Behavioural heterogeneity and equity premium volatility in China

ABSTRACT

This paper develops a simple heterogeneous agents model with fundamentalists and chartists. It examines whether investors’ behavioural heterogeneity is related to excess volatility of equity premium in the Chinese stock market. To calculate deviations between the realized and expected values of the equity premium, we use consumption growth model and dividend growth model to estimate the expected equity premiums of CSI 300 from January 2005 to December 2019. After estimating the heterogeneous agents model using the CSI 300 monthly data, we find that investors’ heterogeneity in price trends and trading strategies can significantly explain excess volatility of equity premium. Our analysis of two significant fluctuations, specifically the booms and busts of 2007–2008 and 2014–2015, further corroborates our key finding that investors’ behavioural heterogeneity plays an important role in explaining excess volatility of equity premium.

KEYWORDS:

• Equity premium volatility
• behavioural heterogeneity
• heterogeneous agents model
• strategies switching

JEL CLASSIFICATION:

• G12
• G10

Disclosure statement

No potential conflict of interest was reported by the authors.

Notes

¹ The econometric form of Equation (1)(1) $x_t=\frac{1}{R^{\ast }}(n_{f,t}{E}_{f,t}[x_{t+1}]+n_{c,t}{E}_{c,t}[x_{t+1}]),$(1) can be expressed as

11 $x_t=\frac{1}{R^{\ast }}(n_{f,t}{\varphi }_f+n_{c,t}{\varphi }_c)x_{t-1}+{\in }_t,{\in }_t\sim N(0,{\sigma }^2)IID,$11

where ${\in }_t$ is the error term, which is unobserved to agents when making expectations $E_{i,t}[x_{t+1}]$.

² Fama and French (2002) estimates the expected equity premium on the S&P 500 portfolio from the dividend growth model. Huang, Zhou, and Zhang (2020) estimates the expected equity premium on the CSI 300 portfolio from the consumption growth model and the dividend growth model, respectively.

Additional information

Funding

This work was supported by the National Natural Science Foundation of China [Grant No. 71790594], Guizhou Key Laboratory of Big Data Statistical Analysis [Grant No. BDSA20200112], GUFE Special Program for Key and Urgently Needed Disciplines [Grant No. 2020ZJXK21], and the Key Program of Guizhou University of Commerce [Grant No. 2020YJZD002].