A Study Of The Influence Of Investor Sentiment On The Early-Warning System Against The Stock Market Crisis

The financial crisis, especially the stock market crisis occurs frequently along the trend of world economic integration and financial liberalization, with the scope and extent evolve gradually from regional and temporary to global and cyclical. The economic costs, as well as the social costs and latent losses caused by the stock market crisis, are unusually large. As the world’s second largest economy, China can not be an exception against these loses. Considering the fact that China’s macroeconomic situation is relatively specific, and the domestic stock market development process is also unique, therefore the monitoring and early warning of potential stock market crises is quite important and urgent.Starting from the point of investor sentiment of the Behavioral Finance, and upon the basis of discussion about the feasibility of measuring the investor sentiment and the necessity of establishing a early warning systems against the stock market crisis, this paper focuses on the impact of investor sentiment on the quality of early warning system against the stock market crisis. The primary job is to build the investor sentiment index of China’s stock market using the principal component analysis. Selecting the economic data and the Shanghai A-share transactions data from July 2005 to June 2015 as the sample, the paper choose several proxy variables include the Shanghai Composite Index, quantity of new accounts, closed-end fund discount, market volume, turnover rate, P/E ratio, Consumer Confidence Index. The investor sentiment index built is proven to have a consistent trend with the stock market and also possess an obvious time leading superiority. Then the paper tested the crisis warning ability of investor sentiment using univariate Logit model. Results showed that there was a significant positive relationship between the investor sentiment and the stock market crisis. To be precise, the probability that the soaring investor sentiment trigger a crisis is about 1.88 times of that when the investor sentiment is low, indicating that sentiment is a significant factor in the stock market crisis warning system. This article furtherly selected macroeconomic variables and stock market variables which have a significant role in early warning to construct a comprehensive crisis warning model for the stock market, also using the univariate Logit model. Finally the empirical results of multivariate warning Logit model shows that adding the investor sentiment into the crisis early warning model based on common macro variables could significantly improve the accuracy of the model of crisis early warning which will reasonably provide an effective early detection and warning against potential crises in the stock market.After a in-depth study on early warning of investor sentiment against the stock market crisis, this paper draw a further conclusion that the investor sentiment possess a time leading superiority in the situation that macroeconomic represented by GDP shows an overall growing development. The overall effectiveness of early warning model is significant. While in the case of deceleration in macroeconomic development, time leading superiority of the investor sentiment is no longer obvious even lag behind the stock market, and the effectiveness of the warning model is significantly reduced. However in such cases if we remove the macroeconomic factor to repeat the empirical process, the effect of early-warning models could be significantly enhanced. The conclusions above could theoretically and practically release an important reference not only for the investors to make rational investment decisions but also for the policy-makers to conduct effective supervision.