March 16, 2020

How Markets as Complex Adaptive Systems Process COVID-19

An institutional client asked at the close of last week what we observe in the market’s behavior – what is the market sensing? Bill Miller’s response reminds us that markets are complex adaptive systems, meaning the market as a whole is more complex and meaningful than just an aggregate of its individual components. We have long studied complex adaptive systems through our 25+ year affiliation with the Santa Fe Institute. Looking at the market’s reaction to the situation of COVID-19 in this context helps us understand and assess the dynamics.


This is one of those periodic storms that upend markets every so often (1974, 1987, 2008, etc). Their path, duration, and resolution depend on the causes and the response by authorities, although they share similarities.

Since the financial crisis of 2008-09, investors have been risk and volatility phobic, consistently overestimating real risk, leading to sharp, scary market declines: 2011, 2016, 2018 in response to events that conjured up visions of 2008, none of which came to pass.

This one, at this stage, is acting more like those in the first paragraph than those in the second, as the speed and depth of the current stock market’s declines have not been seen since 1987 and 2008. So the perceived economic risk is considerably greater than in the other post-crisis declines. The real risk is still unknown.

In broad terms, in this case, we have an exogenous event – the virus – initially causing a moderate shock to global aggregate supply by disrupting the flow of goods from China. That has now morphed into a major disruption in global aggregate demand as the virus spreads around the world and governments attempt to curtail its spread and its effect on public health by shutting schools, canceling all sorts of events, banning large gatherings of people, curtailing travel and so on. The impact could be large enough to knock the global economy into recession.

Markets are discounting mechanisms. They are engaged in real-time information processing, constantly adjusting to changing expectations. In the world of classical economics, they adjust instantaneously and smoothly to new information so as to create an equilibrium between supply and demand. Real markets are a long way from the dynamic stochastic general equilibrium (DSGE) models that dominate academic economics. Being complex adaptive systems, they are nonlinear and constantly changing as circumstances and conditions and information warrant, and those changes can be abrupt, violent, and frightening.

Since the potential outcomes at this early stage of the virus’s global spread span a wide range, from relatively benign to catastrophic (the 1918 flu pandemic), the economic outcomes have a similar spread, which the market must take into account. That is part of why volatility is so high; it reflects the uncertainty and potential impact of that range of outcomes. When the market thinks the authorities don’t get it — Trump’s speech on Wednesday last week — the market reflects that immediately. When it believes proactive measures are being taken — Trump’s remarks [Friday] — that is quickly evident in the market’s reaction.

The market typically overreacts to bad news. Not always, but mostly, and especially when the left tail contains really bad outcomes. In evolutionary theory, there is an understanding of the survival benefit of overestimating risk. A really good, but badly named, new book, The Power of Bad, by John Tierney and Roy Baumeister, covers the psychological literature on this topic. I highly recommend it.

A much more succinct answer to the question than the above is that markets have to price in all the things that can happen. Many more things can happen than will happen, and as those possible outcomes narrow, the market’s accuracy greatly improves.


Any views expressed are subject to change at any time, and Miller Value Partners disclaims any responsibility to update such views. The information presented should not be considered a recommendation to purchase or sell any security and should not be relied upon as investment advice. It should not be assumed that any purchase or sale decisions will be profitable or will equal the performance of any security mentioned. Past performance is no guarantee of future results, and there is no guarantee dividends will be paid or continued. Content may not be reprinted, republished or used in any manner without written consent from Miller Value Partners.

©2020 Miller Value Partners, LLC