Model selection, market capitalization heterogeneity and ESG asset pricing: an empirical study of Chinese A-shares under the LSY four-factor framework
Abstract
This study investigates whether factor-model selection drives the mixed evidence on ESG pricing in China's A-share market, where "green discount" and "green premium" coexist. Taking the Liu–Stambaugh–Yuan (LSY) four-factor model as the benchmark pricing framework and combining Shangdao Ronglv ESG ratings with CSMAR data, it examines model dependence, market capitalization heterogeneity, policy asymmetry, and risk transmission. The results show that ESG pricing conclusions are strongly model-dependent: the estimated ESG premium reverses sign once the local LSY factors and firm fundamentals are controlled, and statistically significant evidence of a premium is concentrated among large-cap stocks. The 2016 green finance policy is associated with dimensionally asymmetric changes in corporate ESG performance, although pre-existing trends limit strict causal identification. The environmental and governance dimensions transmit through opposite stock-price-risk paths, a pattern consistent with a dimensional hedging interpretation. These results imply that credible ESG pricing evidence in the A-share market requires an empirical specification—and above all a pricing benchmark—suited to the Chinese market.