Models and data in finance: les liaisons dangereuses

In my paper, I examine the role that models play, and the relation between models and data, in finance, in particular in stock markets. I discuss several dangerous liaisons between models and data both from a theoretical and a practical viewpoint and their effects on the behaviour of the financial systems and their actors. I argue that these relations and liaisons defy the way traditional philosophy of science accounts for models and the relation between models and data, as stock markets exhibit several dynamics and features that do not fit them. For instance, they challenge the ontological issue about models (the debate about the fictional character of models), or the way model and phenomena and connected, and consequently undermine classical taxonomies as ‘models of data’, ‘model of phenomena’, ‘models of theory’. I then argue that these features open the way to possible exploitations and manipulation of stock market’s dynamics by means of appropriate use of models and data, including ‘reverse finance’, which should be closely analysed in order to contribute to better functioning and serving of the financial systems.