Comments on: "Financial Models Should Come With Health Warnings" http://www.nakedcapitalism.com/2008/03/financial-models-should-come-with.html Sat, 21 Nov 2009 23:52:37 -0500 http://wordpress.org/?v=2.8.4 hourly 1 By: Anonymous http://www.nakedcapitalism.com/2008/03/financial-models-should-come-with.html#comment-4988 Anonymous Tue, 11 Mar 2008 05:51:00 +0000 http://www.nakedcapitalism.com/2008/03/financial-models-should-come-with-health-warnings/#comment-4988 I expect that the mathematical models function in the same way as clothes for the salesman, to make a presentable case. In the old days you could make a sale by saying "the great JP Morgan likes this stock", these days since we are more sophisticated we need to be sold on the Ito calculus and Weiner processes. But (as I think) Mr Taleb implied there isn't a whole lot more to financial modelling than the random walk which one can handle with the binomial theorem. So we have a situation where the mathematicians provide cover and intimidation for the industry with incomprehensible stuff while the other guys clean out. Add to that the evident corruption and cynicism among financial operators and you have stoked up a huge ponzi scheme.<br/><br/>Ivan I expect that the mathematical models function in the same way as clothes for the salesman, to make a presentable case. In the old days you could make a sale by saying “the great JP Morgan likes this stock”, these days since we are more sophisticated we need to be sold on the Ito calculus and Weiner processes. But (as I think) Mr Taleb implied there isn’t a whole lot more to financial modelling than the random walk which one can handle with the binomial theorem. So we have a situation where the mathematicians provide cover and intimidation for the industry with incomprehensible stuff while the other guys clean out. Add to that the evident corruption and cynicism among financial operators and you have stoked up a huge ponzi scheme.

Ivan

]]> By: Anonymous http://www.nakedcapitalism.com/2008/03/financial-models-should-come-with.html#comment-4980 Anonymous Mon, 10 Mar 2008 22:18:00 +0000 http://www.nakedcapitalism.com/2008/03/financial-models-should-come-with-health-warnings/#comment-4980 In a way, it is a result of the scientific background. In Physics, at least, the number of analytical solvable models is limited. A physicist would prefer to use it, even if it's just a rough approximation. The analytical treatment provides, in many cases, insights on the problem.<br/><br/>The problem arises when science must become engineering. And a nice model is not enough, it must also be closer to reality. <br/><br/>But what can you do when your tools can only be applied to an inaccurate model? You try to patch it and hope for the best. In a way, it is a result of the scientific background. In Physics, at least, the number of analytical solvable models is limited. A physicist would prefer to use it, even if it’s just a rough approximation. The analytical treatment provides, in many cases, insights on the problem.

The problem arises when science must become engineering. And a nice model is not enough, it must also be closer to reality.

But what can you do when your tools can only be applied to an inaccurate model? You try to patch it and hope for the best.

]]> By: Lune http://www.nakedcapitalism.com/2008/03/financial-models-should-come-with.html#comment-4973 Lune Mon, 10 Mar 2008 19:22:00 +0000 http://www.nakedcapitalism.com/2008/03/financial-models-should-come-with-health-warnings/#comment-4973 Not being in the financial world myself, I'm curious to what extent finance types have backgrounds in statistics. While you hear about plenty of physics and math PhDs entering finance, I'm not sure their training includes much statistics, to say nothing of the undergrad econ major or newly graduated MBA.<br/><br/>It seems that many of the mistakes being made in the models are actually pretty elementary rules in statistics. It's a little frightening that people making billion dollar models don't understand the appropriate use of normal distributions. Anyone in the industry willing to offer any insight? Not being in the financial world myself, I’m curious to what extent finance types have backgrounds in statistics. While you hear about plenty of physics and math PhDs entering finance, I’m not sure their training includes much statistics, to say nothing of the undergrad econ major or newly graduated MBA.

It seems that many of the mistakes being made in the models are actually pretty elementary rules in statistics. It’s a little frightening that people making billion dollar models don’t understand the appropriate use of normal distributions. Anyone in the industry willing to offer any insight?

]]> By: Anonymous http://www.nakedcapitalism.com/2008/03/financial-models-should-come-with.html#comment-4961 Anonymous Mon, 10 Mar 2008 12:39:00 +0000 http://www.nakedcapitalism.com/2008/03/financial-models-should-come-with-health-warnings/#comment-4961 Biggest assumption:<br/><br/>heads: get a fat bonus,<br/>tails: get a new job.<br/><br/>Model that!<br/><br/>I do this cr@p for a living, and unlike the good professor, I don't get high on my own supply. Biggest assumption:

heads: get a fat bonus,
tails: get a new job.

Model that!

I do this cr@p for a living, and unlike the good professor, I don’t get high on my own supply.

]]> By: Anonymous http://www.nakedcapitalism.com/2008/03/financial-models-should-come-with.html#comment-4958 Anonymous Mon, 10 Mar 2008 09:50:00 +0000 http://www.nakedcapitalism.com/2008/03/financial-models-should-come-with-health-warnings/#comment-4958 Government Wealth Warning: Investing May Seriously Damage Your Wealth. Government Wealth Warning: Investing May Seriously Damage Your Wealth.

]]> By: vlade http://www.nakedcapitalism.com/2008/03/financial-models-should-come-with.html#comment-4956 vlade Mon, 10 Mar 2008 08:38:00 +0000 http://www.nakedcapitalism.com/2008/03/financial-models-should-come-with-health-warnings/#comment-4956 and how many people heed the health warnings on just about anything else?<br/>One of the main problems is that the models can and do drive the market, disengaging it from reality. <br/><br/>And, before anyone says arbitrage, I'd like to point that the arbitrage would then work only if you had unlimited funds. Which is why bubbless are hard to arbitrage away. and how many people heed the health warnings on just about anything else?
One of the main problems is that the models can and do drive the market, disengaging it from reality.

And, before anyone says arbitrage, I’d like to point that the arbitrage would then work only if you had unlimited funds. Which is why bubbless are hard to arbitrage away.

]]>