Multifactor conditional equity premium model: Evidence from China's stock market
利用中国股市数据，通过变量选择方法识别出条件市场方差、规模市场股价和通胀是股权溢价的关键决定因素，发现简单风险-收益关系存在遗漏变量偏误，且该三因子模型并非数据挖掘产物。
There is mixed evidence of a positive relationship between the stock market risk and return. We reexamine this critical implication of asset pricing theory using fresh data from China's stock market, which is largely segmented from the rest of the global financial market. Using formal variable selection methods and a comprehensive set of predictor variables, we identify conditional market variance, scaled market prices, and inflation as crucial determinants of equity premiums. The estimated simple risk-return relationship exhibits downward omitted variable bias, which underlines the importance of considering multiple factors to explain the variation in equity premiums. We cannot wholly attribute the three-factor conditional equity premium model to data mining, as Guo, Sanni, and Yu (2022) select the same model for the U.S. stock market. These findings challenge existing asset pricing models and provide valuable guidance for future theoretical research.