The influence of network platform interaction on corporate total factor productivity: evidence from China stock exchange investor interactive platforms

Purpose The aim of this paper is to study the impact of the questions and answers (Q&A) between investors and enterprises from the China stock exchange investor interactive platforms on the total factor productivity (TFP) of enterprises. Design/methodology/approach To show how the interaction influences the TFP of enterprises, the authors select Q&A records from the interactive platforms related to production, R&D and technology through the Latent Dirichlet Allocation (LDA) topic model and choose A-share listed companies from 2010 to 2019 in China as a sample. To treat the data and test the proposed hypothesis, the authors applied OLS regression and endogeneity testing methods, such as the entropy balance test, Heckman two-stage model and the two-stage least squares regression. Findings This paper finds that interaction between investors and enterprises is positively correlated with TFP, and that improvements in content length and the timeliness of response can promote TFP. Interactive behavior mainly improves the TFP of enterprises by alleviating financing constraints and encouraging enterprises to increase R&D investment. This positive effect is more pronounced in companies with higher agency costs, non-high-tech companies and companies not supported by industrial policy. Originality/value The novelty of the research stands in the application of Python's LDA topic model to screen out Q&A records that are directly related to TFP, such as production, R&D, technology, etc., and measures the degree of information interaction between investors and enterprises from multiple dimensions, such as interaction frequency, content length and the timeliness of response.