Capitalism at a Cross Roads
Considerations of Persistent Disequilibria and Positive Feedback: Finding the Butterfly
April 8, 2012
The twin marvels of truly awful weather and systemic market meltdowns have a lot in common. First they feed on persistent disequilibria in their respective systems – weather systems feed on temperature and pressure differentials which can only be resolved through long time frames for adjustment or short term violence – storms. Similarly, economic and financial disequilibrium tend to be founded on regulatory or information impediments, leading to persistent positive feedback loops – first running markets up, and then, inevitably, leading to collapse.
Awful weather – a self-feeding phenomena
Really chaotic behavior in weather systems and financial markets requires that several positive feedback phenomena – think low interest rates, rising housing prices, systematic mispricing of securities, failed regulation, and systemic level fraud – are required to produce a real financial apocalypse. What makes the process interesting is that these factors may exist in the system functioning perfectly for decades, and only when combined a special brew a la the Three Witches of Macbeth do we get the resultant fiasco.
There has been a steady stream of metaphorical comparisons of Sebastian Junger’s book, The Perfect Storm, with the Great Recession in general, and the meltdown in sub-prime mortgages in particular. It is worth taking a look at what actually made the Perfect Storm unique, and what comparisons are meaningful to the sub-prime meltdown vis-à-vis complexity theory.
If they merge – you get the perfect storm
The Great Recession is a rolling disaster that the world economy seemingly stumbled into in the spring of 2008 led by a housing collapse which began in the sub-prime mortgage market. This rolling economic Verdun was born many years before in a series of ill-conceived and poorly understood policy initiatives championed by consumer advocates and industry groups alike. These reforms – generally characterized as deregulation – were targeted at the seemingly laudable targets of risk mitigation and fairness, and categorically achieved just the opposite. In this paper I take a look at the human faces behind the fiasco, it takes clever and dedicated people to create a mess this big.
A result of negative feedback
Each of these examples is a terrific case study of chaos theory at work. In chaos theory, a small action/change in a complex/dynamic system can have enormously disproportionate impact in the future. The classic example is posited by Lorenzo in his 1972 paper, Predictability: Does the Flap of a Butterfly’s Wings in Brazil set off a Tornado in Texas? What is most important to remember here is that chaos theory is very sensitive to initial conditions, and unpredictable beyond very short periods of time. For those making economic policy this requires a constant vigilance for the undesirable unintended consequences of policy changes – clearly the butterfly did not intend the tornado.
It is to the quest of finding Lorenzo’s butterfly that I dedicate this series of articles.
Considerations of Complexity and Chaos
Complexity is a relatively new area of study, in which the nice equations that we learned in economics and physics are tested against the vagaries of reality. From a practical perspective, it is, at its core a study of the realistic implications of physical and social entropy that we see all around us. To a large degree, complexity theory borrows heavily from the study of biological systems, and applies those lessons to areas as diverse as meteorology to economics. All complex systems share four fundamental characteristics:
1) Autonomous Agents All of the independent actors in any complex system are autonomous agents which produce results in the system. These agents may be storms in a weather system, or stock and bond traders in financial markets. Diversity is a key factor in Autonomous agents – the less diversity, the less complexity and potential variation in outcomes.
2) Connected All Agents in a complex system must be connected – directly or indirectly. Typical examples might include connections via energy sources or communication patterns.
3) Interdependence All Agents in a complex system influence each other through one or more series of connections, and are therefore interdependent, although autonomous.
4) Adaptation Any complex system adapts to external influences in unpredictable ways, and complex social systems are capable of and depend upon learning.
Complex systems are normally open systems like weather, which exist in a thermodynamic gradient and function to dissipate energy in a fundamentally unstable system – a persistent disequilibrium.
Such complex systems frequently produce negative or positive feedback loops, in which actions in the systems in turn impact the system itself – producing more of the original activity. A good example is the self-feeding nature of a hurricane, which continues to grow by deriving energy from heated water, which increases the size and force of the storm, which in turn adds more energy and leads to further growth.
A common positive feedback loop in finance is the positive effect that ample mortgage finance has on real estate prices, feeding more transactions at ever higher values. Such a system is said to be in a persistent disequilibrium until the availability of mortgage financing is curtailed.
Chaotic systems: 1) Are sensitive to initial conditions, 2) Are topologically mixing, and 3) Have dense periodic orbits. Of these three, point one is the object of interest for any amateur historian/scientist – can you figure out where and when the butterfly flapped its wings?
