Photo by Malcolm Ross

Black Swan Events

Nassim Nicholas Taleb is the author of the fascinating book “The Black Swan: The impact of the Highly Improbable”.  He is also a distinguished professor of risk engineering at the Polytechnic Institute of New York University and a former derivatives trader at CS-First Boston.

Drawing on an analogy with early European visitors to Australia who never imagined that a swan could be any colour but white, he developed the Black Swan concept in risk management.

Black Swans are events with three attributes:

1.      they lie outside the realm of regular expectations;

2.      they carry an extreme impact;

3.      human nature makes us concoct explanations for their occurrence after the fact, making them seem predictable – with the benefit of hindsight.

One good example of a black swan event was the near extinction of humans (and other species) about 70,000 years ago when Lake Toba in Sumatra was formed.  This was a mega-volcanic eruption which caused several years of winter globally as ash in the atmosphere obscured the sun.  This may well have been the stimulus for human migrations out of Africa at around that time, as people moved further than before in search of food.

Lake Toba would certainly have been a black swan event at the time.  It was outside the realm of expectations, it carried an extreme impact, and humans may have invented religion at the time to explain the event so as to make it seem predictable – God may destroy us for our sins.

But a future such mega-volcanic event would not really be a black swan event – we know it has happened in the past and that it will happen again.  Similarly, a major asteroid colliding with the earth and killing millions, even causing extinctions, would not be a black swan event.  We know it has happened before and will happen again.

I doubt if there is any such thing as a black swan event nowadays in the sense of lying outside the realm of regular expectations.  But there are many events which we really know will happen but which we are in denial about or simply fail to manage the associated risk.

The Global Financial Crisis of 2008/09 was predicted by a few people but many people including George W Bush and Alan Greenspan simply denied that it was possible.  With hindsight it was inevitable!  So too with the terrible terrorist attacks of September 11 2001 and the devastation of New Orleans by hurricane Katrina in 2005.  Both of these events were predicted but the risk was not managed.

Events that are nowadays dubbed black swan events are really just rare events which have devastating consequences because we have failed to manage risk.

Matthew Le Merle, writing in Issue 64, Autumn 2011, of strategy+business (published by booz & co.) suggests disrupter analysis to help assess the risks of future catastrophic events.  He says that enterprise risk management typically does not evaluate the implications of rare events.  He suggests creating a disrupter list – a list of potential black swan events and asking what would happen to the company if the events occurred.  This is rather like scenario planning.  He recommends that a very wide net be cast when cataloging potential catastrophic events.

I agree that a long list of rare but potentially catastrophic events be identified for such an analysis.  In doing so, we must avoid denial (as many do in the case of climate change) and hubris (a loss of contact with reality and over-estimation of our own abilities – as was the case with the “shock and awe” US invasion of Iraq).

Also, we must be careful not to under-estimate the likelihood of catastrophic events.  How often do we hear the excuse that so-and-so was a once in a century or millennium event?  This is particularly important for investors and stems from using the normal distribution for the occurrence of events rather than, say, the Cauchy distribution which has fatter tails.  In other words, rare events happen more frequently than we sometimes expect.

Charlie Nelson
October 2011

Additional content: January 2012

Scientific American, in the September 2010 edition, has an interesting analysis of the likelihood and severity of rare events.

Some of these could be described as black swans because we have little idea when they may occur; their impact would be devastating; there is little we can do to prevent them; and there is little we can do to survive.  These include:

  • a giant asteroid impact (one per hundred million years) which could kill one quarter of the human population and temporarily destroy civilization;
  • a supervolcano (one per hundred thousand years) which would significantly alter weather patterns for decades, leading to drought and famines;
  • a solar superstorm (one per 300 years) which would knock out power grids and communication systems.

Others could not be described as black swans because we can influence their likelihood and we can prepare plans to survive them if they happen.  These include:

  • nuclear war (one per 300 years) triggered either accidentally or by a rogue state, killing hundreds of millions of people.  More likely is a localized nuclear attack on a city by terrorists.
  • a killer pandemic (one per 60 years) with perhaps one hundred million deaths.
  • runaway global warming (a 50% chance in the next 200 years) which could raise sea levels by 12 metres and create hundreds of millions of refugees.