The impact of herding behaviour of Chinese securities investment funds on firm value: the mediating effects of stock price informativeness and agency cost

ABSTRACT

This paper investigates the impact of herding behaviour of Chinese Securities Investment Funds (SIFs) on firm value by taking stock price informativeness and agency cost as intermediary variables. It constructs LSV model to measure herding behaviour of SIFs, and develops logarithmic conversion method of the goodness-of-fit to characterize stock price informativeness. The stepwise regression method is employed for empirical analysis and it is found that: (i) The herding behaviour of SIFs significantly negatively affects the stock price informativeness, and the former is ‘intentional herding’ that ignores personal private information. (ii) The stock price informativeness is positively related to firm value, while agency cost has a significant negative influence on firm value. (iii) The stock price informativeness and agency cost play significant partial mediating effects between the herding behaviour of SIFs and firm value, respectively.

KEYWORDS:

• Herding behaviour
• stock price informativeness
• agency cost
• firm value

JEL CLASSIFICATION:

• C33
• G34

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹It can be found that if the coefficients $\beta$, $\delta$, ${\beta }_1$, and $\beta { }^{\mathrm{\prime }}$ are both significant, the partial mediating effect of mediating variables is significant.

²Stock price informativeness is an index to measure the degree to which firm-specific information is incorporated into stock price. If the herding behaviour of SIFs is ‘intentional herding’, fund managers will ignore their own private information to follow the investment strategies of others, resulting in the private information they owned is not fully integrated into stock price, thus reducing the stock price informativeness.

³Following Yao and Zou (2016), the lagged term of endogenous explanatory variable (EEV) can be used as instrumental variable for the following two reasons. First, EEV is associated with its lagged values. Second, since the lagged value of EEV has already occurred (from the perspective of the current period, its value has been fixed), it may be ‘predetermined’ and not related to the current disturbance term.

⁴The coefficients $\beta$, $\delta$, ${\beta }_1$, and $\beta { }^{\mathrm{\prime }}$ are −11.9256, −15.7285, 0.0525, and −11.5924, respectively, and they are all significant.

⁵The coefficients $\beta$, $\delta$, ${\beta }_1$, and $\beta { }^{\mathrm{\prime }}$ are −6.3339, 0.1277, 2.8878, and −6.6979, respectively, and they are all significant.

Additional information

Funding

This work was supported by the National Social Science Foundation of China [grant number 20XGL003].