Impact of Ambiguity on Stock Markets

Quantitative studies have provided evidence showing that ambiguity can help to explain the equity premium puzzle and the excess volatility puzzle of the equity market. In addition, it also plays an important role in the 2008 financial crisis. However, empirical studies remain few. Anderson et al. (2009) develop an empirical measure based on the Survey of Professional Forecasters (SPF). The survey data are collected from part of the professionals in the US finance industry, which might result in biased findings. Viale et al. (2014) develop another empirical measure of ambiguity based on the reference model calculated using the smooth transition autoregressive (STAR) model and assumptions about the confidence level of investors. It may be improper to use the STAR model as the reference model because it is difficult to find out a forecasting model that is used by all investors. As such, the first empirical study in Chapter 3 focuses on high-frequency forecasting using linear AR models, exponential smoothing models and nonlinear AR models. The findings suggest that the best-performing forecasting model changes from one period to another and the STAR model cannot beat the AR model, suggesting that the calculation of the ambiguity measure of Viale et al. (2014) is improper. Therefore, the other two empirical studies in Chapters 4 and 5 develop two new empirical ambiguity measures with inspiration from theoretical works. The results support the theoretical proposition that ambiguity can explain the equity premium puzzle and the excess volatility puzzle. In addition, the degree of ambiguity of the equity market can be affected by investors’ expectations on macroeconomic conditions and default risks. On the other hand, Chapter 5 shows that ambiguity plays an important role in the 2008 financial crisis. Last but not least, the thesis also provides an ambiguity indicator for regulators and financial market practitioners.