Does Institutional Knowledge Drive the Information Wedge in Emerging Markets? Evidence from Analysts of Local and Foreign Brokerage Firms in China

Professor, The Chinese University of Hong Kong

Prof. Tianyu Zhang's CV.

Using a topical-modeling technique (Latent Dirichlet Allocation) in textual analysis, we examine the contents of the analyst reports for 403 firms listed in the domestic markets of China. We find that these contents can explain the information advantage of analysts from local brokerages (local analysts) over their counterparts from foreign brokerages (foreign analysts), as captured by their relative forecast accuracy. The local analysts’ superior forecast accuracy is positively associated with the proportion of topics related to politics and government policies in their reports. This positive association is stronger among firms that are either state-owned or receiving government subsidies. However, when the firms that the analysts follow have high overseas sales (top 5% in the sample), foreign analysts can outperform local analysts in forecast accuracy and the foreign analysts’ superior accuracy is greater as they report more on topics pertaining to financial and profitability ratios of the firms.