Bankruptcy has long been an important topic in finance and accounting research. Recent headline bankruptcies have included Enron, Fannie Mae, Freddie Mac, Washington Mutual, Merrill Lynch, and Lehman Brothers. These bankruptcies and their financial fallout have become a serious public concern due to huge influence these companies play in the real economy. Many researchers began investigating bankruptcy predictions back in the early 1970s. However, until recently, most research used prediction models based on traditional statistics. In recent years, however, newly-developed data mining techniques have been applied to various fields, including performance prediction systems. This research applies particle swarm optimization (PSO) to obtain suitable parameter settings for a support vector machine (SVM) model and to select a subset of beneficial features without reducing the classification accuracy rate. Experiments were conducted on an initial sample of 80 electronic companies listed on the Taiwan Stock Exchange Corporation (TSEC).
This paper makes four critical contributions: (1) The results indicate the business cycle factor mainly affects financial prediction performance and has a greater influence than financial ratios. (2) The closer we get to the actual occurrence of financial distress, the higher the accuracy obtained both with and without feature selection under the business cycle approach. For example, PSO-SVM without feature selection provides 89.37% average correct cross-validation for two quarters prior to the occurrence of financial distress. (3) Our empirical results show that PSO integrated with SVM provides better classification accuracy than the Grid search, and genetic algorithm (GA) with SVM approaches for companies as normal or under threat. (4) The PSO-SVM model also provides better prediction accuracy than do the Grid-SVM, GA-SVM, SVM, SOM, and SVR-SOM approaches for seven well-known UCI datasets. Therefore, this paper proposes that the PSO-SVM approach could be a more suitable method for predicting potential financial distress.
Artificial neural network (ANN) model has been used for years to conduct research in stock price prediction for three reasons. First, it has a higher prediction accuracy rate in empirical research. Second, it is not subject to the assumption of having samples from a normal distribution. Third, it can deal with non-linear problems. Nevertheless, the accuracy of prediction relies on the parameter settings of neural network as well as the complexities of problems and the neural network architecture; the results of the analysis could be even more significant with the selection of optimal parameters and network architecture. Currently, as a way of setting parameters, most researchers employed the trial and error method. However, this method is very time-consuming and labor-intensive and may not result in the optimal parameters. Therefore, this research took advantage of a back propagation neural network (BPNN) for the purpose of parameter optimization through constructing a model of stock price prediction, applying design of experiment (DOE) to systematize experiment scheduling, and methods of main effects analysis and interaction analysis. The research used two datasets of financial ratios from 50 blue chip companies in Taiwanese stock market and 40 listed American banks in New York stock exchange as experimental samples. Research results showed that the correlation forecasting, root mean squared error (RMSE), and computing time, which can effectively increase the accuracy of stock price prediction, are better than traditional statistical methods and conventional neural network model.