Aftermath
Page 25
This can be illustrated by the circulatory system of the blue whale. The whale’s heart is the central pump. When blood leaves the heart it flows quickly and smoothly through large blood vessels, like the aorta. Yet the blood vessels must eventually shrink to the size of capillaries that can reach the terminal units of tissue that receive the oxygen in the blood. All parts of the body must receive these blood cells or they die from oxygen depletion. The larger the creature, the more junctions or splits are needed as the blood flows from the aorta to a wide network of capillaries. At each junction, there is some resistance or impedance to the flow (in effect, the heart is pumping against itself) and more energy is required. The large network of arteries and capillaries also causes friction, which puts more stress on the heart and uses more energy. As the energy needed to pump the blood exceeds the energy available in the blood, the whale starts to die of hypoxia. Scientists have estimated that given the distance and energy loss from the heart to the capillaries, the maximum size of a mammal is approximately 250,000 pounds, about the size of a typical blue whale. Stated simply, the blue whale is the largest creature ever seen because it is the largest creature that can ever exist given the constraints of energy outputs, inputs, and hypoxia. Again, there are limits to growth.
What do these biological and structural insights have to do with finance? The creatures and structures studied by Geoffrey West and his colleagues are examples of complex dynamic systems. Such systems include animals, forests, businesses, cities, and the cosmos. Capital markets are one of the best examples of a complex dynamic system. Capital markets exhibit the behaviors and constraints of other biological and man-made complex systems. Limits on size, as represented by the blue whale, also apply to ships, buildings, friendships, nature, and finance. Snow piles up on a mountainside for only so long before the snowpack destabilizes and collapses into an avalanche. A vessel can only be so large before it capsizes and sinks. A bank balance sheet can only be so leveraged before confidence is lost and the bank tumbles into failure. West and others have identified scaling metrics and limits to growth in sundry systems. What are the limits to growth in finance?
While complex systems have unique characteristics (a snowpack is different from a patch of moss, which is different from a whale’s bloodstream), they all have certain dynamics in common. They exhibit diversity, communication among parts, interaction among parts, and adaptive behavior. They exhibit emergent properties; behavior arises that cannot be inferred from perfect knowledge of the system components. They exhibit scale invariance; subunits are near perfect replicas of larger units, like a stream and a river, or a branch and a limb. Energy inputs scale faster than energy outputs, so that new sources of energy (or efficiency) are constantly needed. Most important, the risk of extreme behavior or a systems collapse is a superlinear function of scale.
These complex system dynamics have been expressed theoretically and shown empirically in myriad systems. West and his colleagues have shown how the number of gas stations in a city scales in relation to the population of a city. The slope of the logarithmic power curve that compares population size to the number of gas stations is 0.85. This means that when you double the population size, you increase the number of gas stations by 85 percent. That’s a 15 percent reduction compared to doubling the number of gas stations—an example of economies of scale. Each gas station handles more customers. The scope of the 0.85 exponent is illustrated by the fact that it applies to all cities in all countries, regardless of size. If you inform West of the name and size of a city anywhere, he can tell you how many gas stations that city has with a high degree of accuracy, based solely on the slope of the power curve. There are countless similar examples in natural and man-made systems.
In the gas station case, scientists had a scaling metric (population size of the city) and a specified slope of the power curve (0.85 based on empirical research). What if you had neither? What if you had a complex dynamic system based on composition and behavior but no agreed scaling metrics and no empirical research on the frequency of extreme behavior? That is roughly the situation facing financial risk management today. Capital markets are certainly complex dynamic systems; they exhibit all of the conditions of complexity, including diversity, communication, interaction, and adaptive behavior in the right middling proportions. Capital markets can be scaled up or down, as we saw in 2007 to 2009, and are prone to serious collapse, as we saw in 1987, 1994, 1998, 2000, and 2008. What is unknown is the best way to measure scale and the exponent of the slope of the power curve that links scale to collapse.
