The application of bayesian networks to approximate the probability distribution of returns of equities

This project examines an alternative technique for calculating the probability distribution of returns of equities and assets consisting of equities, such as exchange-traded funds (ETFs) and equity indexes. In particular, Bayesian networks are used, which allow conditioning upon prior information and are capable of evolving and adapting to current market trends through learning processes. Posterior probability return distributions from Bayesian networks are then used to make inferences about the future price movement of a security. This includes foremost the direction of predicted movement, and secondly, the magnitude of the predicted movement. This project explores the application of these return distributions to develop a trading strategy. This strategy will be used to test whether the perceived benefits of Bayesian methods in finance are realized in producing more accurate and insightful return distributions that can indicate, with significance, an asset's return.