Thesis

On stock trading using an exponentially moving average-PI controller in an idealized market

The aim of this study is to explore the application of an exponentially weighted moving average-PI Controller for the purpose of stock trading. Previous research has been conducted on stock trading using a pure PI controller that resulted in a robust expectation property: the expected value of the cumulative profit is positive for most cases. However, by applying the controller proposed in this paper, we are able to generalize previous results to allow for more emphasis on recent data which may carry greater influence in the behavior of the stock price. Thus the exponentially weighted moving average-PI controller more heavily weights recent data. While the use of models to predict the behavior of stock prices are heavily used in industry, they have proved to be unreliable. The use of classical feedback control does not require the use of models, hence we work in a model-free system. As shall be shown, this lends itself to opportunities such as the robust expectation property.

Relationships

Items