The Black Swan of Cairo - How Suppressing Volatility Makes the World Less Predictable and More Dangerous
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
m a y / j u n e 2 o1 1 The Black Swan of Cairo How Suppressing Volatility Makes the World Less Predictable and More Dangerous Nassim Nicholas Taleb and Mark Blyth Volume 9o • Number 3 The contents of Foreign Affairs are copyrighted.©2o11 Council on Foreign Relations, Inc. All rights reserved. Reproduction and distribution of this material is permitted only with the express written consent of Foreign Affairs. Visit www.foreignaffairs.org/permissions for more information.
The Black Swan of Cairo How Suppressing Volatility Makes the World Less Predictable and More Dangerous Nassim Nicholas Taleb and Mark Blyth Why is surprise the permanent condition extremely fragile, while at the same time of the U.S. political and economic elite? In exhibiting no visible risks. In fact, they 2007–8, when the global financial system tend to be too calm and exhibit minimal imploded, the cry that no one could have variability as silent risks accumulate seen this coming was heard everywhere, beneath the surface. Although the stated despite the existence of numerous analyses intention of political leaders and economic showing that a crisis was unavoidable. It policymakers is to stabilize the system by is no surprise that one hears precisely the inhibiting fluctuations, the result tends same response today regarding the current to be the opposite. These artificially con- turmoil in the Middle East. The critical strained systems become prone to “Black issue in both cases is the artificial suppres- Swans”—that is, they become extremely sion of volatility—the ups and downs of vulnerable to large-scale events that lie far life—in the name of stability. It is both mis- from the statistical norm and were largely guided and dangerous to push unobserved unpredictable to a given set of observers. risks further into the statistical tails of the Such environments eventually experi- probability distribution of outcomes and ence massive blowups, catching everyone allow these high-impact, low-probability oª-guard and undoing years of stability “tail risks” to disappear from policymakers’ or, in some cases, ending up far worse than fields of observation. What the world is they were in their initial volatile state. witnessing in Tunisia, Egypt, and Libya Indeed, the longer it takes for the blowup is simply what happens when highly to occur, the worse the resulting harm in constrained systems explode. both economic and political systems. Complex systems that have artificially Seeking to restrict variability seems to suppressed volatility tend to become be good policy (who does not prefer stability Nassim Nicholas Taleb is Distinguished Professor of Risk Engineering at New York University’s Polytechnic Institute and the author of The Black Swan: The Impact of the Highly Improbable. Mark Blyth is Professor of International Political Economy at Brown University. [33]
Nassim Nicholas Taleb and Mark Blyth to chaos?), so it is with very good intentions have been good at fragilizing large corpora- that policymakers unwittingly increase the tions through bailouts, and Democrats have risk of major blowups. And it is the same been good at fragilizing the government. misperception of the properties of natural At the same time, the financial system as a systems that led to both the economic whole exhibited little volatility; it kept get- crisis of 2007–8 and the current turmoil in ting weaker while providing policymakers the Arab world. The policy implications with the illusion of stability, illustrated are identical: to make systems robust, all most notably when Ben Bernanke, who risks must be visible and out in the open— was then a member of the Board of Gover- fluctuat nec mergitur (it fluctuates but does nors of the U.S. Federal Reserve, declared not sink) goes the Latin saying. the era of “the great moderation” in 2004. Just as a robust economic system is one Putatively independent central that encourages early failures (the concepts bankers fell into the same trap. During of “fail small” and “fail fast”), the U.S. gov- the 1990s, U.S. Federal Reserve Chair ernment should stop supporting dictato- Alan Greenspan wanted to iron out the rial regimes for the sake of pseudostability economic cycle’s booms and busts, and he and instead allow political noise to rise to sought to control economic swings with the surface. Making an economy robust interest-rate reductions at the slightest in the face of business swings requires sign of a downward tick in the economic allowing risk to be visible; the same is data. Furthermore, he adapted his eco- true in politics. nomic policy to guarantee bank rescues, with implicit promises of a backstop—the SEDUCED BY STABILITY now infamous “Greenspan put.” These Both the recent financial crisis and the policies proved to have grave delayed side current political crisis in the Middle East eªects. Washington stabilized the market are grounded in the rise of complexity, with bailouts and by allowing certain com- interdependence, and unpredictability. panies to grow “too big to fail.” Because Policymakers in the United Kingdom and policymakers