Kelly criterion for optimizing choices in stock and portfolio markets

Abstract

Kelly criterion is known as a useful tool in the stock market. Yet its efficiency is limited due to its simplicity because it was designed to solve binary gambling problems. Luckily, by switching its coefficients and conclude multiple outcomes in other models, the limitations can be receded or solved. This paper aims at how Kelly criterion work on stock market and portfolio statistics and how to improve it by combining with some other theories. The paper will try to use Ralph Vince’s extension and combine other formulas like the Bayes model and get more improved results. By combining more adjustments, the results are shown to be more sensitive to more real factors and to be more optimized, such as integrating a vast range of data and recognizing the potential risk in the stocks’ variation. In conclusion, it is promising that Kelly criterion can be more and more complete in the future with the existence of new theories that can be its adjustments.