Classification Models for Stock's Performance Prediction: A Case Study of Finance Sector in the Stock Exchange of Thailand

Within stock market, the objective of both stock buyers and sellers is to make profit on price difference based on their expectation on a company's current and future value. There is no investing strategy that is considered to be the best by experts, and investing decision criteria remain contingent upon an individual investor's experiences and bias. To address the challenges, this paper uses financial ratios and company's industry data of stocks in finance sector of the Stock Exchange of Thailand (SET) to construct two classification models predicting (i) whether stock's one-year return will outperform the SET Index and (ii) whether the return will be positive. Various classification techniques, including Logistic Regression (LR), Decision Tree (DT), Linear Discriminant Analysis (LDA), and K-Nearest Neighbor (KNN) are explored, whose performances are evaluated by Area Under the Curve (AUC) of the Receiver Operating Characteristic (ROC) curve. Based on the AUC values, we can conclude that, on average, all methods are comparatively good, with LR and LDA being the most useful classifiers for risk averse investors - as both are not subject to uncertainty due to true positive counting bias.