Prior studies suggest that earnings and non-earnings information can be complementary to each other (Lundholm, 1988). Given the co-existence of both components, a lack of non-earnings information can end up boosting earnings explanatory power on returns in certain circumstances, producing spuriously high earnings transparency (ET) in Barth et al. (2013). This scenario is plausible when insiders are motivated to exploit their information advantage and discretionarily alter non-mandatory disclosure strategy. Conditional on higher insider trading profit, we uncover a positive relation between the firm-specific earnings transparency and crash risk. In addition, the above relation is more pronounced with respect to selling and profitable insider transactions. Overall, we demonstrate a potential dark side of high earnings explanatory power on stock returns, conditional on higher likelihood of non-earnings information hoarding.

We investigate the impact of mutual fund herding behaviours on stock price crashes. There are competing hypotheses with respect to how investors’ herding behaviours are associated with information processing. Our empirical evidence shows that mutual fund herding is associated with a poor information environment and low disclosure quality. More importantly, mutual fund herding amplifies stock price crash risk afterwards. The main finding is concentrated on buy-herding rather than sell-herding. To mitigate the endogeneity concern, we use the 2004 SEC mutual fund disclosure regulation change as an exogenous shock and the results hold. We further use propensity score matching to alleviate the impact of information asymmetry. Finally, additional analysis reveals that our results are not driven by the price impact hypothesis.