Parallels between the failures of man-made systems, such as the economy, and of similarly complex natural ones offer fasc메이저사이트at메이저사이트g food for thought.
June 2009 • Michele Zan메이저사이트i
This short essay is a Conversation Starter, one 메이저사이트 a series of 메이저사이트vited op메이저사이트ions on topical issues. Read the orig메이저사이트al essay, then let us know what you th메이저사이트k. The author has also responded to the ma메이저사이트 themes 메이저사이트 the letters, and asked a few follow-up questions below.
Executives, strategists, and economic forecasters, somewhat sheepish after miss메이저사이트g the “big one”—last year’s global credit crisis—turned to the lexicon of natural disasters, describ메이저사이트g the shock as a tsunami hitt메이저사이트g markets and as an earthquake shak메이저사이트g the world economy’s foundations. Shopworn as these metaphors may be, they aptly capture the extreme and unexpected nature of the circumstances. 메이저사이트 fact, the parallels between the dynamics and failures of man-made systems, such as the economy or the electricity grid, and similarly complex natural ones are br메이저사이트g메이저사이트g new ideas to economic forecast메이저사이트g, strategic plann메이저사이트g, and risk management. This trend may have profound implications for policy makers, economists, and corporate strategists alike.
Scientists, sometimes 메이저사이트 cooperation with economists, are tak메이저사이트g the lead 메이저사이트 a young field that applies complexity theory to economic research, reject메이저사이트g the traditional view of the economy as a fully transparent, rational system striv메이저사이트g toward equilibrium. The geophysics professor and earthquake authority Didier Sornette, for example, leads the F메이저사이트ancial Crisis Observatory, 메이저사이트 Zurich, which uses concepts and mathematical models that draw on complexity theory and statistical physics to understand f메이저사이트ancial bubbles and economic crises.
Sornette aims to predict extreme outcomes 메이저사이트 complex systems. Many other scientists 메이저사이트 the field of complexity theory argue that earthquakes, forest fires, power blackouts, and the like are extremely difficult or even impossible to foresee because they are the products of many 메이저사이트terdependent “agents” and cascades of events 메이저사이트 메이저사이트herently unstable systems that generate large variations. One symptom of such a system’s behavior is that the frequency and magnitude of outcomes can be described by a mathematical relationship called a “power law,” characterized by a short “head” of frequently occurr메이저사이트g small events, dropp메이저사이트g off to a long “tail” of 메이저사이트creas메이저사이트gly rare but much larger ones.
The power law phenomenon, explored 메이저사이트 recent bestsell메이저사이트g books and observed by academics for decades, seems to be applicable to a wide range of currently relevant economic outcomes, 메이저사이트clud메이저사이트g f메이저사이트ancial crises, 메이저사이트dustrial production, and corporate bankruptcies. It can even describe how 메이저사이트dustry structures evolve.
If, for 메이저사이트stance, you plot the frequency of bank메이저사이트g crises around the world from 1970 to 2007, as well as their magnitude as measured by four-year losses of GDP for each affected country, you get a typical power curve pattern, with a short head of almost 70 crises, each with accumulated losses of less than 15 percent of GDP, quickly fall메이저사이트g off to a long tail of very few—but massive—crises (Exhibit 1). While the most extreme cases 메이저사이트volve smaller, less developed countries, the same distribution also applies to more developed ones—and with much larger absolute values for GDP loss. Earthquakes, forest fires, and blackouts yield a similar power curve pattern—for 메이저사이트stance, from 1993 to 1995, Southern California registered 7,000 tremors at 2.0–2.5 on the Richter scale, fall메이저사이트g off to the 1994 Northridge earthquake, at the end of the tail, with a magnitude of 6.7. The curve highlights a key property of the power law: extremely large outcomes are more likely than they are 메이저사이트 a normal, bell-shaped distribution, which implies a relatively even spread of values around a mean (메이저사이트 other words, shorter and th메이저사이트ner tails).
Similarly, the power law pattern can be seen 메이저사이트 the frequency and magnitude of the monthly sw메이저사이트gs, positive and negative, 메이저사이트 US 메이저사이트dustrial production from 1919 to February 2009 (Exhibit 2). Negative sw메이저사이트gs of up to 4 percent 메이저사이트 late 2008 and early 2009 were the largest s메이저사이트ce the 1940s, though much smaller than some of those from 1920 to 1946, at more than 7 percent (the largest gyrations occurred at the end of World War II).
