The Response of Dynamic Herd Behavior to Domestic and U.S. Market Factors: Evidence from the Greater China Stock Markets

ABSTRACT

We investigate the dynamic reaction of stock market herding in China, Hong Kong, and Taiwan to unexpected shocks from domestic and U.S. market factors. In China and Taiwan, herding is more pronounced, and the investors tend to herd with the rising stock market returns. Overconfident investors will herd on the subsequent trading days under market stress. Compared with the response to the domestic market factors, the responses of herding in the Greater China stock market to the U.S. market factors are weaker. After the 2007–8 financial crisis, the U.S. market factors highly explain the forecast error variance of herding in the Shanghai A-share and Taiwan markets.

KEY WORDS:

• financial crisis
• herding
• Kalman-filter-based model
• overconfidence
• VAR model

Notes

1. Balcilar and Demirer (2014), and Balcilar et al. (2013) adopt the Markov-switching model to test the presence of herd behavior under different market states.

2. The terms “Greater China stock market” (Yeh and Lee 2000) and “Greater China Economic Area” (Cheng and Glascock 2005) are both used to describe the Chinese, Hong Kong, and Taiwanese markets.

3. In 2003, the trading value was USD 1.3 trillion for the Hong Kong market, USD 3.7 trillion for the Shanghai stock market, and USD 624 billion for the Taiwan stock market. During the same period, the trading values in the Korean and Singapore markets were USD 1.3 trillion and USD 280.9 billion, respectively. The data are obtained from World Federation of Exchanges Focus Monthly Statistics.

4. We follow on the approach of Longstaff (2010) to decompose the sample period. In Longstaff (2010), the sample period is from 2006 to 2008: the 2006 precrisis period, the 2007 subprime crisis period, and the 2008 global crisis period.

5. To examine market herd behavior, Christie and Huang (1995) propose the cross-sectional standard deviation () method. Because the square return deviation calculated in is sensitive to extreme values, Chang et al. (2000) suggest the cross-sectional absolute deviation () method.

6. Our result concerning the Chinese stock market is consistent with the evidence provided by Chiang et al. (2013) and Tan et al. (2008) and contradicts the report of Demirer and Kutan (2006), where Chinese market herding is insignificant.

7. Before running the VAR regression model, all variables included in the VAR system are examined by the ADF unit root test.

8. We carefully examined the different trading dates between the U.S. and Greater China stock markets.

9. According to Sims (1980), the estimated coefficients in the VAR model are unable to fully reveal the critical interrelationships among variables. Sims analyzes the system’s response to random shocks in the related variables. To prevent the VAR results from being influenced by the variables’ order, we employ the generalized impulses technique proposed by Pesaran and Shin (1998). The results concerning the estimated coefficients of the VAR regression equations are available upon request.

Additional information

Funding

This research is supported by the Taiwan National Science Council (NSC 101-2410-H-390-005).