Submitted by Mark Spitznagel, CIO of Universa Investments LP: a white paper
The Austrians and the Swan:
Birds of a Different Feather
On Induction: If it looks like a swan, swims like a swan…
By now, everyone knows what a tail is. The concept has become rather ubiquitous, even to many for whom tails were considered inconsequential just over a few years ago. But do we really know one when we see one?
To review, a tail event—or, as it has come to be known, a black swan event—is an extreme event that happens with extreme infrequency (or, better yet, has never yet happened at all). The word “tail” refers to the outermost and relatively thin tail-like appendage of a frequency distribution (or probability density function). Stock market returns offer perhaps the best example:
Over the past century-plus there have clearly been sizeable annual losses (of let’s say 20% or more) in the aggregate U.S. stock market, and they have occurred with exceedingly low frequency (in fact only a couple of times). So, by definition, we should be able to call such extreme stock market losses “tail events.”
But can we say this, just because of their visible depiction in an unconditional historical return distribution? Here is a twist on the induction problem (a.k.a. the black swan problem): one of vantage point, which Bertrand Russell famously described exactly one-hundred years ago with his wonderful parable (of yet another bird):
The man who has fed the chicken every day throughout its life at last wrings its neck instead, showing that more refined views as to the uniformity of nature would have been useful to the chicken…The mere fact that something has happened a certain number of times causes animals and men to expect that it will happen again.
Bertrand Russell, The Problems of Philosophy (1912)
My friend and colleague Nassim Taleb incorporates Russell’s chicken parable as the “turkey problem” very nicely in his important book The Black Swan. The other side of the coin, which Nassim also significantly points out, is that we tend to explain away black swans a posteriori, and our task in this paper is to avoid both sides of that coin The common epistemological problem is failing to account for a tail until we see it. But the problem at hand is something of the reverse: We account for visible tails unconditionally, and thus fail to account for when such a tail is not even a tail at all. Sometimes, like from the chicken’s less “refined views as to the uniformity of nature,” what is unexpected to us was, in fact, to be expected.
II. Not Just Bad Luck: The Austrian Case
Perhaps more refined views would be useful to us, as well.
This notion of a “uniform nature” is reminiscent of the neoclassical general equilibrium concept of economics, a static conception of the world devoid of capital and entrepreneurial competition. As also with theories of market efficiency, there is a definite cachet and envy of science and mathematics within economics and finance. The profound failure of this approach—of neoclassical economics in general and Keynesianism in particular—should need no argument here. But perhaps this methodology is also the very source of perceiving stock market tails as just “bad luck.”
Despite the tremendous uncertainty in stock returns, they are most certainly not randomly-generated numbers. Tails would be tricky matters even if they were, as we know from the small sample bias, made worse by the very non-Gaussian distributions which replicate historical return distributions so well. But stock markets are so much richer, grittier, and more complex than that.
The Austrian School of economics gave and still gives us the chief counterpoint to this naïve vew. This is the school of economic thought so-named for the Austrians who first created its principles3, starting with Carl Menger in the late 19th century and most fully developed by Ludwig von Mises in the early 20th century, whose students Friedrich von Hayek and Murray Rothbard continued to make great strides for the school.
To Mises, “What distinguishes the Austrian School and will lend it immortal fame is precisely the fact that it created a theory of economic action and not of economic equilibrium or non-action.” The Austrian approach to the market process is just that: “The market is a process.” Moreover, the epistemological and methodological foundations of the Austrians are based on a priori, logic-based postulates about this process. Economics loses its position as a positivist, experimental science, as “economic statistics is a method of economic history, and not a method from which theoretical insight can be won.” Economic is distinct from noneconomic action—“here there are no constant relationships between quantities.” This approach of course cannot necessarily provide for precise predictions, but rather gives us a universal logical structure with which to understand the market process. Inductive knowledge takes a back seat to deductive knowledge, where general principles lead to specific conclusions (as opposed to specific instances leading to general principles), which are logically ensured by the validity of the principles. What matters most is distinguishing systematic propensities in the entrepreneurial-competitive market process, a structure which would be difficult to impossible to discern by a statistician or historian.
To the Austrians, the process is decidedly non-random, but operates (though in a non-deterministic way, of course) under the incentives of entrepreneurial “error-correction” in the economy. In a never ending series of steps, entrepreneurs homeostatically correct natural market “maladjustments” (as well as distinctly unnatural ones) back to what the Austrians call the evenly rotating economy (henceforth the ERE). This is the same idea as equilibrium, but, importantly, it is never considered reality, but rather merely an imaginary gedanken experiment through which we can understand the market process; it is actually a static point within the process itself, a state that we will never really see. Entrepreneurs continuously move the markets back to the ERE—though it never gets (or at least stays) there. Rothbard called the ERE “a static situation, outside of time,” and “the goal toward which the market moves. But the point at issue is that it is not observable, or real, as are actual market prices.”
