A Neural Network Analysis of Accounting Variables and Stock Price

The Case of Real Estate Companies in the Philippines

Abstract

This research investigates the relative importance of various accounting variables in explaining stock price changes for the top three real estate companies listed on the Philippine Stock Exchange (MEG, SMPH, and ALI) in 2023. Using artificial neural network (ANN) analysis, the study focuses on key accounting variables such as current ratio, return on assets, return on equity, net profit margin, operating profit margin, and debt-to-equity ratio. Data was collected from quarterly financial reports and stock price figures from 2018 to the third quarter of 2023. The ANN model was implemented in SPSS 25 with a feedforward back-propagation multilayer perceptron. Results indicated that the net profit margin is the most significant predictor of stock prices across all three companies, highlighting profitability as a crucial factor. The operating profit margin is particularly important for MEG, while the current ratio is most critical for SMPH. The debt-to-equity ratio showed moderate importance across all companies. These insights highlight the varied influence of accounting variables, with profitability, operational efficiency, leverage, and liquidity playing key roles in stock price determination for different companies, aiding in investment decisions and strategic planning.