Security analysts’ earnings forecasting performance based on information transmission network

• Published in: Physica A Vol. 509; pp. 611 - 619
• Main Authors: Hou, Jianlei, Zhao, Shangmei, Yang, Haijun
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.11.2018
• Subjects: Earnings forecasting performance; Information connection; Network analysis; Research reports; Information connection; Network analysis; Research reports; Earnings forecasting performance
• ISSN: 0378-4371; 1873-2119
• DOI: 10.1016/j.physa.2018.06.093

Security analysts’ research reports are important information resources in financial markets. To investigate the relationship between information connection among analysts and earnings forecasting performance, we construct an information transmission network based on all research reports issued on stocks of Shanghai Stock Exchange and Shenzhen Stock Exchange from 2008 to 2016. Visual network figure presents the change of analysts’ network position during the sample period. Moreover, information connection stability is defined to assess network stability, which is negatively correlated with stock market volatility. Empirical results show that analysts with a more central position in the network will get better earnings forecasting performance with less forecasting volatility and higher forecasting accuracy. •We construct an information transmission network using analysts’ research reports.•Visual diagram shows the change of analysts’ network position from 2008 to 2016.•The degree distribution of the networks obeys right skew power-law distribution.•We assess the information connection stability based on May–Wigner stability theorem.•Analysts with a more central network position have a better forecasting performance.