Today seems an appropriate time to bring up this topic: I’m reading Taleb’s ‘The Black Swan’ and he basically says that all of MPT and it’s main contributors, Markowitz, Sharpe, etc. are “phonies” and their theory amounts to nothing more than “charlatanism”. He rips apart Merton and Scholes for their options pricing model. He also lampoons Paul Samuelson and every other Nobel Prize winner in economics. The problem is that his arguments and view of the world seem to make more sense. What should a young CFA candidate do with this information? Why am I studying these methods if they don’t work and in Taleb’s estimation are a “danger to the world”. JoeD what do I do?
you have to play the game… no matter how @#$%ed up it is. as / if you start managing / trading you will begin to see the crap that they call “free markets” Quorky, CFA
Taleb is just being part of the dialectic, an academic troll. No question though that taking all that stuff literally can be really dangerous since none of it was meant to be taken literally.
Then it seems the formula for success is: 1) learn the wrong way to do it; 2) get in the door armed with a great background on the wrong way to do it; 3) forget everything you learned to avoid being the next big bag holder and losing all your clients’ money. That about sum it up?
If I’m not mistaken, MPT provides a framework with which to perform investment analysis. There has be some fixed rules in order to take a structured and consistent approach to valuation and portfolio management. Yes it’s not perfect but it is the best we’ve got. With regards the three step method, I was just reading book 2 this morning and came across an analysis of post-tax, post-inflation returns for stocks bonds and bills over the last 80 years. Negative returns for T-bills and stuff all return on govt and muni bonds. Plus long periods (20-25 years) of negative real after-tax returns on both bonds and equity. Given we’ve just come off a stellar 20 years for equity investment and the investment market is tanking it doesn’t bode well. Depressing.
DblA Wrote: ------------------------------------------------------- > Yes it’s not perfect but it is the > best we’ve got. This is the exact argument that Taleb claims is so dangerous. Especially if we dogmatically follow the predictions our models make. Does that mean those models are useless, who knows? One might wonder why Lehman was happy to be so exposed to residential real estate. My hunch is that they looked at historical home prices and said to themselves “nationally home prices have never declined so the risk to these investments is low”. Well, we’ve seen where that logic gets you. I’m not sure if I believe everything Taleb is saying in this book, but the last week certainly doesn’t hurt his argument.
Someone once said “all models are wrong, but some are useful”. The models (MPT, CAPM, Black-Schoies, etc…) don’t fully reflect reality. But then again, a road map doesn’t either - if it was to a scale of one inch = one inch, it would be accurate but useless. These models do, however, give some good insights (i.e. covariance is what matters, variance matters to options, etc…). There was recently some pieces I read on Wilmott’s website talking about how when he first got in the business, the B-S model obviously was inaccurate. But at the same time, it could be tweaked enought to make it “close enough.” My sense is that a lot of models are like that. I haven’t read Taleb’s book, but his whole gig is predicated on telling us how current models are wrong. For him to do otherwise wouldn;t be good marketing, and would make him a far less interesting speaker. I haven;t yet read his book, but what does he suggest as an alternative?
That’s one of the frustrating things about his book, he doesn’t give a better alternative, especially to those in large institutions who need to systematically implement investment objectives. For the individual he advocates a barbell strategy: large proportion in very conservative investments and spreading out small highly speculative bets in the hopes of capturing a positive black swan. I think the main lessons are: (1) don’t be the sucker, always be skeptical of your strategy and try look for instances where it fails rather than looking for confirmation that it works; (2) never be too confident in the range of possible outcomes; (3) never theorize – don’t abstract theories about how the world works by looking at historical data. It only takes one (possibly very costly) event to invalidate a theory.
I have a problem with item (3). I do remember the bits where Taleb talks about the seductiveness of a causal narrative, but I’m not sure what one does if you just throw out theories for the sake of being theories. I prefer the idea of being continually skeptical of one’s theories. Figure out how useful they are but try to find where the holes in assumptions are. Then try to be aware of when markets are doing things that don’t line up with your assumptions. When I was an academic, I wrote an article (never published) that was basically a criticism of how economics was trying to redefine itself from “the science of deciding what to produce, how to produce it, and how to distribute it,” to “the science of any decision that involves costs and benefits,” which I felt basically assumed away a whole lot of knowledge of other disciplines, like psychology, sociology, anthropology, political science, etc… One of the segments was about economic theory and the danger of believing your models too literally. The conclusion from that section was: a) Models are useful as a way to ORGANIZE how you understand the world (and data). They are best when they help you understand what things are important, how they fit together and interact. b) Models become dangerous when they are used as a SUBSTITUTE for understanding (or collecting data about) the world. If you start to think that you don’t need to know reality or check up on things because you have a nice theory, then you’re in trouble. Constant self-vigilance is important here.
Models are useful for looking at relative risk probabilities and getting past analysis paralysis. It’s like saying all weather forecasting is useless because it can’t predict every drop of rain 6 months in advance… What does Taleb propose in return? Tea leaves? I’m not attacking, just curious; I haven’t read the book.
