Sunday, May 06, 2007
When catastrophe strikes, blame a black swan
Posted by David Smith at 08:59 AM
Category: David Smith's other articles

A piece about The Black Swan: The Impact of the Highly Improbable, by Nassim Nicholas Taleb. You can buy the book here:

When, on July 7, 2005, four young British men blew themselves up on London’s transport network, killing 52 innocent people and injuring hundreds of others, should we have been fully prepared for it?

Or was this what Nassim Nicholas Taleb describes in his new book as a “black swan” event, something that was thought of as highly improbable beforehand but which we quickly rationalise afterwards as being entirely predictable, as MI5 is discovering to its cost?

Black swan events have fascinated philosophers for centuries. When the Dutch explorer Willem de Vlamingh discovered black swans on the west coast of Australia in 1697 and the news filtered back to Europe, it became an analogy for the things we can and cannot know with certainty.

As John Stuart Mill, the classical economist and philosopher, put it: “No amount of observations of white swans can allow the inference that all swans are white, but the observation of a single black swan is sufficient to refute that conclusion.”

Donald Rumsfeld is no Mill, but his most famous – and most ridiculed quote – was a modern variation on the theme, taken directly from Taleb’s earlier book, Fooled by Randomness.

“Reports that say that something hasn’t happened are always interesting to me, because as we know, there are known knowns; there are things we know we know,” the former US defence secretary famously said. “We also know there are known unknowns; that is to say we know there are some things we do not know. But there are also unknown unknowns – the ones we don’t know we don’t know.”

Black swan events, according to Taleb, are the unknown unknowns, the things we didn’t know we didn’t know until after they have happened. If the July 7 attacks were a true black swan event we should not seek to blame the security services or the police.

If 9/11 was a black swan event, as he argues [he does not touch directly on July 7], then we should forget all the congressional investigations and soul-searching about the US administration’s failure to prevent it. Failing to prevent another would indeed be an act of gross negligence, which is why he says New York was infinitely safer on September 12, 2001 than it was on September 10.

But the characteristic of a black swan event, that it is essentially an “out of the blue” occurrence, means that in practical terms some things cannot be prevented. Before September 11, given what was then a reasonable assessment of the risk of such an attack, nobody would have approved a 24-hour air force patrol in the skies over Manhattan. Even if there had been one, would anybody have given the order to shoot down a passenger plane, even one apparently bent on a destructive course? After 9/11, yes; before it, unlikely.

Niall Ferguson, the historian, suggested that the Virginia Tech massacre, when Cho Seung-Hui turned his guns on fellow students and teachers, killing more than 30 people, was a black swan event. Maybe, though plenty of people were worried enough about him to try to have him committed, and murderous school and campus rampages are not exactly unknown in America.

Where Ferguson was on safer ground, perhaps, was in agreeing with Taleb’s assertion that most history conforms to the black swan hypothesis. Historians spend their time constructing narratives to prove that events nobody foresaw followed a logical and entirely predictable course.

Hindsight is a wonderful thing. Economists are good at explaining market crashes, booms or recessions after the event. They are not nearly so good at telling you about them beforehand, as I would acknowledge. When I went to meet Taleb it was not until I was well on my way that I glanced at the publisher’s letter saying two of his pet hates were economists and people in blue suits. I am the nearest thing this newspaper has to an economist and, yes, I was wearing a blue suit.

Taleb is an engaging, lively and intelligent former Wall Street trader turned professor (in the sciences of uncertainty at the University of Massa-chusetts). His new book, The Black Swan: The Impact of the Highly Improbable (Allen Lane), is a New York Times bestseller and No 4 in the Amazon chart in Britain. It is funny, quirky and thought-provoking.

It appeals to common sense. Taleb talks of two worlds: Mediocristan, where everything is fairly predictable and single events do not change things much; and Extremistan, where the big, unexpected event changes everything. We know how much September 11 changed things. In business and finance, we know how much the rise of Google, largely unpredicted, has changed the landscape.