Candidates for measuring capital markets’ scale include bank asset size, gross notional value of derivatives, concentration of assets in fewer banks, contingent liabilities (clearinghouse credit, guarantees, options, etc.), non-bank lenders, trading volumes, SWIFT message traffic, and other gross metrics. A weighted blend of numerous factors may prove best. Risk factors that could trigger a market collapse include leverage ratios, government debt-to-GDP ratio, government deficit-to-GDP ratio, real interest rates, real growth, nominal growth, contagion, trade wars, currency wars, credit spreads, and geopolitical shocks. A forecaster using predictive analytic science looks for a critical state measured by scale combined with a catalyst of two or more triggers acting as force multipliers in a feedback loop (such as excessive leverage and declining asset values). The difficulty is that while most of these factors are tracked on an individual basis by specialists, there is no effort to synthesize the factors into a mosaic that captures unexpected emergence, amplification, feedback, and contagion. The most challenging dynamic is the extent to which the capital market interacts with itself.
Which brings us back to Godzilla. In studies of scaling limits based on variable exponents for volume versus strength and the branching needed to reach terminal units, West was always careful to specify that the scaling limitations applied, “provided, of course, that the materials they’re made of don’t change so that their densities remain the same.” In other words, a one-hundred-foot Godzilla can be compared to a three-hundred-foot Godzilla (the first is barely possible, the second is impossible) only if the same organic body parts are involved. If the three-hundred-foot Godzilla is a robot made of titanium and copper wire, then a different factor analysis is required. Or, perhaps a three-hundred-foot Godzilla could pose upright if lashed to a steel scaffold.
Are there hidden scaffolds in finance?
A large financial institution like JPMorgan is a complex entity, like capital markets more broadly. It has the markers of diversity, communication, interaction, and adaptive behavior. It exhibits emergent properties such as the sudden appearance of $6 billion in trading losses in May 2012, attributable to actions by a trader, Bruno Iksil, known as the London Whale. Iksil’s trading in credit derivatives expanded for over a year with poor supervision before losses unexpectedly emerged—behavior typical of complex systems.
Just as a blue whale has a terminal unit, the blood cell, a bank has a terminal unit, which is the client. Deals can be for $100 at an ATM or $100 billion at a corporate closing, but in each case the bank faces off with a client. A bank’s blood is money. Just as blood flows through arteries and capillaries, so money flows through payment systems all the way down to cash dispensers. Blood carries energy in the form of oxygen and nutrition and removes waste. Likewise, money is stored energy. You expend energy as labor or capital to earn money, and use money to release energy by hiring workers or investing in plant and equipment. Money is stored energy that powers a capitalist economy. Banks provide the cardiovascular system to move the money around.
Godzilla cannot exist because he would fall of his own weight. No creature larger than a blue whale can exist because the energy needed to move its blood is greater than the energy available. Skyscrapers have different height-volume equations than Godzilla because they substitute steel for bone. Still, they too have limitations on absolute height and volume. What are the limitations of size and scale for a bank? In what way can a bank
fail due to excessive scale in the way an oversized mammal can collapse of its own weight or suffer hypoxia and die?
We don’t have exact metrics on excessive scale in banks because the scaling factors have not been specified and empirical testing has not been conducted. So-called stress tests imposed on banks by the Treasury are mostly for show and involve static capital adequacy, which is never sufficient in extreme stress. The fact that banks can fail due to excessive size has been demonstrated repeatedly. In 2008, Bear Stearns, Fannie Mae, Freddie Mac, and Lehman Brothers failed successively between March 18, 2008, and September 15, 2008, due to overleverage, contagion, and credit spreads—three of our scaling metrics. After Lehman Brothers’ bankruptcy, Morgan Stanley was days away from failure and Goldman Sachs and other major banks were next in line to fail. This is evidence that the individual banks, and the system as a whole, were suffering the institutional equivalent of hypoxia. They were out of oxygen and dying off. Only government intervention in the form of deposit guarantees, money market guarantees, term-lending facilities, and trillions of dollars of currency swaps propped up what was left of the banking system. Think of the big banks as collapsing Godzillas and the Federal government as a steel scaffold put in place to prop up the bloated banks.
