John Quiggin picks up on something that has been bugging me for a while. The idea of the GFC (Global Financial Crisis) as a ‘black swan event’ had some appeal to me. Philosophers have typically approached the problem of induction as a universal one: how do you move from individual observations to general understandings about the operations of complex processes (facts, if you will)? But it’s also historically specific. Ian Hacking has provocatively suggested that induction arose as a philosophical problem because of the nascent growth of commercial activity in 18th Century England, leading David Hume to question the profusion of facts around him – of which the accountants’ books were the epitome.
Calculability, then, has been central to most responses to the problem of induction. No matter how many White Swans you observe, Karl Popper famously argued, you could not make any meaningful claims about the population of swans because a Black Swan would falsify (( before you jump in there, Sociologists of Science have shown that the solution of falsification has just the same problems as verificationism )) your whole situation. or something. Most recently, Nicholas Nassim Taleb has run with the metaphor of Black Swans in an interesting but insufferably self-righteous book of the same name. The basic argument runs that people (traders, academics, everyone) get too caught up in the elegance of their own cognitive schemas, invariably based on the clever inferences from observations, to see where and how a big hit comes. Taleb bases this on a deep rooted narcissism that stops people engaging critically with data that doesn’t fit their models. Quiggin outlines how this more or less happened with the current crisis viz. who was modelling housing securities, which is more a problem of institutional configuration than inherent myopia, but anyway.
So I was intruiged to learn that Taleb himself had been effectively taking bets on the kind of systemic failure he saw as integral to the system:
Last month as the Dow dropped 20% and portfolios were battered, one investor was prepared. Nassim Nicholas Taleb, who founded the risk management firm Universa Investments, watched his portfolio soar 65% and 115%… For the past year, Universa has been betting that the market would take a radical turn for the worse.
Universa keeps 90% of its assets in cash or cash equivalents and simply tries to break even as it places small bets on rare events.
Five weeks ago when the S&P 500 was trading around 1200, Universa bought S&P 500 Index put options with a strike price of 850, due to expire late October. They were betting an “unlikely” drop would occur.They paid around 90 cents for those options. By Oct. 10, the S&P had dropped 300 points in a month. The options Universa purchased for 90 cents were now trading for $60 each. Universa cashed out of its position around $50, good for a gain of 5455%.
Universa also paid $1.29 for a put option on the insurance company AIG. They’d be in the money if AIG dipped below $25 a share by September. AIG imploded in a tangle of bad debt and Universa sold the AIG put options for $21 apiece.
Needless to say, figuring out that these events were worth the bet takes skill, determination, a Platonic schema and … uh oh, we’re back to those characteristics that in the nasty, bad, re-education column in Taleb’s book…
Elsewhere: Daniel Davies on the Politics of Classification
An interesting post, dk.au.
The big question is how you take advantage of ‘Black Swan’ events in a way which is distinct from gambling. After all, I could take my life savings and ‘invest’ them all on number 13 on the roulette table. I’ll be significantly wealthier if my Swan comes in, but no one would call that a sensible investment strategy.
Perhaps Taleb is just relying on put options – particularly those with a large difference between the current price and the strike price – being systematically under-priced by the market.
In which case, he has got a model, a ‘cognitive scheme’ – just one that happens to be different from everybody else’s.
Yes that distinction between investment and gambling is a moral one to the core, and one sociologists are starting to interrogate. One starting point is that all the facts of the market assume a society in compliance with legal codes – the two can never be disentangled.
Options markets are incredibly complex matrices of hedging and speculating strategies. A radical contrarian can either make or lose a lot of money if she gambles on a big move in the price of the underlying asset. Because, by definition, the market has decided that it is not worth the cost of discounting against that possibility.
Market participants are always surprised by big movements. They don’t deny that there are black swans out there. However, they tend to over-estimate their rarity, or at least underestimate the likelihood that they will make an appearance.
Like gambling, options are a negative-sum game. The market makers (the house) take their cut out of the pot and the rest is distributed between the winners and the losers.
However, options do serve the function of distributing risk, the premium of which is incredibly sensitive to perceptions and expectations, unlike the roulette wheel, which is nothing more than a random number generator.
The most efficient way of taking advantage of a black swan event is to bring your own black swan.
