Market manipulation by rumormongers: Evidence from insiders’ stock selling

Using a large sample of data on insiders’ stock selling and rumors about A-share listed companies in China, this study empirically tests whether and how rumors about companies are used to manipulate the market in the context of insiders’ stock selling. We find that the probability of a rumor’s occurrence, especially that of a favorable rumor, significantly increases in the 30 days before the first transaction in a round of insiders’ stock selling and remains high for 30 days afterward, showing clear signs of manipulation. These results are robust to several endogeneity tests. The probability of manipulation via rumor increases with a company’s degree of information asymmetry. In addition, large-scale stock selling, centralized bidding, and transactions involving CEOs or chairmen (or their relatives) have a significantly higher probability of manipulation via rumor, while transactions made by directors, supervisors, or senior executives (but not their relatives) have a significantly lower probability of manipulation via rumor. Further examination shows that using rumor to manipulate the market increases insiders’ transaction returns but leads to stock price reversal in the long term.