It has become highly fashionable to talk about “black swans” since the crisis began in 2008 and now even Scott Sumner talks about it in his recent post “Don’t forget about those black swans”. Ok, Scott is actually not obsessed with black swans, but his headline reminded me how much focus there is on “black swans” these days – especially among central bankers and regulators and to some extent also among market participants.
What is a black swan? The black swan theory was popularized by Nassim Taleb in 2007 book “The Black Swan”. Taleb’s idea basically is that the financial markets underprice the risk of extreme events happening. Taleb obviously felt vindicated when crisis hit in 2008. The extreme event happened and it had clearly not been priced by the market in advance.
Lets go back to back to Scott’s post. Here he quotes Matt Yglesias:
Here’s a fun Intrade price anomaly that showed up this morning. The markets indicate that there’s more than a 3 percent chance that neither Barack Obama nor Mitt Romney will win the presidential election. That’s clearly way too high.
Scott then counter Matt by saying that we should not forget “something unusual happening”:
1. One of the two major candidates is assassinated, and the replacement is elected (as in Mexico’s 1994 election.)
2. Ditto, except one pulls out due to health problems, or scandal.
3. A third party candidate comes out of nowhere to get elected.
Scott is of course right. All this could happen and as a consequence it would obviously be wrong if the market had price a 100% chance that nobody else than Obama or Romney would become US president.
Scott and I tend to think that financial markets are (more or less) efficient and as a consequence we would not be gambling men. Scott nonetheless seem to think that the odds are good:
“But 3% is low odds. It’s basically saying once in ever 130 years you’d expect something really weird to happen in US presidential politics during an election year. That’s a long time! Given all the weird things that have happened, how unlikely is it? Some might counter that none of the three scenarios I’ve outlined have occurred in the US during an election year (my history is weak so I’m not certain.) But mind-bogglingly unusual things have happened on occasion. On November 10, 1972, what kind of odds would Intrade have given on neither Nixon nor Agnew being President on January 1 1975?”
Scott certainly have a good point, but I will not question the market on this one. The market pricing is the best assessment of risk we have. If not there would be money in street ready to pick up for anybody – and there is not. Obviously Taleb disagrees as he believe that markets tend to underprice risk. However, I fundamentally think that Taleb is wrong and I don’t see much evidence that market underprice black swan events. The fact that rare events happen is not evidence that the market on average underprice the likelihood of this events.
The fashion long-shot bias and central banks
In the evidence from betting markets seem to indicate that if anything bettors tend to have a favourite-longshot bias meaning that they tend to overprice the likelihood that the favourite will loose elections or sport games. Said in another way if anything bettors tend to overprice the likelihood of a black swan events. I happen to think that this is not a market problem in markets in general, but it nonetheless indicates that if anything the problem is too much focus on black swan events rather than too little focus on them.
This to a very large extent has been the case of the past 4 years – especially in regard to central banking and banking regulation. There seem to be a near-obsession among some policy makers that a new black swan could turn up. How often have we heard the talk about the major risk of bubbles if interest rates are kept too low too long and most of the new financial regulation being push through across the world these days is justified by reference to the risk of some kind of black swan event.
Media and policy makers in my view have become obsessed with extreme events happening – you will be reminded about that every time you go through the security check in any airport in the world.
The obsession with black swan events is highly problematic as the cost of policy makers obsessing about very unlikely events happening lead them to implement very costly regulation that lead to massive waste of resources. Again just think about how many hours you have spend waiting to get through airport security over the last couple of years and if you think that is bad just think of the cost resulting from excessive new regulation of the global financial markets. So my suggestion is clearly to forget about those black swans!
Finally three book recommendations:
Risk by Dan Gardner that tells about the “politics of fear” (of black swans).
The Myth of the Rational Voter by Bryan Caplan explains why democracy tend to lead to irrationality while market lead to rationality. Said in another way policy makers would be more prone to focus on black swans than market participants in free markets would be.