Not too surprisingly, the description of the inputs to complex systems depends – to a large degree – on the academic background and orientation of the party doing the explanation and evaluation. Physicists tend to describe complexity as requiring constant inputs of energy and developing closed feedback loops limiting the dissipation of energy. That might be a pretty good description of a super-charged storm system like a hurricane.
Economists talk of the primacy of information, learning and adaptation and the central role of competition in the adaptive process. In either case, while you are dealing with radically different systems and agents, the same four fundamental characteristics apply.
The following chart gives a pretty good idea of how different branches of study use complexity to explain the same thing – the driving force in each system.
Looking at the simple chart above, it highlights the complexity of a system involving humans; as we are uniquely driven by all of the above factors.
The Perfect Storm as a Metaphor for the Great Recession
Sebastian Junger did a great job touching the mystery and horror associated with the Great Halloween Storm of 1991 in his book, The Perfect Storm. What made the storm particularly noteworthy was its absolutely unique origin – the merger of 1) the remnants of a blown-out hurricane (Grace), 2) an extra-tropical low pressure system, and 3) a vigorous, early season cold front blowing out over the Gulf Stream – which was still unusually warm for relatively late in the season – around 80 degrees.
While the wind speed of the storm was only around 60-75 mph, it took a very unusual track – heading directly back toward shore, and producing some truly gigantic waves. The tides and waves produced some historic damage including the sinking of the Andrea Gail – which was the story line of both the book and movie.
The factors that make the Perfect Storm a good metaphor for the financial wreckage of the Great Recession are: 1) It was born of the merger of a number of small and seemingly normal phenomena into a super storm, 2) It incorporated a number of previously dissipated tempests and reinvigorated them with the infusion of new energy, 3) It brought together common factors – the Gulf Stream and Canadian air masses – in a unique and enormously powerful and destructive manner, and 4) The storm track was very unusual and dramatically amplified the damage from the storm. This sounds like a pretty reasonable group of factors to degrade complexity into destructive chaos.
The Distant Origins of the Sub-Prime Fiasco
Vast amounts have been written about the origins of the Great Recession. Most of it is correct, and all of it (as far as I can tell) is incomplete. The closest that I have found is the work of McLean & Nocera in their book, All the Devils are Here, which begins with the original securitization of mortgages in the early ‘80’s. They focused on the splendid trio of characters Lewis Ranieri of Solomon Brothers, David Maxwell of Fannie Mae, and Larry Fink of First Boston. Of the three, Ranieri was doubtlessly the most colorful – an up-the-hard-way kid from the streets of Brooklyn, who wanted to be an Italian chef, but was chased out of the kitchen by asthma, and into the mailroom at Salomon Brothers, from which he ultimately rose up to be Vice Chairman.
These three guys were simultaneously fierce competitors and iron-bound allies in negotiations and lobbying the Federal Government and Congress for legal and regulatory accommodation to create the securitized mortgage industry. They were tireless, driven and ultimately successful.
Lewis Ranieri – The father of Mortgage Securitization
McLean & Nocera begin their tale at the right point in time, and with the right characters, but fail to correctly identify the fundamental reason that securitization was perceived and sold to the American people and regulators as an absolute necessity. They correctly identify that the Baby Boomers were coming of age and needed mortgages, and the entire S&L industry which would normally provide them was in desperate straits – hemorrhaging cash and on the verge of mass extinction. The fundamental disequilibrium that caused this fiasco – they never addressed; and that is where all the fun was.
In the late 70’s and early 80’s, Paul Volker, dramatically increased interest rates, successfully reversing the pervasive inflation expectations born of the unilateral abrogation of Bretton Woods and resulting currency collapse under the Nixon Administration. There was at the time a little noted regulation – known as Regulation Q – which limited the amount that S&L’s could pay on deposits. The regulators in Washington refused to allow S&L’s to pay market rates, and deposits fled in droves, forcing the S&L’s into the capital markets for liquidity when they could, or to Government funding when they could not. This financing was at rates up to 18%, the vast majority of their assets paid interest at less than 6% – and they were bleeding cash –and lots of it. This was a system in massive disequilibrium caused solely by the Federal Government.
The S&L’s did what they could to try to attract deposits, including giving away appliances as “gifts” to depositors willing to accept below market interest rates. During this period most banks in the US came to resemble hardware/appliance stores – but with armed guards. It was all very strange, and could have been fixed with a little Reagan era deregulation – simply repeal Regulation Q. That, however, would have eliminated the need for mortgage securitization (and money market funds, also born of the same disequilibrium) – the fundamental underpinning that ultimately fed the sub-prime fiasco – and Wall Street profits for 20 years. The combined lobbying forces of Wall Street and the GSE’s (Fannie, Freddie and Ginnie) in Washington simply overwhelmed the local bankers from the S&L’s, the most successful system of home financing in history was tossed into the trash, and the mortgage backed securities industry was born. When historians undertake to total the costs of the bailout for the Great Recession, they must include the $300 billion that was spent to clean up the S&L mess in the late 80’s and early 90’s.
