| Previous studies on stock market bubbles have developed theoretical models showing that the stock market bubble is a determinant of the stock price, and the presence of bubbles is also supported by empirical evidence. Based on these results, this dissertation further explores the effects of stock market bubbles on fixed investment and financial market.;The first research question focuses on testing the hypothesis that stock returns are more sensitive to investor sentiment during stock market crashes than during stock market booms. The empirical results confirm that sentiment betas are asymmetric across stock market cycles.;The next research question aims to examine the dynamic effects of stock market misvaluation on firm fixed investment. We apply a Bayesian vector autoregression (BVAR) model to calculate the impulse response function of investment to misvaluation shock. And we find that investment responds 47%-55% at maximum annually to one standard deviation of misvaluation.;Finally, we address why stock return volatility is typically higher after the stock market falls than after it rises (referred as asymmetric volatility). By maximizing the likelihood functions of dividends and stock reruns from a quadratic generalized autoregressive conditional heteroscedasticity (QGARCH) model, we decompose this asymmetric volatility into the volatility feedback effect due to dividend news, and the bubble effect explained by bubble news. |