Risk, Uncertain and Profit – Frank Knight’s classic. If you are really interested in the issues of risk and uncertainty then there is no reason at all to read Taleb’s books (I have read both The Black Swan and Fooled by Randomness – they are “fun” and something you can read while you are waitin in line at the security check in the airport, but it is certainly not Nobel prize material). Instead just read Knight’s classic. It is much more insightful. It is actually something that frustrates me a great deal about Taleb’s books – there is really nothing new in what he is saying, but he claims to have come up with everything himself.
The Market Monetarist 
David Eagle
/ May 5, 2012Lars, I disagree. I think we should think about Black Swans. Remember that incident at the 2005 WEAI meetings when Milton Friedman scolded me for asking a hypothetical question about what monetary policy should do if aggregate supply drops by 10%. Perhaps Friedman agreed with you that this was a Black Swan and we should not ask questions about Black Swans. I disagree and so does Friedman’s former student, Robert Barro. Immediately after Friedman’s session in 2005 was the Honorary WEAI Presidential Luncheon and Address. Barro was the Honorary WEAI President. His address was about drops of aggregate supply by more than 20%. Hence, Barro was talking about blacker swans than I was talking about.
I believe that our economic systems must be designed to efficiently deal with Black Swans. You have noted my work with Arrow-Debreu economies. The reason that Arrow-Debreu economies are Pareto efficient is that they have securities that handle all possible contingencies including Black Swans. What I discovered is that if we do not have state-contingent securities, then we can still efficiently handle all the possible contingencies involving aggregate output (at least for people with average relative risk aversion) by one of the following: (i) nominal GDP targeting, or (ii) quasi-real indexing. Think about Black Swan events like a large asteroid hitting the earth or mega volcanos that spew ashes into the air that blocks the sun for years . These Black-Swan events are ones that Dale Domian and I used to motivate our paper, “Sounding the Alarm on Inflation Targeting and Indexing” (http://ideas.repec.org/p/wpa/wuwpma/0312010.html). That paper is an attack on inflation targeting and conventional inflation indexing and argues that, to achieve Pareto efficiency, we need to target instead nominal GDP (or use quasi-real indexing).
As a result, I think Market Monetarists should be arguing in favor of preparing our economies to deal with Black Swan events rather than advocating that we bury our heads in the sand.
Benjamin Cole
/ May 7, 2012Nice blogging of late. The chart on Irish deflation is shocking.
Do markets underprice risk? I think this is a tough question (for macroeconomics) as we don’t have enough events to know.
The Dow Jones Industrial Average, going into 2008, was still below where it was in 1999. PE’s came down a lot in that time frame. So uncertainty was being priced in.
It may be that US investors, like those of Japan, will underprice due to perceived risk, going forward. Why buy equity or property if they are so soft? Why buy an asset that can languish for decades at a stretch?
The Fed, ECB and BoJ are breeding self-fulfilling doom….
Yichuan Wang
/ May 13, 2012I think there is a confusion over what kinds of Black Swans we should be worrying ourselves about. Taleb doesn’t advocate heightened layers of complexity to address Black Swans. Rather, society needs to be simplified to the level of heuristics and principles.
At its core, the Black Swan is more an epistemological argument about how we don’t know what can possibly happen. This actually implies the exact opposite of policy fear-mongering. Instead of trying to get Dodd-Frank through to catch every possible dangerous trade, the financial ecology should be designed to prevent single trades from destroying markets.
And I don’t think the favorite/long-shot argument is that convincing in the world of finance. Given finance is so much more driven by statistical algorithms that assume thin tails and low variability instead of higher variability, then there’s a much higher risk of mispricing long-shots in finance. Additionally, the horse-race example is a classic instance of the ludic fallacy, in which we try to use games to approximate the real world. Horse-racing can’t get that extreme; a horse can get 1st or not get 1st. But financial markets have no upper or lower bound on how well or how badly they do, giving them a much different qualitative properties.