Dave Pollard's environmental philosophy, creative works, business papers and essays.
In search of a better way to live and make a living, and a better understanding of how the world really works.

	Wednesday, February 18, 2009

Fooled by Randomness, Risk and Opportunity Nassim Taleb is a very bright guy. He has some strange social and communication skills (he savagely ridicules those he thinks exemplify foolish qualities, he's annoyingly arrogant, and he comes across as a bit scatterbrained both in TV interviews and in some of his writing). Because of this, it's a bit surprising that his work (The Black Swan, Fooled by Randomness) has become so popular. But, given that he "predicted" the financial crisis, and made a lot of money by betting on it (or, more accurately, betting against it not occurring), he's recently become a celebrity and is being treated as a business guru (he was very visible at Davos this year, where the charts above and below were presented). His books present some interesting, and truly original, thinking. One of my current work projects is to get business leaders to think of environmental sustainability as congruent in the long term with business sustainability, and to think of both in the context of risk management. I recently presented a paper on this subject that I co-authored, in London at a Prince's Trust event on sustainability. It argues that the risk management models currently used in business need to be enhanced to consider: the very high degree of uncertainty of some risks (i.e. on the likelihood/severity matrix above, the positioning of these risks cannot be pinpointed)  the high degree of interrelationship of many risks (in the chart below, some of these interrelationships are plotted; the light green one at the very top is water scarcity (if you want to see this more legibly it's on page 4 of this report)) the fact that responsibility for many risks, especially the more external and uncontrollable risks (e.g. climate change risks), is not specifically assigned to anyone in organizations, and hence tends to fall on the shoulders of the CEO and Board (who generally don't know what to do about them), adaptation as an important risk response strategy (most risk management models consider avoidance and mitigation the only viable response strategies), and opportunity as the (asymmetric) flip-side of risk, something that should always be considered whenever risk is being considered. Taleb's key argument fits well with the above ideas. His thesis for the book is: We favor the visible, the embedded, the personal, the narrated, and the tangible; we scorn the abstract. Everything good (e.g. aesthetics, ethics) and wrong (e.g. being fooled by randomness) with us seems to flow from this. We are wired, he argues, to deal with immediate emergencies (fight or flight, when being pursued by predators), and to optimize the likelihood of procreation. Because of this, our brains, emotions and instincts can be "fooled". Several of these types of foolishness are now getting our species into deep trouble: Hindsight bias: Past events look more logical, causal and correlated than they really are. We therefore reward people (executives, politicians, celebrities, gamblers) for being in the right place at the right time, and don't see that their "success" had nothing to do with them; it was almost entirely good luck.  Learned helplessness: As Gladwell has argued, because of bad and misinterpreted information, we fear the wrong things. We think terrorism is a bigger threat than flying, and think SUVs are safer than convertibles. And we act (inappropriately) accordingly. Being swayed by simplifying rhetoric: We love (and believe) proverbs and terse oversimplifications. In an increasingly complex world, our minds crave simplification, and are suckers for it. Vulnerability to "noise": We are such pattern-seekers that we look for sense in the firehose of information, and find it when it isn't there. We would be better off to turn off the mainstream media, stop paying attention to them, and spend our time in original thought and reflection instead. The induction problem: The past always seems deterministic. We expect the future to be a continuation of the recent past. We "forget" (or never learn) the longer-term historical context. Because we have only ever seen white swans (boom economic times) we don't believe in the possibility of black swans (crashes). So we constantly overreact to the "latest" news (quarterly earnings reports). The key to resilience is to anticipate the possibility of black swans (unlikely but potentially catastrophic events) and to hedge against them, avoid them entirely (become a dentist, not a stockbroker) or be prepared for them. Taleb made his mark by hedging against a market collapse, but the challenge is that some such events are very difficult to hedge against or prepare for, so we tend to shrug them off until they actually occur. The skewness problem: Despite the fact that rare, catastrophic events can more than undo decades of positives, we tend to perceive 'blips' as being of the same magnitude in both directions, so we continue to bet on the positives and, on net, lose. The survivorship bias: We tend to see only the winners/survivors and lose track of the large numbers of losers. There are probably hundreds of people as bright and skilled as Bill Gates or Warren Buffett or Wayne Gretzky or (name your star) who followed exactly the same method or philosophy as these successes, but who, because they were not in the right place at the right time, failed and are now unknown. We are therefore foolish to follow the lead of such successes and expect the same result. When it comes to success in life and other complex situations, the cream actually rises to the top very rarely. This problem is worst at the top levels, because executives and politicians tend to make decisions that are more affected by external events, so the outcome, good or bad, is least likely to be the result of their decision. The anchoring bias: Our perception of where we are is biased by where we have come from. If we have a million