measures and optimism in places that were ultimately inappropriate. In June 2008, in a meeting that can only be described as an "all clear" for the participants, Governor Kroszner quite typically underplayed the significance of what was still rotting underneath:
"In terms of how accommodative monetary policy is, I think actually it would be worthwhile-and maybe at the end of this I might pose a question to Brian-to look at LIBOR-OIS spreads and how much they typically go up during recessionary periods. I know that other risk spreads typically go up, but my understanding is that those typically don't go up as much. Since so many contracts are based off the one-month and three-month LIBOR, that 75 basis points suggests that at least now we might want to take that into account in thinking about where monetary policy stands relative to other times when we would have had a funds rate at roughly this level."
Spreads were not widening during a typical cyclical exchange, as Governor Kroszner implied, the entire systemic framework had already been fatally undermined, and in a manner that threatened not just banks but the entire economy. Something that started out as, and should have been contained as a nasty little recession, escalated into what may be (hopefully) the most significant economic event in nearly a century - a global malaise that is as yet unfinished.
The major part of the problem was both that participants and policymakers were overly-confident in their traditional understanding of the financial system, seemingly unaware of all the banking evolution that occurred while they played Nero with the economy, while also under the spell of recency bias created by their over-dependence on "big data." Econometrics had made such significant progress in developing regressions for nearly everything, but that really didn't mean what they thought. In their quest for simple correlations, the common sense idea of causation, the jagged edges of reality, was left behind as every nook and cranny was overly smoothed away by mathematics.
That has been a trend in centralization that goes way beyond just economics. This 21st century rush to contain and then "analyze" vast amounts of data, to try to turn what was previously just statistical noise into meaning, and not just meaning but full-blown predictions and ultimately control, has been a driving impulse of "technology." The internet plays a central role, as does computing power and linkages in general.
There are, of course, significant problems that if not fully incapacitating, at the very least speak to the lack of sufficient progress in those directions. Some of that goes way back to philosopher David Hume's observation that we can never really know causation since it is fully and eternally beyond the grasp of human experience; at best we can observe correlation and always and everywhere assume causation. Even gravity is an assumption (an extremely high assurance one), durable, forthright and predictable, because no one has ever observed it directly. We assume causation because it has never failed (quantum observations notwithstanding, though ultimately adding weight to the deficiency here) the predictability or replicability tests of true science.
Only a few years ago, "big data" suffered what might have been an irreversible failure of its current incarnation. Google, the championing behemoth of all things internet data, had proclaimed by 2008 the ability to use its algorithms of search topics and data to predict the flu season. Google Flu Trends was built upon statistical theories of correlation, and was even billed to be far superior to the CDC's own efforts.
However, as the Financial Times disparagingly described in March of this year,
"After reliably providing a swift and accurate account of flu outbreaks for several winters, the theory-free, data-rich model had lost its nose for where flu was going. Google's model pointed to a severe outbreak but when the slow-and-steady data from the CDC arrived, they showed that Google's estimates of the spread of flu-like illnesses were overstated by almost a factor of two.
"The problem was that Google did not know - could not begin to know - what linked the search terms with the spread of flu. Google's engineers weren't trying to figure out what caused what. They were merely finding statistical patterns in the data. They cared about ¬correlation rather than causation."
You can almost observe Dr. Edward Lorenz's butterfly effect in Google's failures, but there is something far deeper that is very problematic at this stage of technological attainment. The 2007 eurodollar freeze was as much a failure of "big data" (though perhaps smaller in scale but of the very same purpose and paradigm) as financial flow. At the very center of the housing bubble, the pivot that turned securitization from a niche into a mass producer of unthinkable gravity, was an eerily similar assumption about correlation.
In 2000, David X. Li, then a partner in JP Morgan's RiskMetrics unit, wrote a paper that revolutionized the entire mortgage business and allowed a sadly convenient marriage of that traditional credit basis with structured finance. Wall Street had been, up to that point, limited by pricing deficiencies in structured products from achieving a truly mass scale. Li's application of a Gaussian copula to CDO's and other structured products, particularly mortgage-related, unleashed the derivitivization of finance.
The problem has always been correlation, not as a philosophical journey, but as a practical and statistical matter for pricing. The Gaussian copula, pioneered by Li's theory, turned out correlation by simply assuming, like Google, that it already existed in other indications that simply needed to be "mined" and separated from the "noise." If you are familiar with implied volatility in options trading, the Gaussian copula accomplished something similar with correlation by simple inference - if credit spreads and curves were similarly situated or moved in the same manner, the copula's statistical assumptions just inferred some measured quantity of correlation.
That allowed sophisticated models to begin to price what was previously illiquid by nothing more than inference and really innuendo. The mistake in hindsight is obvious, though it should have been fully apparent contemporarily. The entire system of pricing credit, to the tune of trillions, was based on a data-driven assumption that was as valid as Google Flu Trends. The implications of mispricing, systemically, were enormous, including that violent paradigm begun in August 2007 (this is not to argue that there were not other factors at work, including uninterrupted monetarism, only to show how such intrusions can override natural means of self-correction and self-limitation to become so large in scale and detriment).
