Bridge over ocean
12 July 2018 Financial Analysts Journal Book Review

The End of Theory: Financial Crises, the Failure of Economics, and the Sweep of Human Interaction (a review)

  1. Robert N. Farago, ASIP
The author proposes agent-based modeling as a method for managing risk in a complex system, providing clear, non-technical explanations for the concepts of the “four horsemen” that together help us understand why traditional risk models fail in a crisis. He describes the complexity spectrum, ranging from deterministic mechanical systems to strategic complexity with its competing agents. This book is not a how-to guide but a thought-provoking challenge to the use of models in economics and, by extension, fund management.

The use of mathematical models in economics dates back to 1862, with the introduction of marginal utility by William Stanley Jevons. The misuse of financial models in economics followed soon after. Jevons was determined to link economic cycles to sunspots, but the failure of his exotic models had no consequences. In contrast, the repercussions of the failure of economic models to predict the global financial crisis can still be felt a decade later. Richard Bookstaber explores the limitations of our current risk models and proposes a new approach—agent-based modeling.

Richard Bookstaber is well placed to discuss the topic. He was the first risk manager at Morgan Stanley. Subsequently, he worked at Salomon Brothers, Moore Capital, Bridgewater, and the US Treasury. He currently manages risk at the investment office at the University of California, where he is putting the theories in this book into practice.

The first half of the book explains why the current approach cannot cope when a crisis hits. It introduces the “four horsemen” that explain why the financial system is too complex to model:

  • Computational irreducibility describes “a problem without mathematical shortcuts, where the only way to determine the outcome is to perform each step of the program.” Even problems with seemingly simple rules can prove to be computationally irreducible.

  • Emergent phenomena “occur when the overall effect of individuals’ actions is different from what the individuals are doing. The actions of the system differ from the actions of the agents that comprise the system.” The coordinated movement of a school of fish provides an elegant example of when this phenomenon works well. The deaths caused by stampedes in large crowds of people illustrate what can happen in a crisis.

  • Non-ergodicity is the opposite of ergodicity, which describes a process that does not vary with time or experience. “It follows the same probabilities today as it did in the distant past and will in the distant future.” Our world is shaped by the interaction of humans, who are not always rational decision makers. Our world is non-ergodic. In addition, our decisions alter the future, as explained in George Soros’s theory of reflexivity.

  • Radical uncertainty captures those outcomes that are not only unexpected but also unforecastable. These are the unknown unknowns, the black swans, the unpredictable future events that cannot be modeled. Radical uncertainty is the natural outcome when mathematical irreducibility, emergent phenomena, and non-ergodicity are combined.

How can we manage risk when faced with these four horsemen? The proposed approach replaces complex models with simple rules of thumb, or heuristics. Bookstaber turns to biology to provide examples. The cockroach is one of nature’s survivors. It has a single rule: “The cockroach simply scurries away when the little hairs on its legs vibrate from puffs of air, puffs that might signal an approaching predator, like you. That is all it does.” This is a sub-optimum solution in any one situation. Bookstaber argues that “being optimal in any given environment may not be optimal in the long run.” A simpler solution can prove more robust when a shock occurs.

How can heuristic measures be combined into a practical model? Bookstaber introduces agent-based modeling in the second half of the book. These models “allow for individuals who are plotting their own course, making adjustments along the way, and affecting the world and others through their actions.” They have five components, illustrated with examples from the global financial crisis:

  1. The agents, including the individual investment banks

  2. The environment, in which the cost of funding is key

  3. Heuristics, with leverage the most important single measure

  4. Interactions, a step-by-step analysis of what happens when the cost of funding increases

  5. Dynamics, capturing the impact of the interactions across the system

This breakdown advances risk analysis beyond scenario analysis, playing each scenario forward through time to identify the stress points that emerge.

Agent-based modeling is designed to deal with complexity. Bookstaber accordingly sets out the complexity spectrum. It “ranges from deterministic mechanical systems, to stochastic systems, to types of dynamic systems such as non-linear and adaptive, to those that introduce reflexivity, and finally to systems with strategic complexity, such as those dictated by warfare.” Fund management sits at the extreme end of this spectrum because of its competitive nature.

The story of the Notte Bianca, when the whole of Rome was plunged into darkness, illustrates the need for a new approach to complexity. The Italian electricity system was designed to cope with local failures. So too was the communication system. What engineers failed to take into account was the interaction between the two. Catastrophic failure occurred when attempts to reboot one system caused the other to fail, leading to a negative feedback loop. The financial system can be looked at through the same lens, with an asset layer, a funding layer, and a collateral layer. Here too, unforeseen interactions between the layers caused system-wide failure during the global financial crisis.

This is ultimately a book about financial crises. The message is that “once we are off the rails, we are facing a system that has uncoupled, where the approaches that work in normal times are not off by just a few degrees.” In a crisis, the story changes rapidly, and determining the appropriate response requires agility. John Maynard Keynes’s quote on economic models applies: Economics “is a science of thinking in terms of models joined to the art of choosing models which are relevant to the contemporary world.” For Bookstaber, “the basic message is that when there is a high degree of complexity, you have to figure it out as you go along.”

Bookstaber argues that agent-based modeling provides this agility. “We now have the machinery in place to understand the origins and course of many types of financial crises, namely, those that are based on the combined elements of leverage, concentration of assets, and growing illiquidity.”

The End of Theory is relevant to anyone working in the financial industry. Practitioners looking to implement agent-based models will have to look elsewhere for a practical guide, but all investment professionals will gain useful insights from this rewarding book.

We’re using cookies, but you can turn them off in Privacy Settings.  Otherwise, you are agreeing to our use of cookies.  Accepting cookies does not mean that we are collecting personal data. Learn more in our Privacy Policy.