To be honest, I quite like NNT and find what he says overall enjoyable. I don't really like his perennial self important attitude, that he's rich and he can master french poetry and never wears a tie etc., but I guess that's part of the show. Instead, what I really don't like (or maybe don't get) is that he raises a lot concerns, and all the time the only solution seems to be: don't do any trade.

QuoteOriginally posted by: DevonFangsTo be honest, I quite like NNT and find what he says overall enjoyable. I don't really like his perennial self important attitude, that he's rich and he can master french poetry and never wears a tie etc., but I guess that's part of the show. Instead, what I really don't like (or maybe don't get) is that he raises a lot concerns, and all the time the only solution seems to be: don't do any trade.I have always tended towards rather liking him, but I think his arrogant air has increased since he got more famous. But I dont read him in quite the same way as you. I think his main point is simply that you should not trust people who say they manage, understand or any way can predict uncertainty. The weak form of this, that 'risk' can be managed, is he argues, a mirage because risk in the narrow way in which it is defined in any model, still assumes known unknowns. So I dont think he argues that you should not do any trades. I think he argues that you should not do it on the basis of any of the above because i) either you do understand it, but you do it anyway, which makes you dishonest or ii) you dont understand it, and hence you are 'the mark'. As I understood it, his investment philosophy is based around taking out the marks who think they get risk but dont. The more hyperbole he can put into these arguments the better for his self promotion. Without his image as a meat-eating, ex-trading, intellectual-rentier-slaying, bearded hero he would be just another finance professor in a tweed suit.

QuoteOriginally posted by: DevonFangsTo be honest, I quite like NNT and find what he says overall enjoyable. I don't really like his perennial self important attitude, that he's rich and he can master french poetry and never wears a tie etc., but I guess that's part of the show. Instead, what I really don't like (or maybe don't get) is that he raises a lot concerns, and all the time the only solution seems to be: don't do any trade.His solution is more like starving the beast, which would be disastrous for the stock market & economy while enriching his black swan fund greatly. NNT seeks crisis and failure not only for personal gain, but due to some ingrained resentment or perceived hurt by an illusory cabal of economists and technologists who are at the root of every problem.

QuoteOriginally posted by: DevonFangsTo be honest, I quite like NNT and find what he says overall enjoyable. I don't really like his perennial self important attitude, that he's rich and he can master french poetry and never wears a tie etc., but I guess that's part of the show. Instead, what I really don't like (or maybe don't get) is that he raises a lot concerns, and all the time the only solution seems to be: don't do any trade.He rather warns against what everybody knows in theory but not necessarily applies in practice: that we inevitably underestimate the unpredictable (and often consciously ignore its risks encouraged by the lack of liability), that we like thinking in a Gaussian way and thus can easily fall a victim of the gambler's fallacy so to speak. Of course you can trade - NNT only reminds the old truthI think about him every time I close the wrong trade on time (and every time I don't close the right one too early!) :-)Now he's indeed become more of a philosopher and populariser of the mathematical finance than a practitioner. His latest theories draw on complexity and emergence, and I think it's important that someone tells people that such concepts exist and model the mechanisms of the market. And, c'mon, I sit most of the day (and night...) watching my scripts running trades so I am interested in everything, including what NNT likes to eat and wear. Where were you dining today, DF, at Wasabi or Tortilla?@ what you don't like about NNT quiz: I didn't like some parts of Black Swan because they were too naïve for someone with my background, but it was unarguably an important book...

I just can't wrap my mind around how in his books he's criticizing mathematical finance yet he uses financial math in his papers? Pick one or the other. The income he earns from lectures and books overrides this apparent contradiction.

QuoteOriginally posted by: investor82I just can't wrap my mind around how in his books he's criticizing mathematical finance yet he uses financial math in his papers? Pick one or the other. The income he earns from lectures and books overrides this apparent contradiction.There's really no contradiction in Nassim's use of math for at least two reasons. First, the math does provide some useful insights into price, risk, sensitivities and all that. As long as one remains cognizant of the potential modelling errors (i.e., avoids excessive leverage based on the mistaken assumption that the math looks more accurate than it is) then the approximations can be useful. Second, he uses the math to show how dangerous the math is, especially the "picking-up-pennies-and-losing-dollars" strategies that arise from unintentional or willful ignorance of the errors in the math and biased perceptions of risk and causality.

QuoteOriginally posted by: Traden4AlphaQuoteOriginally posted by: investor82I just can't wrap my mind around how in his books he's criticizing mathematical finance yet he uses financial math in his papers? Pick one or the other. The income he earns from lectures and books overrides this apparent contradiction.There's really no contradiction in Nassim's use of math for at least two reasons. First, the math does provide some useful insights into price, risk, sensitivities and all that. As long as one remains cognizant of the potential modelling errors (i.e., avoids excessive leverage based on the mistaken assumption that the math looks more accurate than it is) then the approximations can be useful. Second, he uses the math to show how dangerous the math is, especially the "picking-up-pennies-and-losing-dollars" strategies that arise from unintentional or willful ignorance of the errors in the math and biased perceptions of risk and causality.There is nothing wrong with the maths; it was just applied in the wrong way to to the wrong problem.

QuoteOriginally posted by: PaulI will restrict myself to commenting on the psychological side of this, not the economics or finance. There is very little chance that he is envious of those you mention! Almost certainly the envy goes the other way. He is extremely successful and rich, and has become so by going against the mainstream...P.He is the mainstream, that's the problem. People have been droning on about all that fat-tails, black-swan nonsense for years.Meanwhile, the philosophy of probability has a history going back centuries which no-one bothers to learn. His 'fooled by randomness' stuff was much better, but hasn't caught on in the same way at all. I can't think why.

QuoteOriginally posted by: spacemonkeyHe is the mainstream, that's the problem. People have been droning on about all that fat-tails, black-swan nonsense for years.Meanwhile, the philosophy of probability has a history going back centuries which no-one bothers to learn. His 'fooled by randomness' stuff was much better, but hasn't caught on in the same way at all. I can't think why.The fat-tails, black swan ideas are NOT nonsense to anyone who must look at and assess distributional assumptions as a part of their job. What IS nonsense is that there are a lot of parrots out there who latch on to Taleb's ideas, sensing that there is something "important" in there, and run around bastardizing the terminology like "fat-tails, non-Gaussian," etc. to others so that non-technical people will believe that they are brilliant. The problem is, the parroters lack the math skills to properly comprehend these ideas and why they are generally right. I agree with the others that Taleb's style really reduces the effectiveness of his messages, which is a shame. Spacemonkey, I definitely understand your point on the Fooled by Randomness ideas being "better"...although he still had a nasty bite when he wrote that, the approach seemed softer and more genuinely philosophical, which is where I think this kind of work needs to remain at the moment for it to be effective. Now Taleb seems very dismissive and impatient (which, as a fan, I can understand), but Taleb fails to recognize that he is at the point where if he just attenuated that attitude, he could catch many more flies with honey than vinegar and in the long run, actually benefit his causes much more effectively.Text

