October 31, 2016
Sir, as an Executive Director at the World Bank, in October 2004, in a formal statement I wrote: “Phrases such as “absolute risk-free arbitrage income opportunities” should be banned in our Knowledge Bank. We believe that much of the world’s financial markets are currently being dangerously overstretched through an exaggerated reliance on intrinsically weak financial models that are based on very short series of statistical evidence and very doubtful volatility assumptions.”
So clearly I cannot but to be in total agreement with the general concept of Cathy O’Neil’s “Weapons of Math Destruction” and as reported by Federica Cocco in “A manual for citizens taking on the machines” October 31.
But, I say “general agreement”, because I am not against math per se, just against math and models when applied to the real world by desk-bound mathematicians (and economists) who have never walked on main street; and are perhaps pursuing ulterior motives.
Take for instance credit evaluations. Do you think there is a lot of buyer interest in data that provides for a perfect credit scoring? Never so. Perfect credit scoring does not produce big profits. Big profits are the result of selling something very risky as very safe. Big profits are the result of convincing a debtor that data proves he is much riskier than what he really is.