Abstract
This article focuses on financial ‘model risk’ supervision as a test case for a reflexive approach to the sociology of contemporary financial markets. Model risk is customarily defined as a statistically significant relation between the expectation of a trading loss by a firm and its strategic use of, somehow flawed, econometric models for asset pricing and market trading purposes. It made its first public appearance in the mid-1990s, as a component of a new generation of financial risk management systems for the financial derivatives industry. Since then it has been assimilated by the most sophisticated national and international financial regulatory bodies. A perfect illustration of the thesis of the progressive ‘embedding of the economy into economics’, the forensic practice of financial reliability trials (backtesting) faces a deep pragmatic dilemma: how to distinguish truly unpredictable error from negligent risk management behaviour in a wildly randomized social environment.

This publication has 2 references indexed in Scilit: