Determinants of industry herding

This thesis provides empirical evidence on the determinants of herding in US and China using both market and industry level data. Herding is examined based on market returns, volatility, trading volume and different market conditions, using the CSAD measure on daily data from 1990 to 2016. The findings for the US market demonstrate that herding does not exist. However, some herding becomes visible at the industry level. The results also demonstrate that there is limited evidence of herding during rising and declining markets days, which is more significant on days with low trading volatility and low trading volume. For different market conditions, the finding shows that herding is present at the market and industry level during the Dot com bubble and the Global Financial Crisis. The results for the Chinese markets provide evidence of herding at both the market and industry level, although it is more prevalent in Shenzhen stock exchange. Evidence further demonstrates that industry herding is more prevalent in the Shenzhen stock exchange when the market is declining, the trading volume is high, and volatility is low. After examining herding during the Asian crisis and Global financial crisis, the results demonstrate herding occurs during both crises at the market and industry level. Finally, the findings demonstrate that at the market level, US returns only has an impact on herding in Shanghai stock exchange. The results have implications for financial market investors and stock market regulatory authorities in both markets. For the US, it is important that investors know the impact of industry herding on specific industries, while regulatory authorities should encourage investors to diversify their sector investments. For the Chinese markets, the findings imply that participants in the Chinese stock markets (sectors) are irrational when they make investment decisions. Therefore, regulatory authorities should consider irrationality in their rule-making processes and market reforms