believed it was better to do the United States have long promoted something than to do nothing, they felt policies aimed at eliminating fluctuation— obligated to heal the economy rather than no more booms and busts in the economy, wait and see if it healed on its own. no more “Iranian surprises” in foreign The foreign policy equivalent is to policy. These policies have almost always support the incumbent no matter what. produced undesirable outcomes. For And just as banks took wild risks thanks to example, the U.S. banking system became Greenspan’s implicit insurance policy, client very fragile following a succession of pro- governments such as Hosni Mubarak’s in gressively larger bailouts and government Egypt for years engaged in overt plunder interventions, particularly after the 1983 thanks to similarly reliable U.S. support. rescue of major banks (ironically, by the Those who seek to prevent volatility on same Reagan administration that trum- the grounds that any and all bumps in the peted free markets). In the United States, road must be avoided paradoxically increase promoting these bad policies has been a the probability that a tail risk will cause a bipartisan eªort throughout. Republicans major explosion. Consider as a thought [34] fore ign affairs . Volume 90 No. 3
The Black Swan of Cairo experiment a man placed in an artificially magnitude of the hidden risks and a mis- sterilized environment for a decade and understanding of the statistical properties then invited to take a ride on a crowded of the system. subway; he would be expected to die What is needed is a system that can quickly. Likewise, preventing small forest prevent the harm done to citizens by the fires can cause larger forest fires to become dishonesty of business elites; the limited devastating. This property is shared by competence of forecasters, economists, and all complex systems. statisticians; and the imperfections of In the realm of economics, price con- regulation, not one that aims to eliminate trols are designed to constrain volatility on these flaws. Humans must try to resist the the grounds that stable prices are a good illusion of control: just as foreign policy thing. But although these controls might should be intelligence-proof (it should work in some rare situations, the long-term minimize its reliance on the competence eªect of any such system is an eventual of information-gathering organizations and extremely costly blowup whose cleanup and the predictions of “experts” in what costs can far exceed the benefits accrued. are inherently unpredictable domains), The risks of a dictatorship, no matter how the economy should be regulator-proof, seemingly stable, are no diªerent, in the given that some regulations simply make long run, from those of an artificially the system itself more fragile. Due to the controlled price. complexity of markets, intricate regulations Such attempts to institutionally engineer simply serve to generate fees for lawyers the world come in two types: those that and profits for sophisticated derivatives conform to the world as it is and those traders who can build complicated financial that attempt to reform the world. The products that skirt those regulations. nature of humans, quite reasonably, is to in- tervene in an eªort to alter their world and DON’T BE A TURKEY the outcomes it produces. But government The life of a turkey before Thanksgiving interventions are laden with unintended— is illustrative: the turkey is fed for 1,000 and unforeseen—consequences, particularly days and every day seems to confirm that in complex systems, so humans must work the farmer cares for it—until the last day, with nature by tolerating systems that when confidence is maximal. The “turkey absorb human imperfections rather than problem” occurs when a naive analysis of seek to change them. stability is derived from the absence of Take, for example, the recent celebrated past variations. Likewise, confidence in documentary on the financial crisis, Inside stability was maximal at the onset of the Job, which blames the crisis on the malfea- financial crisis in 2007. sance and dishonesty of bankers and the The turkey problem for humans is the incompetence of regulators. Although it result of mistaking one environment for is morally satisfying, the film naively over- another. Humans simultaneously inhabit looks the fact that humans have always been two systems: the linear and the complex. dishonest and regulators have always The linear domain is characterized by its been behind the curve. The only diªerence predictability and the low degree of this time around was the unprecedented interaction among its components, which fore ign affairs . May / June 2011 [35]
Nassim Nicholas Taleb and Mark Blyth allows the use of mathematical methods can predict a solar eclipse and the trajectory that make forecasts reliable. In complex of a space vessel, but not the stock market systems, there is an absence of visible or Egyptian political events. All man-made causal links between the elements, masking complex systems have commonalities and a high degree of interdependence and even universalities. Sadly, deceptive calm extremely low predictability. Nonlinear (followed by Black Swan surprises) seems elements are also present, such as those to be one of those properties. commonly known, and generally misun- derstood, as “tipping points.” Imagine THE ERROR OF PREDICTION someone who keeps adding sand to a sand As with a crumbling sand pile, it would be pile without any visible consequence, until foolish to attribute the collapse of a fragile suddenly the entire pile crumbles. It would bridge to the last truck that crossed it, and be foolish to blame the collapse on the last even more foolish to try to predict in grain of sand rather than the structure of the advance which truck might bring it down. pile, but that is what people do consistently, The system is responsible, not the compo- and that is the policy error. nents. But after the financial crisis of U.S. President Barack Obama may 2007–8, many people thought that predict- blame an intelligence failure for the gov- ing the subprime meltdown would have ernment’s not foreseeing the revolution helped. It would not have, since it was a in Egypt (just as former U.S. President symptom of the crisis, not its underlying Jimmy Carter blamed an intelligence cause. Likewise, Obama’s blaming “bad in- failure for his administration’s not fore- telligence” for his administration’s failure to seeing the 1979 Islamic Revolution in Iran), predict the crisis in Egypt is symptomatic but it is the suppressed risk in the statis- of both the misunderstanding of complex tical tails that matters—not the failure to systems and the bad policies involved. see the last grain of sand. As a result of Obama’s mistake illustrates the illusion complicated interdependence and conta- of local causal chains—that is, confusing gion eªects, in all man-made complex catalysts for causes and assuming that one systems, a small number of possible events can know which catalyst will produce which dominate, namely, Black Swans. eªect. The final episode of the upheaval in Engineering, architecture, astronomy, Egypt was unpredictable for all observers, most of physics, and much of common especially those involved. As such, blam- science are linear domains. The complex ing the cia is as foolish as funding it to domain is the realm of the social world, forecast such events. Governments are epidemics, and economics. Crucially, the wasting billions of dollars on attempting linear domain delivers mild variations to predict events that are produced by without large shocks, whereas the complex interdependent systems and are therefore domain delivers massive jumps and gaps. not statistically understandable at the Complex systems are misunderstood, individual level. mostly because humans’ sophistication, As Mark Abdollahian of Sentia Group, obtained over the history of human knowl- one of the contractors who sell predictive edge in the linear domain, does not transfer analytics to the U.S. government, noted properly to the complex domain. Humans regarding Egypt, policymakers should [36] fore ign affairs . Volume 90 No. 3
“think of this like Las Vegas. In blackjack, focus. It is telling that the intelligence if you can do four percent better than the analysts made the same mistake as the average, you’re making real money.” But risk-management systems that failed to the analogy is spurious. There is no “four predict the economic crisis—and oªered percent better” on Egypt. This is not just the exact same excuses when they failed. money wasted but the construction of a Political and economic “tail events” are false confidence based on an erroneous unpredictable, and their probabilities are fore ign affairs . May / June 2011 [37]
Nassim Nicholas Taleb and Mark Blyth not scientifically measurable. No matter tend to be comparatively mild. For exam- how many dollars are spent on research, ple, a shift in the ruling coalition from predicting revolutions is not the same as Christian parties to Hezbollah is not such counting cards; humans will never be able a consequential jump in terms of the to turn politics into the tractable random- country’s economic and political stability. ness of blackjack. Switching equilibrium, with control of Most explanations being oªered for the the government changing from one party current turmoil in the Middle East follow to another, in such systems acts as a shock the “catalysts as causes” confusion. The absorber. Since a single party cannot have riots in Tunisia and Egypt were initially total and more than temporary control, the attributed to rising commodity prices, not possibility of a large jump in the regime to stifling and unpopular dictatorships. type is constrained. But Bahrain and Libya are countries with In contrast, consider Iran and Iraq. high gdps that can aªord to import grain Mohammad Reza Shah Pahlavi and Sad- and other commodities. Again, the focus is dam Hussein both constrained volatility wrong even if the logic is comforting. It is by any means necessary. In Iran, when the system and its fragility, not events, the shah was toppled, the shift of power to that must be studied—what physicists call Ayatollah Ruhollah Khomeini was a huge, “percolation theory,” in which the proper- unforeseeable jump. After the fact, analysts ties of the terrain are studied rather than could construct convincing accounts about those of a single