dollars after living our lives poor, we think we are rich. But if we once had two million, we feel poor, a failure. Because of these biases, we are, he argues, very poor at assessing risks (both their likelihood and severity). I would say (as someone who has struggled with large organizations that have a strong, unacknowledged bias against true innovation) we are equally poor at assessing opportunities (both their likelihood of success and their consequences if they do succeed), because of these same biases. So, looking at the risks in the top chart above, Taleb would probably say that (1) we are probably underestimating the consequences of many of the risks on the left (low perceived likelihood) side of this chart, (2) we are missing a raft of risks on this left side that we have forgotten can occur or can't even imagine occurring, and (3) the combined probability of at least one (and probably several, possibly interdependent) of these 'individually low-likelihood' risks occurring is very high, and that occurrence, far more than the higher-probability "known" risks on the right side of the chart, is what we should really be considering, and preparing for. This is at least easier when we know what those risks are. We can anticipate the consequences of another disastrous war, next time in Iran or North Korea (and Obama's decision to escalate the war in Afghanistan today signals that he is also incapable of learning the lessons of history, which is an ominous sign). We know, from the sad lesson of Katrina, that the consequence of natural catastrophes in the 21st century will be to abandon afflicted cities to die (we simply cannot afford to rebuild them), just as we abandon old buildings and factories. From the 1970s and 2008, we have an inkling of the consequence of huge oil price spikes (though the gnomes of Davos still cannot get themselves to acknowledge that the real risk is not a price spike, but the end of oil as the engine of our economy). From the great blackouts we've experienced, we're reluctantly aware that the decaying and neglected infrastructure in our cities everywhere is going to cause us enormous problems, but because we consider it (for now) a low-likelihood catastrophe (and because we can't afford to fix it) we just put it out of our minds. Same thing with pandemics and water crises: we know they're coming, and that they will both cause a horrific economic downturn (and the indirect economic consequences will probably kill more people than the diseases and droughts will kill directly), but because they're still 'unlikely' in any year (and hence 'unlikely' to occur in the 10-year horizon of the charts above), we do nothing. Another real problem is all the risks that are not even on the chart. What if the real political crisis is not war in Iran, but the collapse of Mexico? There are plenty of warning signs for this, but we haven't even begun to consider what happens when organized crime takes over an anarchic state right beside us, and fifty million people seek asylum elsewhere. What if the real terrorism risk in not an 'international' threat but a bunch of whacked-out individuals who manage to produce (not as hard as you might think) weaponized anthrax and use it as a carrier for smallpox? What about a good old fashioned nuclear war between India and Pakistan? I can imagine dozens of risks, some of which have a long history of occurring but not recently (think Mao and ask why a populist coup in China is not on the risk list above), that belong in the upper left corner of this chart. They are all perhaps 'unlikely', but taken together, their probability is as high as the probability of an attack on the US was a year before 9/11, and their consequence is likely to be much greater. My pick for 'breakout' risk of the year? Food crisis (notice they call it "food price volatility": the gnomes can't quite get themselves to use the real 'f' word famine). It's in the upper left (#1) but there's lots of evidence it should be in the upper right. Unless they've read something about history, people think famine is something that only happens in Africa and Asia. But then, last year the gnomes only gave "asset price collapse" (their euphamism for global depression) a 20% chance of occurring in the next decade, and still don't think that it's much more likely than that. Another weakness in our analysis of risks is that we tend to view them all as temporary 'events' that need to be mitigated and survived, until things "return to normal". But just as some innovations (what Christensen and Raynor call "disruptive innovations") permanently change the business landscape, some risks (climate change, the end of oil) will, when they occur, usher in permanent structural changes in our world and in our economic and political systems. That's something Taleb doesn't deal with, but which I hope to continue to write about. The real value of scenario planning, simulations and other adaptation risk response strategies is not so much that they help us anticipate system shocks (though they can do that), as that they help us prepare for permanent shifts in our world, and help us learn to cope with complexity. PS: Taleb's advice is to shun the mainstream media, avoid self-help books and advice that would presume to make us who we are not, never complain (it's no one's "fault" so there is no point), never take compliments or harsh criticism too seriously (they say more about the speaker than about you), avoid superstition, never gamble, always be skeptical of apparent causality, and try to avoid path-dependent decisions (those you make unaware of your anchoring bias, described above). He also suggests not scheduling your life tightly, since he's observed that "time optimizers" are generally more unhappy because they set up more opportunities for "failure". I actually find his advice (despite his warning against taking advice from anyone) more interesting than a lot of his explanation of how we are fooled by (im)probabilities. Category: Complexity 4:44:34 PM  trackback []  comment []

© Copyright 2009 Dave Pollard. Last update: 3/2/09; 11:26:57 PM.