Markets have had a hectic week. As you read these words, the U.S. stock market may be in freefall, or it may be in a sharp recovery (and by the time you finish reading, it could well be moving in the other direction.) Little more than a week ago, the VIX was below 14; on Monday, it briefly rose above 45¹. How are we supposed to make sense of such behavior?
Different type of uncertainty; different behavior; different math
Part of the reason that commentators find it so difficult to explain these occasional bouts of extreme market turbulence is that investors have been conditioned to think of markets (and economic systems in general) as analagous to physical systems. Random walks; Black-Scholes; utility maximization; modern portfolio theory: all of these common models of market behavior derive from the mathematics of physics.
However, in a 2010 paper called “WARNING: Physics Envy May Be Hazardous To Your Wealth!” Andrew Lo and Mark Mueller of MIT argue that “As impressive as the achievements of modern physics are, physical systems are inherently simpler and more stable than economic systems, hence deduction based on a few fundamental postulates is likely to be more successful in the former case than in the latter.” By which they mean: economies and market are not physical systems, and they cannot be explained using the same math.
This argument is based on the nature of the uncertainty in each case. In their paper, Lo and Mueller develop a hierarchy of five levels of uncertainty. Some readers will recognize this as an extension of Frank Knight’s famous differentiation between risk (by which he meant risk that is measurable) and uncertainty (which is not.) In Lo and Mueller’s extended hierarchy, uncertainty in physical systems is described as “fully reducible” – which means that if you have enough data, then it’s possible to accurately model the randomness within the system. But the uncertainty in economic systems is not like that: there’s a limit to how much we can work out about what’s driving the behavior.
They go on to observe that “the failure of quantitative models in economics is almost always the result of a mismatch between the type of uncertainty in effect and the methods used to manage it”.
Ray Dalio has the wrong analogy
Even some of the most successful investors are prone to thinking in terms of physics. Ray Dalio—founder and CIO of Bridgewater Associates—has produced a video that opens with the statement “The economy works like a simple machine”. There’s no disputing Ray’s track record as an investor and a businessman. But I think he’s picked the wrong analogy. A machine is exactly what an economy isn’t.
An economy is driven by billions of individual decisions. And while the individual decisions may be simple enough, the way that they interact with one another results in system-level behavior that is far from a straightforward mapping of the individual actions. The classic example of this “more is different” phenomenon is the ant colony: the basic insight being that even though ant behavior is driven by a few simple rules, and the ant colony behavior is an aggregation of the individual behavior, the colony itself is a highly sophisticated entity. Examples abound: a city as the emergent product of millions of people; the human brain as the highly complex and sophisticated product of the interaction of neurons. We don’t think of these as machines.
And when investment markets start behaving as if they have a life of their own, the shortcomings of the “simple machine” analogy become clear. But stuff like this happens all the time in nature. As Lo and Mueller put it: “model-building in the social sciences should be much less informed by mathematical aesthetics, and much more by pragmatism in the face of partially reducible uncertainty.” Translation: if you want to make sense of the turbulent markets of the past few days, don’t ask a physicist; ask a biologist.