How financial models become self-negating

Taylor Spears
Work package: 
WP 8
Publication number: 
01 January 2012

ABSTRACT: Mathematical models are essential to the practice of both financial valuation and risk management. Yet in certain circumstances, the use of a model by market participants can cause markets to behave in ways contrary to the model's predictions. Models are thus codified bodies of knowledge that under certain circumstances are capable of becoming “self-negating”. Recently economists and sociologists have noted this property of financial models, yet there has been little systematic classification of social and market conditions under which models can become self-falsifying, and how these conditions interact with the technical properties of models themselves. The purpose of this paper is to review the relevant economics and sociology literature and build an initial taxonomy for this purpose. The paper focuses on two broad classes of models: statistical models and arbitrage-free models. The primary implication of this research is that regulators and central banks must be attentive to the ways in which different classes of financial models can lead to market instability in order to effectively intervene where it is appropriate to do so.

FINNOV DP8.2769.0 KB