Handle with care: Challenges in company-level emissions data for assessing financial risks from climate change

Climate data play an important role for market actors and regulators to assess climate-related vulnerability. The most important quantitative class of such data are carbon emissions as almost all metrics to analyse carbon exposure relate to carbon emissions of companies and countries. This paper provides a detailed analysis of the quality of carbon emission data, points out the most common data flaws, and offers suggestions for a robust empirical analysis. Using a large data set of company-level carbon emissions, we show that year-by-year analysis of the consistency of company emissions is required to identify data flaws. Also, we find that economic and carbon data are not perfectly synchronised. As all carbon-emission metrics suffer from similar data inconsistencies robustness of results is not achieved by using several such metrics. Thus, our findings serve as a warning for the reliability of emission data reporting and their unreflected use in empirical analyses.