QuoteOriginally posted by: LineOfBestFitQuoteOriginally posted by: spacemonkeyHe is the mainstream, that's the problem. People have been droning on about all that fat-tails, black-swan nonsense for years.Meanwhile, the philosophy of probability has a history going back centuries which no-one bothers to learn. His 'fooled by randomness' stuff was much better, but hasn't caught on in the same way at all. I can't think why.The fat-tails, black swan ideas are NOT nonsense to anyone who must look at and assess distributional assumptions as a part of their job. What IS nonsense is that there are a lot of parrots out there who latch on to Taleb's ideas, sensing that there is something "important" in there, and run around bastardizing the terminology like "fat-tails, non-Gaussian," etc. to others so that non-technical people will believe that they are brilliant. The problem is, the parroters lack the math skills to properly comprehend these ideas and why they are generally right.BTW, are black swans subject to Pareto scaling (power laws)?I agree with the others that Taleb's style really reduces the effectiveness of his messages, which is a shame. Spacemonkey, I definitely understand your point on the Fooled by Randomness ideas being "better"...although he still had a nasty bite when he wrote that, the approach seemed softer and more genuinely philosophical, which is where I think this kind of work needs to remain at the moment for it to be effective. Now Taleb seems very dismissive and impatient (which, as a fan, I can understand), but Taleb fails to recognize that he is at the point where if he just attenuated that attitude, he could catch many more flies with honey than vinegar and in the long run, actually benefit his causes much more effectively.TextAnd Gaussian distribution was Black, Scholes and Mertons' idea? Who was Gauss then?Fat tails and black swans (a.k.a. the distribution outliers) are not Taleb's ideas - he just pointed out their existence in financial markets, and openly warned against them. (Just like the selfish gene is not Dawkin's idea, but the adaptation of the work of biologists like Hamilton, Wilson, Haldane et al. - a digression.) Taleb's got a brilliant open mind capable of very critical thinking and deep understanding of problems. He shares them in his books and articles using different approaches and presenting them from different perspectives, and they all form a consistent picture - none of them is nonsense. But it's like with Karol Marx: when you read Das Kapital, you keep nodding your head agreeing with his insightful unveiling criticism of capitalism. But in The Manifesto he only proves that there're no straightforward solution to the problems.I would like to ask a question to some experienced quants. It seems quite easy (to a newbie like me) to propose alternative models taking into account the fat tails. From what I understand people want to have fast analytical solutions, which is why they like the Gaussian distribution so much. Cannot one start from fitting the data with stable distributions with the stability parameter lower than 2 (to account for the fat tails), then regularise it in a standard way, and put in the old models? The exponential regularisation factor will not complicate the differential equations; it will introduce to the models a new parameter charcterising the kurtosis (i.e. the prob of extremely high losses or gains). Or why not simply stay with the old Gaussian models (between the rare large events the noise distribution is pretty Gaussian), but from a simple fit estimate the tails fatness as the prob of extremely high losses or gains?QuoteOriginally posted by: spacemonkeyQuoteOriginally posted by: PaulI will restrict myself to commenting on the psychological side of this, not the economics or finance. There is very little chance that he is envious of those you mention! Almost certainly the envy goes the other way. He is extremely successful and rich, and has become so by going against the mainstream...P.He is the mainstream, that's the problem. People have been droning on about all that fat-tails, black-swan nonsense for years.Meanwhile, the philosophy of probability has a history going back centuries which no-one bothers to learn. His 'fooled by randomness' stuff was much better, but hasn't caught on in the same way at all. I can't think why.Maybe because the black swan appeared and slapped us on the face with its wet cold foot? As for the artifacts described in the other book, there's a body of statistical methods guarding us against them. They will surely become important in finance with the growing interest in "data mining".

QuoteOriginally posted by: investor82His solution is more like starving the beast, which would be disastrous for the stock market & economy while enriching his black swan fund greatly. NNT seeks crisis and failure not only for personal gain, but due to some ingrained resentment or perceived hurt by an illusory cabal of economists and technologists who are at the root of every problem.Never trust a man who runs a hedge fund.

QuoteI would like to ask a question to some experienced quants. It seems quite easy (to a newbie like me) to propose alternative models taking into account the fat tails.I am writing just two lines because I am afraid to derail the thread (It's better if you search for other threads or open a new one with this separate question if you need a more detailed answer).There are lots of ways to take in account fat tails and in fact there is a lot of literature about this. The problem is that taking in account fat tails is only useful if you have the correct ones! If we knew (or if it existed) the data generating process of financial data we could write the distribution and do all the correct modelling accounting skewness and kurtosys. The problem is that the DGP changes all the time, so even estimating the empirical distribution (this has been done since the beginning of modelling derivative prices) doesn't necessarily give us the current distribution (it may be for a while before a new regime change).