element of the terrain. how killing Iranian Communists, driving When dealing with a system that is the left into exile, demobilizing the demo- inherently unpredictable, what should be cratic opposition, and driving all dissent done? Diªerentiating between two types into the mosque had made Khomeini’s of countries is useful. In the first, changes rise inevitable. In Iraq, the United States in government do not lead to meaningful removed the lid and was actually surprised diªerences in political outcomes (since to find that the regime did not jump from political tensions are out in the open). hyperconstraint to something like France. In the second type, changes in govern- But this was impossible to predict ahead ment lead to both drastic and deeply of time due to the nature of the system unpredictable changes. itself. What can be said, however, is that Consider that Italy, with its much- the more constrained the volatility, the maligned “cabinet instability,” is economi- bigger the regime jump is likely to be. From cally and politically stable despite having the French Revolution to the triumph of had more than 60 governments since the Bolsheviks, history is replete with World War II (indeed, one may say Italy’s such examples, and yet somehow humans stability is because of these switches of remain unable to process what they mean. government). Similarly, in spite of consis- tently bad press, Lebanon is a relatively THE FEAR OF RANDOMNESS safe bet in terms of how far governments Humans fear randomness—a healthy can jump from equilibrium; in spite of ancestral trait inherited from a diªerent all the noise, shifting alliances, and street environment. Whereas in the past, which protests, changes in government there was a more linear world, this trait enhanced [38] fore ign affairs . Volume 90 No. 3
The Black Swan of Cairo fitness and increased chances of survival, imposition of peace through repeated it can have the reverse eªect in today’s punishment lies at the heart of many complex world, making volatility take the seemingly intractable conflicts, including shape of nasty Black Swans hiding behind the Israeli-Palestinian stalemate. Further- deceptive periods of “great moderation.” more, dealing with seemingly reliable This is not to say that any and all volatility high-level o⁄cials rather than the people should be embraced. Insurance should themselves prevents any peace treaty signed not be banned, for example. from being robust. The Romans were wise But alongside the “catalysts as causes” enough to know that only a free man under confusion sit two mental biases: the illusion Roman law could be trusted to engage in a of control and the action bias (the illusion contract; by extension, only a free people that doing something is always better than can be trusted to abide by a treaty. Treaties doing nothing). This leads to the desire to that are negotiated with the consent of a impose man-made solutions. Greenspan’s broad swath of the populations on both actions were harmful, but it would have sides of a conflict tend to survive. Just as no been hard to justify inaction in a democracy central bank is powerful enough to dictate where the incentive is to always promise a stability, no superpower can be powerful better outcome than the other guy, regard- enough to guarantee solid peace alone. less of the actual, delayed cost. U.S. policy toward the Middle East has Variation is information. When there historically, and especially since 9/11, been is no variation, there is no information. unduly focused on the repression of any This explains the cia’s failure to predict and all political fluctuations in the name the Egyptian revolution and, a generation of preventing “Islamic fundamentalism”— before, the Iranian Revolution—in both a trope that Mubarak repeated until his cases, the revolutionaries themselves did last moments in power and that Libyan not have a clear idea of their relative leader Muammar al-Qaddafi continues strength with respect to the regime they to emphasize today, blaming Osama were hoping to topple. So rather than sub- bin Laden for what has befallen him. This sidize and praise as a “force for stability” is wrong. The West and its autocratic Arab every tin-pot potentate on the planet, allies have strengthened Islamic funda- the U.S. government should encourage mentalists by forcing them underground, countries to let information flow upward and even more so by killing them. through the transparency that comes with As Jean-Jacques Rousseau put it, “A political agitation. It should not fear fluc- little bit of agitation gives motivation to tuations per se, since allowing them to be the soul, and what really makes the species in the open, as Italy and Lebanon both prosper is not peace so much as freedom.” show in diªerent ways, creates the stability With freedom comes some unpredictable of small jumps. fluctuation. This is one of life’s packages: As Seneca wrote in De clementia, there is no freedom without noise—and “Repeated punishment, while it crushes no stability without volatility.∂ the hatred of a few, stirs the hatred of all . . . just as trees that have been trimmed throw out again countless branches.” The fore ign affairs . May / June 2011 [39]
You can also read