A New Approach in Financial Modelling with the Aid of Artificial Neural Networks

Abstract

In recent years there has been significant interest in Artificial Neural Networks (ANNs) and their suitability for the solution of many real-world problems. When simulating complex systems which display highly nonlinear behaviour, traditional methods prove insufficient either due to their oversimplified problem description or due to the lack of available computer resources. ANNs have been shown to be quite successful when applied to such problems whose dynamics may not be explicitly understood. The use of ANNs has become more widespread with increasing computing processing power. In this work we present a brief overview of the ANN model developed and its validation and concentrate on specific applications in the area of financial modelling, where the network manages to reconstruct with quite high accuracy the behaviour of a very complex problem, such as the prediction of the performance of a bank stock. This constitutes the first step towards the creation of a framework which enables mathematical models to benefit from the ability of ANNs to resolve complex trends.