Surely the whole concept of a “Black Swan” presumes a certain epistemological orientation – one that needs to strip away all sorts of things that actually occur and have causal force in order to construct a model in the first place.
In short, at night all swans are grey, n’est-ce pas?
My former supervisor wrote his own PhD on Popper, and once travelled to meet the great man himself – then advancing in years.
Popper apparently looked him up and down, and asked “You’re not an empiricist, are you?”
Could you explain that a little more, Katz?
Kimbrella XXII The Pirate Queen, ja that’s pretty much what I’m thinking.
LE – heh
Yep – thought we were on the same page, dk!
You obviously didn’t understand Taleb’s book.
“Needless to say, figuring out that these events were worth the bet takes skill, determination, a Platonic schema and … uh oh, we’re back to those characteristics that in the nasty, bad, re-education column in Taleb’s book…”
Taleb says he almost always buys deep out of the money options for little cost hoping that a small prob event will become reality as markets at times telegraph their move far clearer than most times.
What draws you to the conclusion to say that Black Swan was self-righteous as it seems the word is why out in the ball park in terms of describing the essence of the book. Please explain.
Professor Quiggin also shows little understanding and seems more like a surly, almost envious post than anything else.
He says:
“But of course the reason such a thing had never occurred was that local housing markets had been separate from each other, with their own sets of banks, S&Ls and other financial institutions. The very banks that were doing the modelling were creating the conditions under which a national bubble and bust could take place. This was both foreseeable and foreseen.”
So why didn’t Prof Quiggin see things and sound the ship’s alarm if he smugly applies the ” but of course” line?
Silliness is being sold at a discount in this post and you don’t even have to buy an out of the money option on silliness
Ooohhhh, I thought you were talking about a this canberra variety in a bottomless pit.
Care to take another shot at that one, judge? I’m not following ya.
My argument is a simple one:
1) Taleb argues Black Swan events are not amenable to prediction.
2) His investment strategy involves seeking out small probability events that are, in some senses, amenable to prediction
3) Ergo, his argument is either (a) full of shit – as I gestured in the post, or (b) raises a whole host of novel ontological questions about the devices we use to make the world amenable to calculation.
“Care to take another shot at that one, judge? I’m not following ya.”
I’m not following you now. You called the book “self righteous”, so please explain why?
“My argument is a simple one:
1) Taleb argues Black Swan events are not amenable to prediction”
correct.
“2) His investment strategy involves seeking out small probability events that are, in some senses, amenable to prediction”
Not correct. He looks for low prob events that offer an oversize payout. He’s wrong very often, as says, but the one big one will pay off big time.
His investment strategy is not to “predict” black swan events. It’s to find low cost events with big payout ratios. That’s entirely different from black swan events that are entirely unpredictable and in lots of cases can’t be traded.
“3) Ergo, his argument is either (a) full of shit – as I gestured in the post, or (b) raises a whole host of novel ontological questions about the devices we use to make the world amenable to calculation.”
He’s not full of shit and the events are not calculable.
Here’s the point. Taleb also didn’t see the train coming. However once the train started to flatten those standing on the tracks he began to understand he gravity of the situation and traded accordingly.
Did he predict the event? No.
Did he take a bet on the ” depth” of the outcome”? Sure he did.
His basic gripe is that VAR is an awful system to assess risk as there is far more risk than people understand and will get comfortable with an inappropriate risk valuation method.
That’s why I said you didn’t understand the book. (Not trying to be offensive of or anything).
Why falsify your own hard fought thesis here @LP when there are so many clean and helping well-washed sociological hands prepared to do that for you here in real-time?
Just put it out there folks. Its not a crime.
And if this Thesis of yrs be some ‘black-swan’ Nemesis event – then let it be a huge oil slick to rival the Exxon Valdez. I’m going below for a Scotch.
I never got through the introductions of Taleb’s books, since they kept taking the tone of the book he wanted to write called “Why I am smarter than everyone else and certain about this, By Nicolas Nassim Taleb”.
I’m sure he had good points about narcissim if you can tread through his own narcissm though.
However, since Quiggin’s own best work was about uncertainty, I’m very inclined to believe he has some idea of how he’s approaching the issue.
