以財務比率及特徵降維預測股票報酬率

The Prediction of Stock Returns Using Financial Ratios and Feature Dimensionality Reduction

廖萬里(Liew Ban Lee)

指導教授 ： 周宗南

朝陽科技大學/管理學院/財務金融系/碩士(2016年)

並列摘要

Value investing is one of the most popular investment strategy for investors to search for the undervalued stocks based on their financial reports and balance sheets. However, the numerous metrics derived from the financial statements are not easy for the investor to analyze and determine the financial health of a company. The main purpose of this study is to employ feature extraction to identify a smaller number of financial ratios for the prediction of stock return which reflects the quality of a company.

Four Data Mining Models approaches, including Multilayer Perceptron Model, Meta Regression Model, Random Forest Model and Random Tree Models were incorporated with feature extraction to evaluate the forecast performance of five different industries in Taiwan. The results demonstrated that the prediction errors were improved for Multilayer Perceptron Model and Meta Regression Model by the feature extraction strategy which reducing the original 16 variables into 5 variables. Finally, this paper concluded that the feature extraction strategy might improve some prediction error, but not suitable in every data mining model.

並列關鍵字

Financial Ratios ； Feature Extraction ； Data mining model

參考文獻

1. Abarbanell, J. S., & Bushee, B. J. (1997). Fundamental analysis, future earnings, and stock prices. Journal of Accounting Research, 35(1), 1-24.

2. Abarbanell, J. S., & Bushee, B. J. (1998). Abnormal returns to a fundamental analysis strategy. Accounting Review, 19-45.

3. Albert, A. A., Blas, N. G., & de Mingo López, L. F. (2015). Natural combination to trade in the stock market. Soft Computing, 1-18.

4. Baradi, N. K., & Mohapatra, S. (2015). The Use of Technical and Fundamental Tools By Indian Stock Brokers. International Journal of Business Analytics (IJBAN), 2(1), 60-73

5. Ballings, M., Van den Poel, D., Hespeels, N., & Gryp, R. (2015). Evaluating multiple classifiers for stock price direction prediction. Expert Systems with Applications, 42(20), 7046-7056.