Order Book Based Decision Support Model for Traders

Abstract

Abstract Stock market trading is a complicated task when considering the profit earning because stock market behavior is dynamic and complex. Kelly criterion is an optimal strategy which is used for betting systems and also for stock market investing or trading. There are several models which have modeled Kelly Criterion to the stock market. However, those models are only theoretical models with lack of practical validity. Practical behavior, dynamicity of the stock market and giving clear decisions to traders are not concerned in the existing models implemented using Kelly criterion to stock market. Also validation of models with real stock market have not being conducted. In this thesis a dynamic computational model of Kelly criterion for stock traders is built. Selecting one suitable model from the above models and order book and trades data are used in order to solve the above stipulated problems. Four cases are introduced on how short term returns behave like normal. Then model considered normal distribution of the short term returns. After investigating the trades with chosen Kelly criterion to stock market model, the methodology and equations on how to handle trades with the existing Kelly model of stock market is introduced. Introduction of short term traders or day traders to the Kelly model of stock market is carried out. Also, after investigating trades with the chosen Kelly criterion to stock market model, introduction of methodology and equations on how to use order book with the existing Kelly model of stock market is conducted. Then, validation of the model with the real stock market behavior and comparison of the accuracy of the model without modifications to the model with order book is conducted. Developed dynamic model provides stock buy or sell decisions to a day trader.