Research On Investors Oriented Financial Failure Forecasting Of Listed Firms

With the rapid development of society and economy in China after reform and opening, people’s growing wealth leads to expanding demand of financial products. The China securities market, especially the stock market, has become one important option of financial products for Chinese because of return and liquidity. From the open channels, the investors can obtain the financial information and recommendation of listed firms from securities’ analysts. However, most investors are lack of investment knowledge, experience and skills. They cannot make investment decisions efficiently with that information above. It is easy for them to loss or go bankruptcy in investments. There are a lot of investors. If those people fail in investments, it not only affects the development of personal family life, but also affects the stability of the society when it becomes seriously.In view of the above-mentioned facts and the basis of previous studies, this paper is finished from the perspective of investors. The information about listed firms that can be obtained by investors is treated as the cutpoint. Then, using support vector machine, soft set theory, fuzzy-soft set theory, uni-int decision making, and empirical research, quantitative analysis, qualitative analysis and comparative analysis together, we have done the research on investors oriented to select forecasting variables for listed firms. And the research of financial information oriented forecasting model of financial failure for listed firms, the research of recommendation oriented forecasting model of financial failure for listed firms, and the research of financial information and recommendation oriented forecasting model of financial failure for listed firms have been finished respectively.First, the information that investors can obtain about listed firms is different sample sizes of financial information and recommendation. This paper emphasizes selection of quantitative financial information forecasting variables for financial failure. The frequently-used models for financial ratios selection——Expert System and Statistical Analysis methods have their own difficulties in application. Considering fast changes in managerial environment and investment environment of a company, a novel Logit Model and soft set theory oriented parameters reduction method(NSS) is proposed to select financial ratios, aim at optimizing forecasting results and satisfying demands of investors. By this way, we select 9 financial ratios from early warning index. Sales net profit rate, management fee rate, turnover of total capital and comprehensive lever are the first time to be considered as forecasting indexes of financial failure for listed firms. Empirical results shows that based on truthful data of Chinese listed firms, the soft set oriented parameters reduction method(NSS) can efficiently improve failure forecasting results under different sample sizes. Combined with the recommendation, this paper constructs the forecasting index set for financial failure for listed firms.Secondly, some investors prefer making investment decisions with financial information of listed firms, but they may be in the confusion that how to deal it with different sample sizes and different moving trend of various financial ratios, so a novel financial information-oriented forecasting model based on Logit model(LR) and Support Vector Machine(SVM) of financial failure for listed firms is proposed. With the help of scatter diagram methods, financial ratios are divided into two groups according to the different moving trends: linear moving trend ratios and nonlinear moving trend ratios. LR and SVM are employed to deal with linear moving trend ratios and nonlinear moving trend ratios respectively. And the forecasts are combined with Residual Support Vector Machine(RSVM). Empirical results show that based on truthful data of Chinese listed firms, compared to other frequently-used forecasting models, the proposed model can effectively improve the forecasting performance and assist investors to make decisions with financial information, especially when the sample sizes are small.Again, some investors prefer making investment decisions with recommendations of listed firms, but they may be in the confusion that how to deal with different sample sizes and the recommendation because so less attention is paid to this topic that the number of suitable forecasting models is small, a novel recommendation combining fuzzy soft set theory and D-S evidence theory forecasting model of financial failure for listed firms is proposed. It is proposed firstly to select securities analysts who supply the recommendation about each listed firms. Then, fuzzy soft set method is employed to compute the degree of membership. So, the qualitative recommendation is translated into the quantitative information. The recommendation can be integrated with the D-S evidence theory. After the search of optimal threshold value, the forecasting model of financial failure for listed firms is finished. Empirical results show that based on truthful data of Chinese listed firms, the proposed forecasting method can effectively improve the forecasting performance with the recommendation and all sample sizes based on the perspective of investors and assist investors to make decisions with the recommendation and all sample sizes.Finally, based on the research above, in consideration that some investors prefer using financial information and recommendation together to make investment decisions, but they may be in the confusion how to deal with the financial information and recommendation together because so less attention is paid on this topic that the number of suitable forecasting models is small, a novel financial information and recommendation oriented forecasting model combining soft set theory, fuzzy soft set theory and uni-int decision making is proposed. Empirical results show that based on truthful data of Chinese listed firms, compared to other forecasting models, the proposed method can settle with, forecast, then go together with combination forecasting, and adopt decision making theory that deals with uni-int status data and different sample sizes to effectively improve the forecasting performance with financial information and recommendation together with all simple sizes and help investors make investment decisions.