There are, however, a couple of things about the black swan hypothesis that suggest we should not be entirely swept away by it. The discovery of black swans turned what had been a common misconception on its head, but it simply replaced one certainty – all swans are white – with another – there are both black and white swans.

Before 9/11, or before July 7, nobody was saying there would never be terrorist attacks on New York, Washington and London. Indeed, most experts said there would be. New York had its own history; the World Trade Center attack of 1993. So did London, thanks to Irish terrorists. From September 11, 2001 until July 7, 2005, we heard little but warnings about the dangers of a large-scale terror attack on London. There was surprise in the timing and manner of the attack but not that it happened. Perhaps there was more surprise in that something similar was attempted two weeks later.

My argument would be that we predict too much, not too little. The world is full of predictions of big, epoch-changing events that never happen. When one does, the people who predicted it are not slow in coming forward – 9/11, or something like it, was predicted, but so were hundreds of other attacks, some even more devastating.

You could argue that the whole of the dotcom bubble of the 1990s was a prediction of the Google phenomenon. The problem was not that people failed to see a Google coming; it was that they saw too many. Faced with a plethora of predictions, the only thing we can do is try to form a reasonable view of the risk attached to them. Sometimes we will get it wrong; often we will get it right.

I am not saying we can predict everything. What I am saying is that it is sometimes too easy to throw up our hands, admit our limitations and put everything into the black swan category. Many of the biggest challenges of our time are based on big events that, according to the experts, are entirely predictable. Global warming, say scientists, will have predictable effects on the ability of the planet to sustain life. The black swan event, perhaps, would be if they turn out to be wrong.

From The Sunday Times, May 6 2007

Comments

I saw a review on the Guardian online and this section really stood out;

Taleb is a fan of the Polish-born French mathematician Benoit Mandelbrot, who gives short shrift to those who believe financial markets resemble a bell curve, with modest movements the norm and violent moves infinitesimally rare. Looking at the daily movements of the DJIA from 1916 to 2003, Mandelbrot said that according to the neat bell curve analysis, there should have been 58 days when the Dow moved more than 3.4%, when in fact there were 1,001.

Instead of just six days when there were movements of more than 4.5% there were 366. Only once in every 300,000 years should there have been a day when the Dow moved by 7% or more, but it happened 48 times. "Extreme price swings are the norm in financial markets - not aberrations that can be ignored. Price movements do not follow the well-mannered bell curve assumed by modern finance; they follow a more violent curve that makes an investor's ride much bumpier," Mandelbrot says in his book The (Mis)Behaviour of Markets (Profile Books). "A sound trading strategy would build this cold, hard fact into its foundations".

The argument that markets are efficient to a certain extent relies on them to be rational and the phenomenon that Mandelbrot was intrigued by gives some pause for thought about just how rational markets are.

I'm not sure that I'd agree that markets are necessarily efficient or rational; perhaps 'less inefficient than the alternatives' and 'mostly rational' in the same way as The Hitch Hikers Guid to the Galaxy describes Earth as 'mostly harmless'.

Posted by: Jonathan at May 6, 2007 12:57 PM

Taleb isn't saying we shouldn't predict, he's saying we should know the limitations of our predictions. For global warming the models will predict modest increases in temperature, there are scenarios however where feedback loops could drive extreme temperature changes that exterminate all life on earth.

We can't predict how likely that is, because our models will never be accurate enough and because we're straying into long tail, black swan unpredictability. So the best case is warming, the worst case is total annihilation.

This means there is no downside to acting now. Best case we avert the crisis, worst case we reduce the impact.

In fact there is considerable upside, because if there is a positive 'black swan' - a positive but unpredictable event - then we've reduced our reliance on fossil fuels, headed off peak oil and given ourselves cleaner air and lower impact on the environment. Bonus!

It's not about whether or not we can predict, it's about investing to protect against downside risk, while maximising upside risk.

Posted by: Nick Jenkins at May 20, 2011 04:39 AM
Post a comment









Remember personal info?








    •