The central figures in those failures have not been expunged or even re-arranged; instead they've been heightened. There is a growing movement in central banking toward even greater control, driven by data and mathematical assumptions, at the targeted expense of free markets and free opinion.
While the ECB's recent move to a negative rate on its deposit account garnered the most attention, it was actually the companion program that was most significant, in my opinion. Mario Draghi rolled out the designs for what they are calling the T-LTRO's; T meaning "targeted." Central bank doctrine going all the way back to the days of Walter Bagehot was universal in its extreme reluctance to go beyond crisis liquidity management. Perhaps when orthodox economics began attempting a form of soft central planning that limitation was breached, but this is something altogether more severe and dangerously momentous.
Under the Bagehot regime, central banks were left with the job of "currency elasticity", to use the Federal Reserve's founding misdirection. That meant a generic "flood" of currency to combat a banking insufficiency. The theory was simple - the flood would end the banking problem and therefore the economy would mend. However, the geographic and simultaneous "floods" of currency the world over after badly misunderstanding the significance of August 9, 2007, has produced nothing along those lines. Banking panic may have disappeared from the front pages, but irregularities remain hidden under that cresting tide of monetarism.
Not the least of which is the failure of any major economy to experience anything like a recovery. Because of that, and there is more than a little hint of desperation in the ECB's move, Draghi's regime has decided it will now engage in direct allocation of financial resources. The ECB will create "liquidity" based on targeted lending to small and medium businesses in an effort to dissuade further "market" erosion in that segment - that market judgment has been deemed insufficient and must be overridden, apparently, by direct and central decree.
In June, the PBOC followed the ECB's footprint and very quietly implemented something called the "pledged supplementary lending" program (PSL). The idea is to, again, target specific sectors of credit markets. No longer will the central bank simply flood ledger money into the system and let it be sorted out by "market" forces, however badly restrained by other central bank means, the PBOC will actively engage in targets and allocation.
"PSLs, if deployed as China Business News described, would supplement the central bank's existing set of targeted tools for managing interest rates and liquidity, helping its ongoing campaign to rely more on precision firepower in its money markets to ensure capital is routed to productive uses, as opposed to opening the capital floodgate by reducing system-wide bank reserve requirement ratios (RRR) or reducing benchmark long-term interest rates."
The commentary above sounds reassuringly innocuous, as if there is nothing at all problematic about shifting toward more heavy reliance on centralization. Perhaps that is the nature of China that makes it seem not at all out of place, but this is not, again, isolated. Recent history has given us every reason to be highly suspicious not of market forces but the central bank intrusions into them. There is no basis whatsoever to simply assume that a central authority can, let alone should, break over a century and a half of useful and limiting tradition and begin to actively and fully allocate financial resources in an economy. This is well beyond cronyism (which is the inevitable destructive ends of non-economic decision-making) into the actual and active destruction of capitalism. Profit considerations are to be set aside and now all that will matter are concepts of social advancement (and then social "justice" or whatever political expediency arises).
This is a recipe for total economic inefficiency, as if no one currently making or reporting on policy has ever read or intuitively understood the immense power for truly free-market advancement expressed in I, Pencil.
"There is a fact still more astounding: The absence of a master mind, of anyone dictating or forcibly directing these countless actions which bring me into being. No trace of such a person can be found. Instead, we find the invisible hand at work.
"... Since only God can make a tree, I insist that only God could make me. Man can no more direct these millions of know-hows to bring me into being than he can put molecules together to create a tree."
Man cannot still "direct these millions of know-hows" but he is about to try in a manner entirely new to economic and financial experience. The absence of a "master mind" is slowly being "rectified" in the changing facilities of central banking, taking place everywhere failures continue to mount. Let's not forget Janet Yellen's very recent expressions and likely intentions of if not direct approval something very close to it. Central bankers decry the levels upon which bastardized "markets" are allocating resources, without seeing the bastardization for what it is, so there is now a move to implement minimum standards and narrowed targets, including specifications for credit and growth (new GSE leadership and directives fit comfortably within this observation). It may be limited now to the PBOC, ECB and Bank of Japan, but will there be enough resistance, based rightly on skepticism, in the US to avoid a similar fate?
Maybe this next step was inevitable, as the "flood" approach will always contain all manner of harmful side effects. However, it was always believed that those would be more than offset by beneficial gains, the accretion of "aggregate demand" that would fix so many of these problems. Instead of seeing that as an inseparable flaw of monetarism, the ratchet of bureaucracy tightens yet further under the guise of data and correlations of "science." The defining variable, for the central banker, is supposed to be precision based on their ability to gather and analyze data, as if central planners can identify the "correct" means of redistribution (theft) to create economic momentum. Yet, there are innumerable episodes that demonstrate, including those I have described, with total conclusivity no such precision exists. There is infinite wisdom in the dispersed nature of free market systems that confounds all attempts at harnessing it for the means of nothing more than total control. That used to be recognized as a universal truth, and I have no doubt it will be again at some point that seemingly draws closer with every malaise-tinged bubbly tick.