Moreover, “a firm earns entrepreneurial profits when its return is more than interest, suffers entrepreneurial losses when its return is less…there are no entrepreneurial profits or losses in the ERE.” So “there is always competitive pressure, then, driving toward a uniform rate of interest in the economy.” Rents, as they are called, are driven by output prices and are capitalized in the price of capital—enforcing a tendency toward a mere interest return on invested capital. We must keep in mind that capitalists purchase capital goods in exchange for expected future goods, “the capital goods for which he pays are way stations on the route to the final product—the consumers’ good.” From initial investment to completion, production (including of higher order factors) requires time.
By about one hundred years ago, the Austrians gave us an a priori script for the process of boom and bust that would repeatedly follow from repeated inflationary credit expansions. Without this artificial credit, entrepreneurial profit and loss (“errors”) would remain a natural part of the process, except that, for the most part, they would naturally happen quite independently of one-another.
Central to the process is the “price of time": the interest rate market. This market conveys tremendous information to entrepreneurs due to the aggregate time preference (or the degree to which people prefer present versus future satisfaction) which determines it and is reflected in it. Interest rates are indeed the coordinating mechanism for capital investment in factors of production.
Non-Austrian economists typically depict capital as homogeneous, as opposed to the Austrians’ temporally heterogeneous and complex view of the capital structure. We see this in the impact of interest rate changes. Low rates entice entrepreneurs to engage in otherwise insufficiently profitable longer production periods, as consumers’ lower time preference means they prefer to wait for later consumption in the future, and thus their additional savings are what move rates lower; high rates tell entrepreneurs that consumers want to consume more now, and the dearth of savings and accompanying higher rates make longer-term production projects unattractive and should be ignored in order to attend to the consumers’ current wants. The present value of marginal higher order (longer production) goods is disproportionately impacted by changes in their discount rates, as more of their present value is due to their value further in the future.
Variability in time preferences changes interest and capital formation. If lower time preference and higher savings and lower interest rates created higher valuations in earlier-stage capital (factors of production) which initiates a capital investment boom, this newfound excess profitability would be neutralized by lower demand for present consumption goods and lower valuations in that later-stage capital. (John Maynard Keynes’ favored paradox of thrift is completely wrong, as it ignores the effect on capital investment of increased savings, and resulting productivity—and ignores the destructiveness of inflation, as well.)
But there is an enormous difference between changes in aggregate time preference and central bank interest rate manipulation. Where this is all heading: The Austrian theory of capital and interest leads to the logical explication of the boom and bust cycle. To the logic of the Austrians, extreme stock market loss, or busts—correlated entrepreneurial errors, as we say—are not a feature of natural free markets. Rather, it is entirely a result of central bank intervention. When a central bank lowers interest rates, what essentially happens is a dislocation in the market’s ability to coordinate production. The lower rates make otherwise marginal capital (having marginal return on capital) suddenly profitable, resulting in net capital investment in higher-order capital goods, and persistent market maladjustments.
Despite the signals given off by the lower interest rates, the balance between consumption and savings hasn’t changed, and the result is an across-the-board expansion—rather than just capital goods at the expense of consumption goods. What the new owners of capital will find is that savings are unavailable later in the production process. These economic cross currents—more hunger for investment by entrepreneurs seizing perceived capital investment opportunities, and consumers not feeding that hunger with savings, but rather actually consuming more—creates a situation of extreme unsustainable malinvestment that ultimately must be liquidated.
The only way out of the misallocated, malinvestment of capital, is a buildup of actual resources (wealth) in the economy in order to support it. This could result from lower time preferences (but as we know compressed interest rates actually inhibit savings)—or of course by accumulated reinvested profits over time (but of course time will not be on the side of marginal malinvested capital earning economic losses).
Credit expansion raises capital investment in the short run, only to see the broad inevitable collapse of the capital structure. Eventually the economic profit from capital investment and the lengthening of the production structure are disrupted, as the low interest rates that made such otherwise unprofitable, longer term investment attractive disappear. As reality sets in, and as time preferences dominate the interest rates again (even central banks cannot keep asset valuations rising forever), projects become untenable and must be abandoned. Despite the illusory signs from the interest rate market, the economy cannot support all of the central bank-distorted capital structure, and the boom becomes visibly unsustainable.
“In short,” wrote Rothbard, “and this is a highly important point to grasp, the depression is the ‘recovery’ process, and the end of the depression heralds the return to normal, and to optimum efficiency. The depression, then, far from being an evil scourge, is the necessary and beneficial return of the economy to normal after the distortions imposed by the boom. The boom, then, requires a ‘bust.’”
Aggregate, correlated economic loss—the correlated entrepreneurial errors in the eyes of the Austrians—is not a random event, not bad luck, and not a tail. Rather, it is the result of distortions and imbalances in the aggregate capital structure which are untenable. When it comes to an end, by necessity, it does so ferociously due to the surprise by entrepreneurs across the economy as they discover that they have all committed investment errors. Rather than serving their homeostatic function of correcting market maladjustments back to the ERE, the market adjusts itself abruptly when they all liquidate.
What follows—to those who see only the “uniformity of nature”—is a dreaded tail event.
The paper continues - Read on below (full pdf)