Financial Risk Management Limits: Big-data and Risk of Modelling and Analysis

Abstract

The explosion of information in the financial industry, combined with an increase in storage, calculation and analysis capacities, explains the financial industry's infatuation with Big Data. This technology allows financial organizations to use more data and to perform a much larger and faster methodical analysis, to minimize risks and to accompany future fluctuations generated by the different national and international economic conditions. Nevertheless, this development could not only reduce the security of operations, but it also represents a major challenge in the analysis and modeling of financial risks when predicting the fluctuation of a financial product or making decisions at the managerial level. In this paper, we highlight some of the challenges related to the processing and modeling of these mega-data, relying on a critical literature review, so as to be able to criticize traditional models of stock price fluctuations (financial market normality). We also discuss some resolutions to statistical methods and approaches that can lead to an adequate quantitative management of financial risks. We propose the use of the bootstrap methodology in the financial area, which presents a technique that intervenes precisely when the scope of the problem considered is not covered by classical methods or when the conditions of application of the previous methods are no longer valid.

Keywords: Big Data; financial modeling; financial risk management; asymptotic methods; the bootstrap.

JEL Classification: G17

Paper type: Theoretical Research