|Author Name||CHEN Cheng (University of Hong Kong) / SENGA Tatsuro (Fellow, RIETI) / SUN Chang (University of Hong Kong) / ZHANG Hongyong (Fellow, RIETI)|
|Creation Date/NO.||March 2018 18-E-010|
|Research Project||Studies on Firm Management and Internationalization under the Growing Fluidity of the Japanese Economy|
|Download / Links|
This paper uses a unique dataset of Japanese multinational affiliates, which contains information on sales forecasts, to detect information imperfection and learning in the international market. We document three stylized facts concerning affiliates' forecasts. First, forecast errors (FEs) of sales decline with the affiliate's age. Second, if the parent firm has previous export experience to the region where its affiliate is set up, the entering affiliate starts with a smaller absolute value of FEs. Third, FEs of sales are positively correlated over time and this positive correlation becomes stronger when the affiliates are located further away from Japan. In total, we view these facts as direct evidence for the existence of imperfect information and learning in the international market. We then build up and quantify a dynamic industry equilibrium model of trade and foreign direct investment (FDI), which features information rigidity and learning, in order to explain the documented facts. Counterfactual analysis shows that the variance of time-invariant demand draws and that of transitory shocks have qualitatively different and quantitatively important implications for dynamic patterns of trade/multinational production, dynamic selection, and aggregate productivity.