... and you have just a little bit of data about the fat tails, since they are composed of RARE large events ;-)The difference between finance and physics is that in physics you usually have the underlying microscopic theory which tells you that you should have fat tails and they should have this distribution, and the theory is confirmed by a number of controlled experiments. There much less of such solid foundation in finance, which means you have to rely on statistical observations more. Hopefully as we gather more and more data about everyone, we will be able to understand social dynamics better (time to apply for a grant from NSA!).In short, finance is at the stage of XVIIth century physics when it comes to explaining WHY markets behave the way they do.

QuoteOriginally posted by: UltravioletI would like to ask a question to some experienced quants. It seems quite easy (to a newbie like me) to propose alternative models taking into account the fat tails.Proposing a model is a piece o cake, fitting it to data is already sophisticated. Fitting a model to nonexistent data (and black swans happen very very rarely by definition) is a real challenge.

Thanks for your answers, gc and katastrofa.QuoteOriginally posted by: katastrofa... and you have just a little bit of data about the fat tails, since they are composed of RARE large events ;-)There is the extreme value theory to tackle such stuff, I believe. But even without resorting to it, why can't we determine the stability parameter of the distribution in the current regime from e.g. width at half maximum? It's different for different tails' fatness, it scales differently with changing volatility, ...There's also another reasoning suggesting that we shouldn't need so much data. The market as a whole is a closed system (if one doesn't count external events like Tunguska meteorite) - or at least let's make such an operational assumption. The market's "illusionary" openness results from the finite speed of the information transfer - I don't know what is the decision of another trader until he/she makes his/her trade, even if we are the part of the same market. With the information spread, the system converges to a stationary state. It is unique equilibrium, in which the BS theory applies, we can talk about e.g. "the true prices", etc. (I'm still trying to figure out how one could make any money under such conditions ) Indeed, we cannot gather data from the whole market to model it (if we did, we would in fact turn it into the BSM steady state). Rather, because of the finite speed of the information transfer, the market consists of subsystems with independent equilibria. Their dynamics is Gaussian within their boundaries, while the spread of the information between them is associated with jumps accounting for the fat tails. I would focus on one of such subsystems that I am particularly interested in. The rest of the market I could model as some external world to which I am coupled. The coupling would be characterized by some sort of a power density function specific to any chosen subsystem, and it could be constructed from historical data. I could see how strongly coupled/"exposed" I am to some "frequencies": I could be extremely fragile to some evens and oblivious to other - isn't it enough to hedge/stress-test my portfolio? Again, there's a whole field of mathematics and physics to model such problems and ready to use methods.QuoteThe difference between finance and physics is that in physics you usually have the underlying microscopic theory which tells you that you should have fat tails and they should have this distribution, and the theory is confirmed by a number of controlled experiments. There much less of such solid foundation in finance, which means you have to rely on statistical observations more. Hopefully as we gather more and more data about everyone, we will be able to understand social dynamics better (time to apply for a grant from NSA!).The common notion that physicists have reliable data is hardly ever the truth. Every good honest experimentalist will tell you so. Physics is challenged by the complexity of realistic systems (which is no lesser than in the case of financial markets). It often happens that experiments are not replicable or only realize us how complex are the problems we investigate and show a small part of it. What works relies on very basic effects or, like in medicine, on statistics.QuoteIn short, finance is at the stage of XVIIth century physics when it comes to explaining WHY markets behave the way they do.They didn't have Gaussian distributions in the XVII century, cupcake.Disclaimer: I'm fully aware of my ignorance in this field and that my ideas are either not original or make no sense Just got a bit excited about it and curious to hear some remarks from the "engineers".