An econophysics study on financial markets

Abstract

The methods used in the study of statistical physics and non-linear dynamics have helped understand several phenomena in economic systems. One of the primary interests of economic theory in recent times has been to understand the behaviour of financial markets in times of stress. In this context, the emergence of frustration in the correlation network of stocks can indicate some interesting phenomena. During the course of this project, the Hong Kong stock exchange (HSX), NASDAQ and New York Stock Exchange (NYSE) were particularly looked into. Although HSX and NASDAQ did not show breakdown in structural balance, such breakdown was seen in NYSE during the great recession of 2007-2009. In addition to this empirical study, an attempt to understand the pricing of financial assets using agent-based models was also undertaken. Based on models already in literature, it was attempted to develop reinforcement learning models to portray the action of financial traders in the market.