6 ms·
Gaussian vs. Mandelbrotian: The Great Intellectual Fraud
- zzzmarcus 17y agoFull version of the chapter here: http://issuu.com/azureo/docs/the_black_swan/1?mode=a_p http://issuu.com/azureo/docs/the_black_swan/1?mode=a_p Issuu doesn't have a deep-linking feature.
- jwecker 17y agoHighly, highly recommend this book to anyone who hasn't taken the time yet. Not all that revolutionary for anyone firmly grounded in the real ins and outs of probability, but surprisingly revolutionary for those who know just enough probability to be dangerous, so to speak.
- osipov 17y agoGiven the audience on HN I recommend the very readable "Misbehavior of Markets" by Mandelbrot. Everything worth reading in Taleb's book is a dumbed down rehash of Mandelbrot's ideas.
- geezer 17y ago"Not all that revolutionary for anyone firmly grounded in the real ins and outs of probability" The fact that those familiar with the subject matter do not think this is revolutionary is exactly the problem Taleb is pointing to. He is calling it "the great intellectual fraud" for a reason. Domain experts obviously do not believe "intellectual fraud" and "bell curve" should be in the same sentence
- viggity 17y agoIf anyone wants to know what the ten deutschmark bill looks like: http://www.cunymath.cuny.edu/images/sidebar/gauss_10_DM.jpg http://www.cunymath.cuny.edu/images/sidebar/gauss_10_DM.jpg
- duncanj 17y agoI could totally hang out with that guy. Taleb's a fool.
- perkoff 17y agoWhy is Taleb getting all this mainstream attention? The finance guys have been aware of fail tail distributions all along. As Eugene Fama points out on his website... http://www.dimensional.com/famafrench/2009/03/qa-confidence-in-the-bell-curve.html http://www.dimensional.com/famafrench/2009/03/qa-confidence-... "Half of my 1964 Ph.D. thesis is tests of market efficiency, and the other half is a detailed examination of the distribution of stock returns. Mandelbrot is right. The distribution is fat-tailed relative to the normal distribution. In other words, extreme returns occur much more often than would be expected if returns were normal. There was lots of interest in this issue for about ten years. Then academics lost interest. The reason is that most of what we do in terms of portfolio theory and models of risk and expected return works for Mandelbrot's stable distribution class, as well as for the normal distribution (which is in fact a member of the stable class)."
- oakmac 17y agoThe finance guys may have been aware of it, but their actions were clearly not in line with this knowledge as evidence by the year 2008. From the article you linked: "None of this implies, however, that the existence of outliers undermines modern portfolio theory or asset pricing theory." In fact, that's exactly what it does. This is what happens when you build houses on top of sand.
- perkoff 17y agoBig crashes will always occur, but I would not blame the recent crash on the gaussian models. I'd rather say that (almost) everyone underappreciated the risks connected to the real estate prices. Taleb pushes for a strategy that consists of buying a lot of very safe assets and blending them with bets on "extreme events" (like buying far out-of-the-money put options). Is that a viable long-term strategy? I have my doubts, since there are no evidence suggesting that 'uncertain' strategies have greater returns that more quantified ones.
- jwecker 17y agoSince Taleb started thinking through this stuff he's been able to cash out on two big crashes- just in the last 10 years. The second (current) one came after the black swan was published. It's almost erie reading it now. In any case, he's not advocating that everyone use that as a trading strategy. What he's advocating is that people be aware of the nature of the underlying system and stop fooling ourselves into thinking that it's "gaussian + weird things that are obvious in retrospect."
- mmt 17y ago"Take a random sample of any two people from the U.S. population who jointly earn $1 million per annum." This is where the author loses me. Where is the randomness of the sample, when there are two items which are interdependent? Later, he criticizes standard deviation as applied to stocks and bonds (decidedly non-random data), finding fault with the bell curve, rather than the misapplication. Does this chapter make any more sense in the context of the entire book?
- DenisM 17y agoI recommend to read the book. Taleb's point is that applying gaussian instruments to mandelbrotian data is not only wrong, it has a peculiar characteristic of being mostly right all the way until it goes horribly, horribly wrong. The time lag is crucial at it allows people to accumulate risks without realizing what they do.
- mbrubeck 17y ago"Where is the randomness of the sample, when there are two items which are interdependent?" Given all pairs of people whose join income is $1M, select a pair at random. "All pairs" is an unusual population to select from, but mathematically it's a perfectly valid way to define a random variable.
- mmt 17y agoThis makes sense, to be, so, perhaps, I'm confused by how he would then go on to talk about the individual values that make up the pair. Is that, then, exactly his point? That a random sample of a group can't be modeled like a group of random samples, yet this is what's being done?
- waldrews 17y agoSigh. Him again. Yes, he's unfortunately worth reading, for the drama and stimulation and calling attention to things that are known but not always fully appreciated. But I wish he wasn't worth reading. His hero Mandelbrot on the other hand? A real scholar with radical ideas, not a provocateur. The important ideas in his book are well known by thinkers, if not all practitioners, in the fields he criticizes. But, oh, the insults! And so much of what he says is "Such and such models have flaws! Throw them away and do nothing at all until you have perfect models!" That's not terribly useful if you are, say, running an insurance firm. Or doing science. Anyway, here's the issue of American Statistician duly taking him to task on the technicalities, though unfortunately without the bombast and mud-slinging. http://pubs.amstat.org/toc/tas/61/3 http://pubs.amstat.org/toc/tas/61/3 Also could somebody explain to me why hacker types like to name-check Popper but never Kuhn? Is it because the Star Trek TNG episode with the Binars actually got it right?
- moss 17y agoRegarding name-checking of Popper vs. Kuhn: Kuhn and Popper both make interesting observations about the development of science. However, Popper's observations are more directly applicable to the task of actually doing science. He gives a useful framework for developing and testing theories, where Kuhn's work has more to say about how theories become widely accepted. So Popper appeals to the pragmatism of most hackers. Kuhn is also, sadly, very popular with crackpots. Almost any fringe scientist will eagerly explain to you that his ideas represent "a new paradigm", and that those who doubt him are simply trapped in old ways of thinking. So a desire to avoid guilt by association probably also plays into it. That's my take on it anyway. (I know it's kind of a tangent, but it seemed like an interesting thing to think through).
- dspeyer 17y agoGoogle Books cut off before he got to explaining what the fraud was. Does anyone know? He seemed to be complaining about assuming distributions were normal without checking, a simple mistakes that is warned against in any introductory statistics class.