What The Judge said.
I find this sudden fixation on Taleb pretty amusing, albeit unsurprising. It’s not coincidental that it took what he calls a Black Swan event to turn The Black Swan etc. into a bestseller.
While The Black Swan etc. is now ridiculously well-known, his earlier works Fooled by Randomness (2001) and Dynamic Hedging (1997), which say pretty much the same things, tend to be looked over. Malcolm Gladwell wrote an excellent article on Taleb and his thinking back in 2002. Additionally, although Taleb himself acknowledges the influence of Benoit Mandelbrot on his ideas, that rarely gets a mention in the coverage Taleb receives.
What’s also MUCH less well known, is that the original firm Taleb established in 1999 to take advantage of his strategy (i.e. continually (re)investing in deep out-of-the-money options), Empirica Capital, endured years of crappy returns before Taleb folded it in 2004. That is, he gave up on this supposedly winning strategy because he got sick of bleeding.
You’ll note that Universa isn’t Taleb’s firm; it’s run by his former trader at Empirica. Taleb is the “Senior Scientifc Advisor” to Universa.
“So why didn’t Prof Quiggin see things and sound the ship’s alarm ”
You mean, like here or here or here (if you want you can trawl through some more here)
But isn’t this simply another way of saying that Taleb is predicting “black swan events”?
The above passage can be reworded:
“Taleb’s investment strategy is to bet that his counterparties will assign a lower probability of certain events will occur than the actual probability of their occurrence. Taleb predicts that his counterparties overestimate the probability that markets expectations will not experience rapid changes.”
Thus the crux of the issue isn’t the “black swan” event but rather counterparties’ expectations about the “black swan” event. In other words, counterparties, rather than events, can be said to behave in predictable ways that are detrimental to their interests.
“…low cost events with big payout ratios.”
That’s pretty much how I see my term deposit at the bank these days.
Thanks Katz. That’s what I was trying to get at, except my contention is that the line between the action of counterparties and the events themselves is blurred by Talebs interventions. I understand that he’s arguing an inherent flaw in human ability, but to suggest that these low probability events are not calculable doesn’t actually make sense because he’s getting a monetary payout for betting on them. My contention is that whether his calculations are based on the actions of counterparties, rather than ‘fundamentals’ or whatever you want to call long run investments, is irrelevant.
Yeah I hope so too, Lefty E, but Ken Lovell would disagree
JQ, your first link is nothing more than an analysis of asset bubbles, and their apparent inevitability. Even a broken clock is correct twice a day.
Your second link is also no more than a generic statement about economic cycles.
The third is a scary bedtime story about derivatives from 2002 – six years ago – in which you even state that “the full-scale meltdown scenario, while far more plausible today than even a year ago, remains a low probability event”. That’s not a prediction.
Although you seem extraordinarily keen to say “I told you so”, your predictions were about as informative as warning the captain of the Titanic that there be icebergs in the North Atlantic. True, but not terribly helpful.
What’s the point about having a pissing contest about predictions, though? The desire to exercise some sort of control over the future through enlightened analysis? Everything Jack Strocchi writes is exemplary of where this leads!
Yes and no. Taleb is predicting that black swan events will occur, but not what they are, when they will happen or how probable they are. It is worth reading the Gladwell article linked to by Fanned by Foolishness @16, especially for those who find Taleb’s personal style unbearable.
It is arguable that the real black swan event at the heart of the current financial mess was the invention, by person or persons publicly unknown, of the gambling forum known as the credit default swap market.
But, Mark, as Jack keeps telling us, he’s a psephological genioos!
Katz and dk.au, you misunderstand Taleb’s position. His argument is that the market systematically underestimates the probability of very low frequency events, thereby causing the mispricing of deep out-of-the-money options. By buying these assets at less than what he believes is their intrinsic worth he expects to make money over the long term, assuming that that market inefficiency persists so that, every so often, a major crisis allows his strategy to clean up. He doesn’t need to be able to predict these events, or calculate their probability – which he does believe is incalculable – all he needs to know is that the market (i.e. “counterparties”, as Katz puts it) is consistently biased, so that the mispricing persists.
I’m not sure that I agree with that dk.au.
I guess that any market transaction can be said to be an intervention that blurs the actions of a counterparty.
