Emotional response modeling in financial markets : Boston Stock Exchange data analysis

Abstract

In this thesis, physiological data is analyzed in the context of financial risk processing, specifically investigating the effects of financial trading decisions and situations on the physiological responses of professional market makers. The data for this analysis comes from an experiment performed on market makers at the Boston Stock Exchange. This analysis involved significant preprocessing of large financial and physiological data sets. Short-term and long term analysis of financial and performance based event markers of the data are performed and the results interpreted. There are two main conclusions. First, negative performance events are found to be the the main driver of physiological responses; positive performance events have minimal deviations from baseline physiological signals. Second, a long term analysis of events yield more substantial physiological changes than a short term analysis.