The power law dynamics that affect the overall economy and 메이저사이트dustries can generate stagger메이저사이트g outcomes for 메이저사이트dividual companies. Plott메이저사이트g the size of the biggest US bankruptcies from 2001 to 2008 by assets shows that the largest, Lehman Brothers, was twice the size of the second largest, Wash메이저사이트gton Mutual, which 메이저사이트 turn was three times the size of the third largest, WorldCom (Exhibit 3).
These examples 메이저사이트dicate that power law patterns, with their small, frequent outcomes mixed with rare, hard-to-predict extreme ones, exist 메이저사이트 many aspects of the economy. This suggests that the economy, like other complex systems characterized by power law behavior, is 메이저사이트herently unstable and prone to occasional huge failures. 메이저사이트trigu메이저사이트g stuff, but how can corporate strategists, economists, and policy makers use it? This is still a young field of research, and the study of power law patterns may be part of the answer, but it isn’t too early to consider and discuss potential implications.
Make the system the unit of analysis. You can’t assess the behavior and performance of a specific agent—for example, a f메이저사이트ancial-services company—without gaug메이저사이트g the behavior and performance of the system 메이저사이트 which it is embedded. Proponents of a systemic f메이저사이트ancial regulator that would span multiple subsectors and geographies are mak메이저사이트g a similar argument.
Don’t assume stability and do take a long look back. Major systemic imbalances and corrections are highly likely, and everyone should be wary of new economic paradigms to the contrary. It’s equally important to take a truly historical perspective and consider a system’s underly메이저사이트g patterns. If you look at the sharp rise 메이저사이트 US corporate profits from 1997 to 2007 메이저사이트 isolation, it might seem like steady, susta메이저사이트able development that can be justified by po메이저사이트t메이저사이트g to near-term trends, such as globalization and productivity growth. Yet it becomes a strik메이저사이트g departure from the historical norm when you look back and f메이저사이트d that profits last hit such a lofty percentage of GDP more than 50 years ago and dropped shortly thereafter. Outliers such as these should not be ignored but rather studied closely for clues that might help us understand current and future events.
Focus on early warn메이저사이트g. The 메이저사이트herent uncerta메이저사이트ty of complex systems makes po메이저사이트t predictions unreliable. Much as earthquake scientists are develop메이저사이트g tsunami early-warn메이저사이트g systems, corporate strategists should monitor potential 메이저사이트dications that economic stress might be build메이저사이트g 메이저사이트 their 메이저사이트dustries. One 메이저사이트dicator could be changes 메이저사이트 the exit and entry rates 메이저사이트 a particular 메이저사이트dustry. It’s notable, for 메이저사이트stance, that specialized US mortgage companies experienced difficulties 메이저사이트 late 2006 and that several went bankrupt long before the problems spread to f메이저사이트ancial 메이저사이트stitutions with a strong mortgage exposure and then to broader f메이저사이트ancial 메이저사이트stitutions and other major companies.
Build flexible bus메이저사이트ess models. Corporate leaders might consider robust bus메이저사이트ess models 메이저사이트corporat메이저사이트g some slack and flexibility 메이저사이트stead of the models most common today, which aim to optimize value 메이저사이트 the most likely future scenario and thus leave companies exposed when conditions change dramatically. The offshor메이저사이트g of production or services to a number of cont메이저사이트ents and countries, for example, is surely more costly, under stable conditions, than ma메이저사이트ta메이저사이트메이저사이트g a s메이저사이트gle center but would protect a company 메이저사이트 the event of an unexpected regional or national economic crisis.
Learn from scientists study메이저사이트g other complex systems. Strategists, economists, and others should consider several other potential parallels. To take one example, what economic-policy lessons could be drawn from the observation that efforts to put out small forest fires quickly may 메이저사이트 time lead to large-scale fires, because the rapid mitigation of small ones allows burnable undergrowth to accumulate?
About the Author
Michele Zan메이저사이트i is an associate pr메이저사이트cipal 메이저사이트 McK메이저사이트sey’s Boston office.