Once the scaffolding around the banks is in place it cannot be removed unless the banks shrink or the risk of collapse is accepted. The Fed cannot escape the room. Reduced derivatives exposure does not reduce systemic risk when bank derivatives are simply moved to a clearinghouse whose credit is backed up by the same banks. Increased bank capital does not mitigate risk when leverage is still too high and credit risk is understated by flawed models. At most, the added capital buys a few days to work out a rescue. The Fed scaffold around banks like JPMorgan cannot be removed. Worse, it allows Godzilla banks to grow larger, to the point when they collapse of their own weight and take the scaffold down with it.
Middle Class Agonistes
How often have peers lamented the “death of the middle class”? The terms “death” and “middle class” are not defined, yet everyone gets the point. The rich are undeniably richer. The poor are struggling to make ends meet. Meanwhile, the middle-class member works hard, supports a family, pays the majority of all federal income taxes, gets little federal support during her working years, and seems to bear the weight of society.
The middle class is not disappearing, there may be 100 million members of the U.S. middle-class depending on the definition. Yet they are struggling to hold onto their status; members feel that they are clinging to middle-class rank by their fingernails. This insecurity of the U.S. middle class holds warnings for investors. The economic prognosis is poor for members of the middle class and the broader society. The ability of individual members of the middle class to reverse this trend is limited by a lack of political power and elite indifference.
There is no standard definition of middle class. Still, several academic models stand out. One model, developed by Leonard Beeghley in his work The Structure of Social Stratification in the United States (2016), breaks U.S. society into four groups: the rich (5 percent), the middle class (45 percent), the working class (40 percent), and the poor (10 percent).3 In this ontology, the status of the rich (annual incomes over $350,000) and the poor (living below the poverty line) are self-evident. However, the two middle categories might all be considered middle class by some analysts. The rich have a net worth of $1 million or more. However, most of that is in illiquid home equity. Individuals with home equity of $1 million likely consider themselves “upper middle class” if not quite rich, depending on their zip code. Likewise, Beeghley’s working class includes those making $40,000 per year, a group that might consider itself “lower middle class.” If these expanded definitions are used, the middle class might be 89 percent of the population with a 1 percent super-rich group and 10 percent in poverty.
Another ontology is offered by William Thompson, Mica Thompson, and Joseph Hickey in their work Society in Focus (2017).4 The Thompson, Thompson, and Hickey rankings adhere more closely to the foregoing revised version of the Beeghley rankings. Thompson, Thompson, and Hickey have an upper class (1 percent) and a lower class (20 percent). In between are an upper middle class (15 percent), lower middle class (32 percent), and a working class (32 percent). The upper class has incomes of $500,000 per year or more. The upper middle class earns between $75,000 and $499,000 per year. The lower middle class earns between $35,000 and $74,000 per year. The working class earns between $16,000 and $35,000 per year. The lower class has little earned income; they receive government transfer payments or are in poorly paid positions.
Finally, Dennis Gilbert offers a third ontology in his book The American Class Structure in an Age of Growing Inequality (2015).5 Gilbert offers six levels of society instead of the customary five. He describes a capitalist class (1 percent), an upper middle class (14 percent), a lower middle class (30 percent), a working class (30 percent), the working poor (13 percent), and the underclass (12 percent). Gilbert relies on job descriptions and educational attainment rather than income levels to identify his classes, although it’s straightforward to attach estimated income levels to each group. The capitalist class consists of CEOs and politicians. The upper middle class consists of professionals and middle managers. The lower middle class consists of semiprofessionals and craftsmen. The working class consists of blue-collar workers. The working poor consists of low-level clerks. The underclass is generally not in the labor force and receives government transfer payments.
