Identification of high performing equities using financial characteristic attributes

Long-term investors are interested in identifying the characteristics of companies that are likely to triple in value over the next five years, which equates to return of approximately 25 percent per year over the period. Such companies are known as compounders due to their high compounded rate of return. This paper reports on an analysis of corporate and market data undertaken with the goal of identifying specific characteristics that tend to separate compounders from other equities. We consider a wide ranging set of characteristics designed to bear on corporate financial performance. Characteristics are constructed using historical stock prices and financial reports from the years December 1992 to June 2014. The analysis includes both high and standard market return equities. Classification techniques such as Naïve Bayes, logistic regression, and random forests are then applied to the characteristics to try and identify high return equities. The research has the goal of identifying the characteristics of high return equities so that in the future these characteristics can be used to filter investment candidates using classification techniques and models.