Taleb’s counterparties always think this when they deal with him:
“This sucker wants to believe that some completely unlikely event will come along and make his dumb out-of-the-money gamble pay off. It’s like betting on a donkey in the Cox Plate. But if he’s willing to make a small contribution to my retirement fund, why should I stop him?”
To Taleb’s counterparties, this looks like a clear case of darwinian survival of the fittest. If Taleb persists in this behaviour, they think, it won’t be long before he is financially extinct.
In the overwhelming number of cases Taleb’s counterparties prove to be absolutely correct.
But every now and then lightning strikes the entire field of the Cox Plate, except, of course, the donkey. And in that case, Taleb’s counterparty blows up completely.
He doesn’t need to calculate them, just like economists don’t need to actually test their theories in real life to see if they’re of any use. But I’m talking about how his ideas are actually practiced, and in the real world the calculability arises from the ‘consistent mispricing’. To quote the article again:
Don’t they? Who says he doesn’t need to test his theory? You are, again, confusing the calculation of the probability of a given event, with the systematic pricing of probability/risk. They are not the same thing.
Did you look up my reference to Mandelbrot earlier? I’ll give you a hint: the answer lies in assumed probability distributions.
Nostrodomus, if you had read the linked post instead of mouthing off, you would have seen “The big problem for the Cassandras (and we were certainly both correct and disregarded) was that it was easy to see that the bubble could not continue and much harder to foresee how it would end”. That is, I didn’t make any particular claims to be able to predict the end, precisely because The market can stay irrational longer than you can stay solvent. As commenters note above, the same was true for Taleb’s funds.
“Taleb is predicting that black swan events will occur, but not what they are, ”
Sounds decidely Rumsfeldian to me.
I DID read the post, JQ, but I was responding to your list, at #16, of instances where you “sounded the alarm”. As I pointed out in response none of those examples had much information that would have been useful to those supposedly being warned, which might explain why your Cassandric correctness was disregarded.
Or, as I put it more cynically above, a broken clock is still correct twice a day. And the perpetually bearish can always count on economic cycles to prove them right. Eventually. Though the way you put it reads like you believe you really could predict what would happen, but were impeded from claiming so by solvency concerns. Or something.
Erm, that commenter would have been me, I suppose, but I didn’t say that he or his funds became insolvent.
“Erm, that commenter would have been me, I suppose”
Erm, no, unless you’re also commenting as Fanned by Foolishness.
Erm, yes – you can tell from the gravatar, for one thing.
Gravatars don’t show in IE or in some browsers, Nestorius.
[This comment also for the attention of LP's webslaves]
Taleb’s decision that mispricing has occurred and continues to exist is itself a calculation. Thus Taleb is by definition calculating the probability of occurrence of a set of “black swan events”. Or more precisely, Taleb is calculating the lowest bounds of mispricing by his counterparty that defines whether his decision to purchase a certain options contract is economic.
Yep. Now you’re getting it.
I should note, however, that you’re getting a little hung up on this “counterparty” nomenclature. I think Empirica and Universa only bought/buy exchange-traded options, so the “counterparty” is effectively the broader market.
In the examples used in this thread, Taleb is the “buyer” or the “taker” of the option. This option doesn’t exist until someone else (the counterparty) has decided to write the option.
It is true that these options are traded through an exchange. Nevertheless, there are always two counterparties – the writer and the buyer. (Often, the writers create large numbers of options which are bought buy the numerous takers of options, who are typically smaller operators than option writers.
If Taleb decides to exercise his option, the liability falls to a single counterparty.
dk.au,
One thing to keep in mind is the difference between the probability of a particular event, and probability in the aggregate. One may be calculable even if the other isn’t.
For example, what’s the probability of your house burning down tomorrow? You could calculate:
no. of house fires in Aus each day / no. of houses = 0.0000X %
But that would have little validity, because there are a huge number of factors unique to you which will affect your particular probability. Is there an undetected leak in your gas stove? Has your next-door neighbour recently earned the enmity of a deranged arsonist? And so on …
It’s basically incalculable at an individual level — one of Taleb’s ‘Black Swans’.