Can you identify the motivations?
Among the Devils – Consider David X. Li
Cry, ‘Havoc!’ and let slip the dogs of war. (3.1.268)
William Shakespeare, Julius Caesar
Much has been written about David X. Li and his pesky little formula that nearly destroyed the world. Li left China in 1987 (a singularly inauspicious start date) to come to the West to study the ways of finance – and apparently he mastered the art of the fomenting uncontrolled market crashes. He spent several years after arriving in Canada studying business and then statistics, ultimately earning a PhD with a focus on actuarial issues. In 2000 he published the seemingly simple paper entitled, “On Default Correlation: A Copula Approach”, and the fuse was lit on the real estate bomb that exploded in full force in 2008. The formula that was the highlight of the paper, relied on an old actuarial principal of correlation – specifically developed around the predictive value of the demise of one marital spouse on the demise of the second. This coupling or copula was the fundamental concept that was extended to bundles of assets – mortgages, auto loans and the like – to the value of credit derivative swaps (CDS’s) used to insure against the failure of the assets in question, or similar securitized assets.
The Formula Credited with the Sub-Prime Crisis
A Mathematical Expression of a Positive Feedback Loop
The Harbinger of Financial Chaos
In Mr. Li’s defense, he did not initially actively promote his formula as a solution for the pricing of portfolios of sub-prime mortgages or other financial dross, but once it started he participated full on. The formula is a dramatic simplification, relying on a constant of correlation to describe the future movement of asset values and performance. Nothing in any complex system is ever correlated by a constant value. The very concept of a concept of a constant of correlation is an oxymoron in a complex system. I guess that for Mr. Li financial markets in junk securities would follow a smooth and even course to maturity – despite all of the evidence to the contrary. In any event, Wall Street bought in hook, line and sinker – and entered into the headlong rush to the apocalypse.
I have only hit the highlights of this story, and would recommend to the reader checking out Felix Salomon’s terrific article in Wired for a more thorough and entirely entertaining read on the subject, http://www.wired.com/techbiz/it/magazine/17-03/wp_quant?currentPage=all .
4.5x Growth following a pesky little equation
A Bolt of Lightning into the Primordial Goo
My high school biology teacher described the origins of life as caused by a bolt of lightning into the primordial goo. I always loved the imagery, and the obvious link to Mary Shelley’s Frankenstein brought the fantastic to life. This was the same effect that David Li’s little equation had on the primordial goo waiting in the sea of mortgage finance. Securitization allowed the direct sale of risk and CDS’s allowed the insuring against them whether or not you even owned the underlying risk. Li’s equation allowed Wall Street to quantify the risk of a portfolio of assets without even knowing what was inside. Better still, the rating agencies bought into all of it, and were rating portfolios sight unseen. The wonder of it all.
This equation led directly to the muscle-bound thugs running around poor communities selling mortgages the same way that they used to run numbers. Like the Canadian cold front teaming up with defunct Hurricane Grace and the energy in the Gulf Stream in the Perfect Storm, David Li’s equation hit the primordial goo of cheap money, dozy regulators, off-exchange CDS’s, compromised rating agencies and securitization machines craving product. In both cases, the positive feedback loop was spectacular, and the results disastrous.
Time for a Regulatory Re-think
Alan Greenspan used to say that he could not recognize a bubble until it popped. While I find this to be nonsensical, I’ll humor the position by suggesting that in each of the articles in this series I have clearly identified that there has been disproportionate growth – or shrinkage – in a major class of financial instruments presaging every financial catastrophe that I’ve examined. In each case there has been some change in the environment which created a fundamentally unstable system – and led directly to either a positive or negative feedback loop. These aren’t hard to identify, even with the use of retrospective data as favored by regulators.
From a practical perspective, the purpose of financial regulation is always two-fold; 1) Isolate the transaction into a singular event equally understood by both parties, and 2) Create the greatest level of information parity between the parties. The purpose of the derivative markets is to confound the first, and gaining an information edge is the second oldest profession. The pace of change, asset concentration and venue selection opportunities make old static regulatory regimes and strategies almost worthless.
Positive and negative feedback loops are more common in finance today than ever before, but they can be identified simply by looking at differential rates of growth of asset classes in the economy. In short, if it is busted – fix it.
Can you hear the butterfly wings flapping yet?
Reaping the Whirlwind