Abstract:This paper investigates the relationship between idiosyncratic volatility and the risk of a stock price crash in China. Using a large sample of Chinese listed firms from 2007 to 2018, the evidence shows that stock price crash risk is significantly positively associated with idiosyncratic volatility. Moreover, firms with a higher level of idiosyncratic volatility are more likely to experience a crash in stock price. The results are robust to alternative measures of idiosyncratic volatility and stock price crash risk and the findings advance our understanding about the role of short-selling constraints in managing the risk of a stock price crash.
What problem does this paper attempt to address?
This paper aims to explore the relationship between idiosyncratic volatility (IVOL) and the risk of stock price crashes. Specifically, the author uses a large sample of data from Chinese A - share listed companies from 2007 to 2018 to study how idiosyncratic volatility affects the risk of stock price crashes. The main finding of the paper is that there is a significant positive correlation between the risk of stock price crashes and idiosyncratic volatility, that is, companies with higher idiosyncratic volatility are more likely to experience stock price crashes. In addition, the research results are still robust when using different measurement methods of idiosyncratic volatility and stock price crash risk, and further analysis shows that short - selling restrictions play an important role in managing the risk of stock price crashes.
### Main research content
1. **Definition and measurement of idiosyncratic volatility (IVOL)**:
- Use the Fama - French three - factor model to estimate the idiosyncratic volatility of individual stocks.
- Calculate the standard deviation of residuals as a measure of IVOL.
2. **Measurement of stock price crash risk**:
- Use two main indicators: down - up volatility (DUVOL) and negative conditional skewness (NCSKEW).
- DUVOL is defined as:
\[
\text{DUVOL}(j,t)=\ln\left(\frac{(n_u - 1)\sum_{\text{down}}W_{j,t}^2}{(n_d - 1)\sum_{\text{up}}W_{j,t}^2}\right)
\]
- NCSKEW is defined as:
\[
\text{NCSKEW}_{j,t}=-\frac{n(n - 1)^3}{2}\frac{\sum_{\text{down}}W_{j,t}^3}{(n - 2)(n - 1)\left(\sum_{\text{up}}W_{j,t}^2\right)^{3/2}}
\]
3. **Sample selection and data processing**:
- The sample includes companies listed on the Shanghai and Shenzhen A - shares from 2007 to 2018.
- Exclude companies with incomplete data, those under special treatment, and financial and utility companies.
- The final sample contains 8,217 company - year observations.
4. **Empirical analysis**:
- Use multiple regression analysis to analyze the impact of idiosyncratic volatility on the risk of stock price crashes.
- Control variables include company size (Size), book - to - market ratio (Bm), turnover rate (Turnover), the return rate of the previous year (Ret), leverage ratio (Lev), and return on net assets in the same period (Roe).
- The regression equation is:
\[
\text{Crash}=\alpha+\beta\text{IVOL}_{j,t}+\sum_k\gamma_k\text{Control}_{k,j,t - 1}+\theta_t+\delta_j+\epsilon_{j,t}
\]
5. **Robustness test**:
- Use different measurement methods of idiosyncratic volatility (such as the Carhart four - factor model and the Fama - French five - factor model) for regression analysis.
- Use alternative measurement methods of stock price crash risk (such as the negative jump risk based on weekly returns exceeding 3.09 standard deviations) for regression analysis.
### Conclusion
The paper finds that there is a significant positive correlation between idiosyncratic volatility and the risk of stock price crashes, and this conclusion still holds under different measurement methods and control variables. The research also emphasizes the role of short - selling restrictions in exacerbating the risk of stock price crashes, providing an important reference for investors and regulatory agencies.