Blame me? Blame you? Blame Math?

This morning in my email inbox: Amazon recommends Lecturing Birds Flying. Can Mathematical Theories Destroy the Financial Market?, Pablo Triana. I haven't read the book but it refers to N. Taleb's Fooled by Randomness and Black Swan, which I read. Book description: Leading and contrarian thinkers have been talking for years about the conflicts between theoretical and real finance. Nassim Taleb first addressed the issue in his technical treatise on options, Dynamic Hedging. Now, in Lecturing Birds on Flying, Pablo Triana moves the conversation to a narrative that anyone can follow.
In FbR Taleb seconds Unmberto Eco's  insight: It is not important what we know, it is much more important what we DON'T know?  

Yes, true, but most of the mathematicians know that mathematical models do only approximate reality. BUT, if you try to understand, and better solve, a problem you are lucky, if you find an analytic solution (a mathematical model, which can be evaluated "correctly"), if not try statistical analysis and if you even cannot find adequate statistical schemes you might be forced to use data-driven approaches, like machine learning. Usually the approximation quality decreases, but also the understandability. Mathematics does not only compute it also speaks. 
I hold Introduction into Mathematical Finance (only available in German at the moment), an educational book Andreas Binder wrote recently (together with Albrecher, Mayer) in hands. It has 160 pages. Without all the mathematical expressions it probably required 10.000 pages to explain the complex content from yield-duration-convexity to intensity modeling in credit risk theories? And the authors mention with each model about its assumptions, limitations and oversight (including dynamic hedging, ..).
BTW, IMO, the crisis is not a Black-Scholes crisis, it is a crisis of highly correlated high-volume-high-velocity deal spaces (herd behaviour) and leverage. Agreed, we shall not lecture birds flying, but risk management shall watch them flying and understand swarm behaviour ... quantitatively

No comments:

Post a Comment