There are various other income distribution studies available and other class definitions used by economists and social scientists. The most common way of approaching the problem is to divide society into five tiers or quintiles, where each tier has exactly the same number of members and the tiers are divided by income levels without regard to job descriptions or education. This method shows that as of 2016, the highest 20 percent of U.S. households received 51.5 percent of total income. The next 20 percent of households received 22.9 percent of total income. The middle 20 percent received 14.2 percent of total income. The next 20 percent received 8.3 percent of total income, and the bottom 20 percent of households received 3.1 percent of total income. Put differently, the top 40 percent of households receives 74.4 percent of the total income while the bottom 60 percent of households receives only 25.6 percent of total income.
This quintile analysis shines a harsh light on several realities of income distribution in the United States today. The first is that these figures represent household income, which depends on the size of the household. If these figures were computed for individuals, income concentration would be even more skewed in favor of the rich. The second reality is that the U.S. trend is toward greater income inequality. The lowest 60 percent of households earned 32.3 percent of total income in 1970 versus 25.6 percent of total income today. That’s a stunning 21 percent drop in the share of the lowest 60 percent in the past forty-eight years. Meanwhile the highest 20 percent saw a 19 percent gain in their share (from 43.3 percent to 51.5 percent) in the same time period. In the United States today, the adage that the rich get richer and the poor get poorer has seldom been more true.
Whether you analyze the U.S. economy by quintiles, income brackets, job descriptions, or on a highly granular basis with regard to the top 0.01 percent, the result is the same—incomes and net worth in the United States exhibit a high degree of concentration of income and assets among the wealthiest, with a far larger group struggling to hang on to their piece of the American dream.
This data can be used to develop a simpler definition of middle class. Divide the population into rich, upper middle class, lower middle class, working class, and poor. The rich are the top 1 percent, with incomes of $500,000 per year or more. The upper middle class have incomes from $100,000 to $500,000 per year. The lower middle class have incomes from $35,000 to $100,000 per year. The working class have incomes from $15,000 to $35,000 per year. The poor have incomes below $15,000 pe
r year and receive government assistance. This class structure shows the middle class is 85 percent of the total population or 270 million Americans.
Several clarifications are needed. The first is that the figure of 270 million persons includes homemaker spouses and children. With an average family size of 3.5 persons, the number of middle class workers is closer to 80 million, still a large number but only 25 percent of the total U.S. population. The second clarification involves taxation. The income numbers cited above are pretax. When a 30 percent statutory marginal tax rate is applied, the after-tax income figures drop considerably from the pretax levels. A $200,000 income drops to $140,000 after tax, and a $100,000 income drops to $70,000 after tax. Those after-tax numbers are far more modest than the pretax figures.
The pre-tax–after-tax distinction is important because the tax burden falls disproportionately on the middle class. Pew Research Center data reveals that the middle class (as defined above) pays over 60 percent of all income taxes. The poor pay almost no income tax because of low rates, exemptions, and credits. The rich pay 38.3 percent of all income taxes, yet make over 50 percent of total income. This lower effective tax rate compared to the middle class is due to income deferral plans and preferential rates on capital gains. The middle class is right to feel overtaxed relative to both the rich and the poor.
A struggling middle class has more to do with the future than the present. While today’s numbers testify to the presence of a large middle class, their mood is pessimistic. There is a feeling that children will not do as well as their parents. There is a feeling of job insecurity. There is a feeling of overtaxation relative to other echelons of society. Above all, there is a feeling of a rigged game in which the rich share inside information, the poor are subsidized, and the middle class does all the work and receives no respect from elites or political leadership. None of these feelings is misplaced. Burdens placed on the middle class have never been greater, even as society’s rewards are snatched by super-rich investors or recipients of government assistance.