INSIGHT by Jitendra Aswani, a corporate governance fellow at Harvard University
Today, organizations such as Securities and Exchange Commission (SEC), investment companies, media, and others have considerable interest in the disclosure and eventual reduction of US firms’ carbon emissions. Investors and investment companies are interested because they want to know whether emissions reduction by portfolio firms can contribute to greater expected stock returns and better operating performance. Catering to such demand, influential papers in this area finds strong associations between emissions and fundamental measures of firms’ financial performance such as stock returns, operating profitability, and Tobin’s Q. The carbon emissions literature cumulatively proposes two economic arguments – regulatory risk and investors preferences, for positive association between emissions and stock returns. Regulatory risk because if the government likely to take action to combat climate change in “bad” (high-emissions) states of the world, then there is a risk of an increase in the cost of capital for high-emissions firms. This risk captures factors such as potential carbon taxes or mandated remedial pollution clean-up costs. On the other hand, high-emissions firms can also become risky with increased cost of capital if the investors preferences changed as some investors may choose to shun companies in “brown” industries, on the grounds that firms in such industries cause substantial harm to society. If a large enough set of investors choose to avoid high-carbon stocks, then, as in the case of sin stocks (such as stocks of alcohol, tobacco, gambling and similar firms), it should follow that “brown” stocks earn excess returns because a subset of investors shuns them.
In my recently published article in Review of Finance, “Are Carbon Emissions Associated with Stock Returns?”, co-authored with Shivaram Rajgopal (Columbia University) and Aneesh Raghunandan (London School of Economics), we delve deeper into the collective body of evidence regarding emissions and valuation in our current study and argue that past papers in this area rely on (i) specific research design choices, most notably a reliance on unscaled emissions, to draw conclusions rather than a measure of emissions relative to firm sales (i.e., emissions intensity), and (i) the assumption that vendor-estimated carbon emissions are accurate in the sense that they do not systematically differ from firm disclosed carbon emissions, we delve deeper into the collective body of evidence regarding emissions and valuation in our current study and argue that past papers in this area rely on (i) specific research design choices, most notably a reliance on unscaled emissions, to draw conclusions rather than a measure of emissions relative to firm sales (i.e., emissions intensity), and (i) the assumption that vendor-estimated carbon emissions are accurate in the sense that they do not systematically differ from firm disclosed carbon emissions.
Key takeaways from this paper are as follows:
1. Concern over differences in vendor-estimated emission numbers and actual disclosed emission numbers.
2. From a firm and investor’s perspective, carbon intensity rather than unscaled emissions is a better measure to evaluate carbon risk. Unscaled emissions would be an appropriate measure from the regulatory perspective.
| Why is there concern over vendor-estimated emission numbers?
Our primary finding highlights that the association between stock returns and emissions in the United States, as previously documented in research, is primarily driven by emissions estimates provided by vendors, rather than actual emissions disclosed by companies. While a strong relationship exists between vendor-estimated emissions and stock returns, there is limited evidence of a connection between emissions and stock returns for firms that disclose their actual emissions values. Our empirical analysis reveals systematic discrepancies between emissions figures disclosed by vendors and those estimated by companies. This finding holds significant importance for both researchers and practitioners since more than 70% of emissions data in commonly used US emissions databases are estimated by vendors instead of being voluntarily disclosed by firms. Furthermore, in recent years, there has been a substantial expansion in data coverage (e.g., since 2016 in the Trucost database we utilized). However, the vast majority of this expanded coverage relies on vendor-estimated emissions rather than emissions disclosed by companies. Consequently, studies focused on recent years, during which carbon risk has gained greater political significance, are particularly susceptible to this issue.
>>click to zoom in | Figure 1: Relationship between raw material used, goods produced, and carbon emitted in the production process.
| Why to prefer carbon intensity over unscaled emissions?
Another issue arises in prior work: an emphasis on the relation between unscaled carbon emissions (i.e., total carbon emitted) and returns. While total emissions may be a good measure of economy-wide carbon risk, this is not a suitable measurement choice for understanding individual firms’ carbon risk; using a firm’s emissions intensity (the ratio of emissions to revenue) is a better way of measuring this. Carbon emissions arise from a firm’s core operations, therefore total emissions primarily show how much raw material it has used or how many goods it has produced and sold (see Figure 1). To this end, within-firm variation in unscaled carbon emissions is almost entirely driven by variation in units of goods produced and sold (Figure 2). Hence, we argue that on a standalone basis, a relation between unscaled emissions and stock returns can only be interpreted as evidence of a relation between a firm’s revenue and its stock market performance. Conversely, emissions intensity—the ratio of emissions to net sales, a metric also commonly used in practice to assess progress toward decarbonization without sacrificing output—better captures a firm’s emissions performance by avoiding mechanical correlations with firm size.
>>click to zoom in | Figure 2: Relation between Emissions and Sale
To test whether carbon emissions are actually associated with stock returns and financial performance, we used data from a sample of 2,729 US firms (and European firms for external validation) from 2005-2019. We obtained emissions information from Trucost, which we merged with CRSP and Compustat for information on stock returns and firm characteristics. Using this merged data, we analyzed the relation between emission measures (intensity versus unscaled emissions) and stock returns accounted for vendor-estimated vs firm-disclosed actual emissions. we did not find any evidence of a relationship between Scope 1, 2, or 3 emissions and stock market or fundamental financial performance.
Our findings suggest that responsible investors, policy makers, and academics may want to be cautious in interpreting correlations between a firm’s carbon emissions and valuation constructs (e.g., stock returns) or fundamental accounting data (operating profitability).
We take no position on whether disclosing and/or cutting emissions is desirable or not. We only argue that calculating emissions intensity (emissions by revenue) is a more suitable way to capture a firm’s carbon risk and that, based on this measurement, there is no correlation between carbon emissions and stock returns (i.e., no evidence of a carbon premium yet). Nevertheless, we warn that increase in carbon prices can impact this relation. Researchers and practitioners should also be careful in using vendor-estimated emissions figures that may fundamentally alter inferences about the link between emissions and financial performance.
However, we believe that firms disclosing their emissions on a widespread scale, as well as greater truth in advertising by emissions-related data vendors, will mitigate this latter issue.
Jitendra Aswani is a corporate governance fellow at Harvard University. He has a Ph.D. in Finance from Gabelli School of Business, Fordham University. Before joining the Fordham, he did his M.Phil. (Economics) from Indira Gandhi Institute of Development Research (IGIDR) and B.S. (Chemical Engineering) from National Institute of Technology (Jaipur) in India. He has also worked for the Indian Oil Corporation (the national oil company of India) for a small stint.
His research focuses on corporate governance, ESG / CSR, corporate finance, debt markets, and behavioral finance. He has presented his research at well-known finance and economics conferences such as The Econometric Society meetings, SFS Cavalcade North America, Western Finance Associate (WFA), American Finance Association (AFA, Doctoral Session), MIT Asia Conference, Financial Management Association (FMA), China International Conference in Finance (CICF), Royal Economic Society (RES) Annual Meeting, Northeast Universities Development Consortium (NEUDC), UN PRI Academic Conference, and similar others. His work has published in top finance journals such as Review of Finance.
| All opinions expressed are those of the author and/or quoted sources. investESG.eu is an independent and neutral platform dedicated to generating debate around ESG investing topics.