Overconfidence, market liquidity and speculative bubbles
Guang-ping SHI,Xiao-xing LIU,Deng-bao YAO,Xu ZHANG
DOI: https://doi.org/10.13587/j.cnki.jieem.2018.03.008
2018-01-01
Abstract:The violent fluctuations in the stock market, such as the one that took place in June 2015 in Chinese stock market, and another serious stock market crash in May 2007, have raised controversy about the existence of stock market bubbles in China. The evolution mechanism and influence factors of asset price bubbles have been an important research focus in the fields of economy and finance. Many theoretical researches, such as rational and irrational speculative bubbles models, and empirical researches based on Markov regime-switching model, have been used to explain and test the determinants and existence conditions of asset price bubbles. Investors' overconfidence and market liquidity have important impact on stock market bubbles, although the existing theoretical studies have found that the investors' overconfidence has a positive effect on stock market speculative bubbles. However, it hasn't been empirically tested. Although there are only a few researches about the influence of liquidity on stock bubbles that focuses on empirical analysis, systematic theoretical analysis is still scarce. This paper systematically examines the contribution of investors' overconfidence and market liquidity to the stock market speculative bubbles. Specifically, this study theoretically examines how varying degrees of investors' overconfidence and market liquidity influence stock market speculative bubbles as well as the inherent mechanism. A variables extending VNS three-regime-switching model is constructed to describe the dynamic evolution mechanism of the bubbles. Finally, empirical tests of the Chinese stock market bubbles are also made based on the proposed model. The research methods and main conclusions are included as follows:First, a theoretical framework about the impact of investors' overconfidence and market liquidity on the stock market speculative bubbles is established. By setting overconfidence parameter and market liquidity parameter, the expression of stock market speculation bubbles is deduced. This finding reveals that both investors' overconfidence and market liquidity have positive effect on stock market speculative bubbles. The bubble's size increases with investors' overconfidence, and decreases with market illiquidity, but the effect of influence on bubbles varies with the degree of investors' overconfidence and market liquidity. Second, macro-economic variables, such as industrial added value, consumer price index and 7-day inter-bank lending rate, are chosen to measure the fundamental of stock market. Then we examine linkages between stock market bubbles and macro-economic variables by co-integration regression tests as well as build error correction model to extract the bubbles. The idea of measuring investors' overconfidence is separating the trading volume changes due to overconfidence from changes due to other factors. The overconfidence is calculated by estimating ARIMAX model. The empirical evidence indicates that the estimates of bubbles and overconfidence according to the above methods are consistent with the practice. Third, the empirical design focuses on the three-regime-switching model with time-varying transition probabilities by including two supplementary overconfidence and market liquidity components, in order to capture the dynamic process of the Chinese stock market bubbles, and test the explanatory power of overconfidence and market liquidity for the regime-switching. Empirical analysis is carried out using the related monthly data of CSI 300 Index from April 2005 to June 2015,and the results show that compared with the other existent Hamilton and VNS models, the variables extending VNS three-regime-switching model has a better result on describing the dynamic process of bubbles. It is obvious that the bubbles in CSI 300 Index can be distinguished into three regimes, namely, dormant, explosive and collapsing. In addition, the partition of different regimes is much more clearly both before and after the 2008 financial crisis. The investors' overconfidence and market liquidity have the significant explanatory power for regime switching. The investors' overconfidence increases the probability of stock market bubble transforming from dormant to explosive regime. The negative market liquidity increases the probability of stock market bubble collapsing.