**3. Data, Variables, and the Empirical Model**

#### *3.1. Data Source*

The research sample is obtained from the Chinese Stock "A" markets (Shanghai Stock Exchange and Shenzhen Stock Exchange). The sample period ranges from 2000 to 2019. Data to measure companies' market value (denoted by "MV") and control variables are obtained from the WIND system and the China Stock Market & Accounting Research (CSMAR) system. We consider 2013 the start date of China's CET policy, as Liu et al. (2015) did [51]. We take Beijing, Shanghai, Tianjin, Chongqing, Hubei Province, Guangdong

Province, and Shenzhen as trading pilots. The carbon price data are obtained by initially crawling the daily transaction price data of the carbon market from the exchange websites of each pilot region and then taking the annual average value. The patent and green patent data of listed companies used to measure their innovative activities come from the Chinese Research Data Service (CNRDS) platform, containing multiple sub-databases. The ones used in this study are the China Innovation Research Database (CIRD) and the Green Patent Research Database (GPRD). We derive the patent data of the listed companies from the CIRD. The green patent data of listed companies conform to the GPRD. We use the evaluation data of listed companies' social responsibility reports by Hexun index to represent companies' carbon disclosure [52]. Finally, we eliminate the sample data of listed companies in the financial industry, as their businesses do not involve substantial carbon emissions [42,53]. Listed companies marked with ST, \*ST, or PT are deleted from the sample. Additionally, we exclude the sample data of listed companies whose asset-liability ratio is greater than one and eliminate incomplete or omitted sample data, such as company registration location [47]. Ultimately, the final sample consists of 3283 firms.
