Stock Selection Using Association Rules on Cash Flow and Accrual Financial Indicators
Amontep Wijitcharoen, Praisan Padungweang, Sanit Sirisawatvatana, Bunthit Watanapa · 2018
Traditionally, decision makers use accrual-based financial indicators, which are based on historical information, to evaluate stocks. This passive information can provide only partial intelligence to the decision makers. Taking cash-based insights as integral part of analytic would yield a more complete information to robustly indicate the performance of a company. In this study, we propose a systematic framework for financial information mining that actively considers the interaction among traditional financial ratios and cash flow signaling in identifying stocks that outperform the average market return in the arriving period. Association detection method is adopted to ensure the mined rules be reliable and practical. The experimental results show that cash flow is a needed integral component in all mined rules that effectively indicate outranked stocks.