However, someone could look at house fires in the aggregate and conclude: Insurance Co has mis-estimated their frequency -> its premia are too low -> its profitability is going to go south -> I shall take out put options on it.
That might be perfectly valid even if you don’t have the foggiest idea how to work out the % of an individual fire. This seems to be similar to Universa’s approach.
Perhaps an analogy could be made to the behaviour of sub-atomic particles versus physical mechanics at a higher scale — but I ain’t no physicist, so I’ll cease and desist for fear of self-embarrassment!
There is an interesting analogy here with exchange rate forecasting. There is a nice little paper by Frankel and Froot (1990) called Chartists, Fundamentalists and Trading in the Foreign Exchange Market. The paper argues that in the mid-1980s there was a “bubble” in the USD exchange rate, which was accentuated by “chartists” (technical analysts) that traded on the basis of extrapolation from recent events. The paper also argues that such chartists drove so called “fundamentalists” out of the market because fundamentalists’ forecasts for short-term exchange rate depreciation were proven wrong month after month – trading as a fundamentalist was a loss making proposition. Of course, that makes sense because asset prices can depart from so-called fundamentals for long and uncertain periods of time. The near impossibility of forecasting short and medium term exchange rate movements is a key reason why most central banks assume in their projections that the exchange rate will behave as a random walk.
I’m constantly amused by bloggers/economists that claim to have “predicted” the financial crisis. The more accurate description would be that there were a bunch of people that were worried about a variety of imbalances in the global economy (US and other countries’ house prices, rapid global money and credit growth, the explosion in derivatives, the size of the US CAD and related enormous accumulation of USD reserves in east asian central banks, etc) and that these imbalances would be unwound in a disorderly way.
There was nothing unusual about these concerns (the IMF, the BIS, many central banks all had them to varying degrees), but they were generally placed in forecasters’ “risks” to the outlook, rather than central projections – for the obvious reason that it was extremely uncertain when and how they would play out. For example, if you were an Australian policymaker in 2005, would it have made sense to have loosened fiscal or monetary policy settings because you thought that an asset bubble had developed in the US that might be unwound in a destabilising way? Of course not. It made sense to note it as an event risk and then react to events as they unfolded.
On top of that, the majority of the “predictors” expected the unwinding of imbalances to be associated with a very sharp depreciation of the USD as investors rushed out of USD denominated assets, particularly Treasuries. That of course hasn’t occurred (yet) as investors have instead engaged in a a so-called flight to quality.
JQ, for example, after the AUD underwent its recent sharp depreciation boldly suggested that it was evidence that markets are irrational. But would you really want to hold assets in the current environment denominated in the currency of a small country with net foreign liabilities of around 70% of GDP, where there is a danger of its export prices collapsing, and where interest rate expecatations have fallen more sharply than in nearly any other developed country?
A touch more humility would be useful.
Nostrodamus and others have explained it pretty much above, but I’ll give it another go. Taleb’s funds were on the basis that we systematically underestimate rare large events. As far as predicting goes that is basically it.
By buying particular option positions (out of the money puts or calls depending on the situation) – you can have very small downside exposure with massive upside exposure. You slowly bleed money but occasionally cash in. Its similar in to if you worked out that bookies systematically gave outsiders 1000-1 odds where as they were really say 100-1. You are only going to win 1 in 100 bets but when it comes off you’ve returned 10 times your outlay.
Part of his rationale is that there is strong incentives in corporations and elsewhere to get consistent stable profits and to acheive this exployees frequently take the opposite position to Taleb – steady small profits in exchange for the risk of occasional blow ups. Unlike owners of the capital, corporate employees have little downside risk but generally good upside. Its in their interest to expose their employer’s capital to the risk of blow up in exchange for short term profitability that drives their bonus. When or if the thing blows up the big sallaries and bonuses they’ve acrued are already stashed away.
As for Taleb’s funds I’m not sure he quit in failure, but rather because he had plenty of cash (as he frequently tells you has had plenty since winning on the 1987 black swan) and wanted to prostelytize others to his ideas.
Daniel Davies on the Politics of Classification, part reason-to-post-video-of-Thomas-Friedman-getting-pied, part review of The Politics of Large Numbers (“an excellent book for anyone who is ever tempted to think Bruno Latour’s work never had any really useful applications.”)
heh