Market Analysis for Risk Management and Regulation: An Artificial Intelligence Approach
Roy S. Freedman, Joseph Mathai · 1995
The key points in this chapter are: 1. Purely statistical market analysis techniques for determining risk are incomplete because, in the real world, risk is effected by non-statistical phenomena such as news. Quantitative models can ignore such "shocks," but analysts still need to explain their models in the face of such shocks. 2. Risk must include the effects of all relationships. This is only partially accounted by models using correlation statistics. 3. Market analysis techniques used for risk management have much in common with the techniques used in market surveillance for regulation. In this paper, we provide a framework that can be applied either to risk management for trading or to the regulation of markets. Our framework lets analysts define patterns of market behavior and detect new or hidden relationships between subjects in order to evaluate risk and discover information flows. Introduction From the information technology perspective, Wall Street represents a complex and ...