Just a small example of the problem. We start out to create a model for stock prices that we can use to price derivatives. In the usual way of things, we think locally and say that stock price paths are Markov (i.e., where they are going depends on where they are without regard to how they got there). Add on a few assumptions like finite variances and continuous sample paths and we suddenly jump to normality. We didn’t think of normality because it was easy; we thought of normality because we like Markov. So now we have three assumptions that led to normality but we don’t like normality so which are we going to abandon? a) Markov - If I post that I believe in technical analysis, I get a bunch of people younger than my daughter telling me I am naive. “Markets are efficient and that means technical analysis can’t work.” Hmmm. b) Finite variances - This is a good candidate to throw away, but if variances aren’t finite we are all playing an infinite risk game. That means we all ought to be thinking about what it is that limits our risk (since all things are finite) and categorizing investments according to risk truncations or something. c) Continuous sample paths - Obviously not true but every continuous sample path problem in the world has been solved by adding in a jump diffusion process and it doesn’t help Taleb one bit. Taleb, of course, knows all this but if you can sell strawmen and make good coin, you should probably do that.
bchadwick Wrote: ------------------------------------------------------- > > When I was an academic, I wrote an article (never > published) that was basically a criticism of how > economics was trying to redefine itself from “the > science of deciding what to produce, how to > produce it, and how to distribute it,” to “the > science of any decision that involves costs and > benefits,” which I felt basically assumed away a > whole lot of knowledge of other disciplines, like > psychology, sociology, anthropology, political > science, etc… Nice premise (at the least because it pleases me ideologically).
JoeyDVivre Wrote: ------------------------------------------------------- > Taleb, of course, knows all this but if you can > sell strawmen and make good coin, you should > probably do that. I think you’ve hit the nail on the head. From what I know (admittedly, it’s not my areas of expertise), Taleb’s stuff hasn’t gotten too much play among academics. It could be (and this is his story) that the “existing order” is resistant to change. Or, alternately it could be because he’s full of horse hockey. But he’s a good writer and speaker. So he makes money by pushing his story to the masses. More power too him. Just take him with a large grain of salt. The fact that he hasn’t gotten a lot of academic love isn’t the end of the argument. But it is a significant data point.
…by reading The Black Swan, I seem to have jumped into the middle of an all out academic war that I had no clue existed. I suppose I should read about the ideas on the other side before making any judgements. No matter who/what’s to blame, it’s a little unnerving that we’ve witnessed two “100 year floods” in the span of 8 years. If you looked at equity fundamentals before the dot com bust and MBS spreads before this latest meltdown, it’s obvious that market participants were grossly underpricing risk. Maybe it’s just the psychology of bubbles – get in and out in time to find a greater fool – or maybe the quantitative risk assessments were sending false signals. I have no idea…
jbaldyga Wrote: ------------------------------------------------------- > …by reading The Black Swan, I seem to have > jumped into the middle of an all out academic war > that I had no clue existed. I suppose I should > read about the ideas on the other side before > making any judgements. > Except it’s not. Taleb has nothing new to say to academics. > No matter who/what’s to blame, it’s a little > unnerving that we’ve witnessed two “100 year > floods” in the span of 8 years. If you looked at > equity fundamentals before the dot com bust and > MBS spreads before this latest meltdown, it’s > obvious that market participants were grossly > underpricing risk. > You should expect 100 year floods pretty often. It’s a 100 year flood in a particular market and there are lots of markets. They are ultimately healthy as they fix up risk pricing problems in the future. > Maybe it’s just the psychology of bubbles – get > in and out in time to find a greater fool – or > maybe the quantitative risk assessments were > sending false signals. I have no idea…
JoeyDVivre Wrote: ------------------------------------------------------- > You should expect 100 year floods pretty often. > It’s a 100 year flood in a particular market and > there are lots of markets. They are ultimately > healthy as they fix up risk pricing problems in > the future. I find it hard to believe that different markets are incapable of taking lessons from one another. I mean, investment professionals are pretty dumb, but hopefully not that dumb.
Unfortunately, there are lots of ways to be dumb. We keep discovering new ones, but somehow they always involve leverage.
…and Wall Street “financial innovations”.
blame the rating agencies: http://www.bloomberg.com/apps/news?pid=20601109&sid=ax3vfya_Vtdo&refer=home
He swings too far the other way from the free market fundamentalists. Efficient market hypothesis is a theory that describes way world works with perfect information. Then you take into account uncertainty and assymetrical information. So you know market prices aren’t perfect and are susceptible to being “wrong” for extended periods. None of which means a bunch of 29 year old bank examiners or other regulatory jihadists will do any better at spotting the dangerous trends. Don’t leverage to the hilt unless you are willing to take the big losses with the big gains, and don’t bail out the gamblers with the innocent bystanders. And don’t pretend that cutting some highroller’s salary in half is going to solve any problems. Anyone really think the government, meaning an 8 hour a day (unless he/she taking a mental health day or a diversity training or 6 sigma or cheese stealing seminar) graduate of Our Lady of the Automatic Degree is going to do a better job monitoring this stuff with other people’s money on the line than the folks with skin in the game and their own money on the line? Hell of a fun book though. His first one is even better if you’re just generally intellectually curious and not looking for a guru with all the answers.