What could explain the lack of value added by SBTi membership on the stock exchange? One could argue that external stakeholders reward companies that have joined credible VEPs due to concerns about greenwashing.²⁸. To be credible, VEP sponsors should have technical capabilities, have no conflicts of interest with their members, and have credible monitoring and compliance functions in the program design. Despite being sensitive to these issues, the SBTi has been criticized in recent years for the validity of its verification methods. For example, a recent study points out that the SBTi allows companies to set a base year themselves – artificially high base year emissions could make it easier for companies to claim emissions reductions in the future^29,30Our results may therefore reflect the stock market’s skepticism towards the SBTi’s promises.

However, we recognise that companies could benefit from the reputational advantages of Climate-VEP membership in ways other than their stock market returns. Future work should examine whether SBTi membership impacts revenue, employee retention or regulatory costs. However, our findings provide a note of caution to proponents of voluntary commitments who may have overstated their profit impact. This poses a problem for climate leaders who have tried to use economic logic to mobilise business support for climate action.

Even if stock markets do not support voluntary climate pledges by companies, we find that they do not hurt stock prices either, as some conservatives argue. Based on the argument that climate action hurts profits, several U.S. states such as Texas, Florida, Louisiana, West Virginia and Mississippi have banned their state pension funds from investing in ESG-linked assets. Yet the lack of a negative stock market reaction may not be enough to motivate companies to seriously invest in climate action on purely economic grounds. One could argue that the political rationale must be found outside of established financial criteria.³¹ in terms of social and political legitimacy^32,33Future work should investigate whether non-financial legitimacy could be sufficient to motivate voluntary climate protection actions by companies without economic gains.

Although we examine SBTi-related climate actions that focus on emissions reduction, we recognize that other corporate climate policies could aim to increase the resilience of companies’ operations or supply chains to climate-related disruptions.³⁴Some work suggests that bond rating agencies reward cities that participate in VEPs that focus on adaptation, as opposed to VEPs that focus on mitigation.³⁵the rationale being that adaptation creates private benefits, while mitigation benefits bring public benefits. Future work should examine whether the stock market reaction follows a similar pattern to bond ratings.

Our study has several limitations. As mentioned above, we focus only on S&P 500 companies. Future work should go beyond these companies, especially because they are subject to less external scrutiny and media exposure. In addition, it is less clear whether medium-sized or even small companies would have the capacity and willingness to voluntarily take climate action given the considerable costs involved. If the smaller companies are connected to larger companies through supply chains, voluntary climate action could be spread through supply chain links, as documented in the case of other VEPs such as ISO 14001 and Responsible Care.^11.36.

VEPs can create positive benefits for their members or protect them from negative events (such as product recalls, industrial accidents, lobbying, regulatory issues, lawsuits, etc.), the so-called reputational cushion. Scholars have investigated whether or not industry-wide VEPs can provide a reputational cushion. Future work should address whether cross-industry climate VEPs such as SBTi provide reputational insurance to companies on climate issues.

Because our main results rely on the matching techniques, they may highly depend on the covariates used for matching prior to estimation. Future research should expand the range of covariates for matching while minimizing the bias from missing observations through a multiple imputation approach.³⁷.

Data and methods

S&P 500 companies and their SBTi membership status

We focus on S&P 500 companies because, given their size, they are likely to think carefully about how their corporate reputation affects stock prices.³⁸Regarding their SBTi membership status, companies can join (or not join) the SBTi in several ways:

Stage 1: 1.5 degree target: Companies commit to setting emission reduction targets that are consistent with limiting global temperature rise to below 1.5 degrees Celsius compared to pre-industrial levels (102 companies).

Stage 2: Well below 2°C target: Companies commit to setting emission reduction targets consistent with limiting global temperature increase to well below 2°C compared to pre-industrial levels (25 companies).

Stage 3: 2°C target: Companies commit to setting emission reduction targets consistent with limiting global temperature increase to 2°C compared to pre-industrial levels (5 companies).

Level 4: Committed: Companies join the SBTi by committing to undertake target verification in the near future but are still in the process (56 companies).

Level 5: None of the above: Companies have not joined the SBTi (312 companies).

Empirical strategy

We analyze stock prices based on the SBTi membership status of companies in three ways. First, we use event study analysis^39.40a widely used approach to study stock market performance in response to a specific corporate event. Our goal is to assess whether SBTi members’ stock prices increase after joining this climate VEP. In particular, in line with the literature, we focus on abnormal returns of stock prices while accounting for general market price fluctuations.

Since event study analysis does not allow for comparison between firms (i.e., does not allow for comparison of firms with and without SBTi membership), we use two additional approaches to estimate the impact of SBTi membership on firms’ stock prices. Unlike event study analysis, which uses daily stock prices, we focus on quarterly stock prices because most firm indicators (control variables in our models) are reported quarterly. In the second approach, we use linear regression after matching observations (i.e., firms that joined SBTi with those that did not) on their covariates using a coarsened exact match⁴¹In the third approach, we use a weighted two-sided fixed effect model^42.43We obtained information on all covariates, including quarterly stock prices, from the COMPUSTAT—Capital IQ North database of the Wharton Research Data Service.

There are several advantages to using matching techniques and weighted two-sided fixed effects models. SBTi membership imposes significant joining and compliance costs for firms. This creates a selection problem: firms with certain unobserved attributes may be more able and willing to seek SBTi verification. Therefore, simply comparing stock prices across firms may pose an endogeneity problem. Matching techniques attempt to alleviate this problem by ensuring homogeneity across comparison groups based on observed covariates. Because simple matching does not account for the time variance of observed covariates, we use weighted two-sided fixed effects using propensity score weighting techniques with unit and time-varying covariates.

Compared to ordinary two-sided fixed-effects models, the weighted model allows researchers to estimate a difference-in-differences estimator with multi-period treatment status across observations. Because the S&P 500 firms in our sample joined the SBTi at different time periods, we consider the weighted two-sided fixed-effects model to be better than ordinary two-sided fixed-effects models.

In line with the literature, we compare and control for several firm-specific characteristics. First, stock prices might reflect firm size⁴⁴. Therefore, we use total assets, total liabilities and total market value as comparison and control variables. Second, stock prices fluctuate based on the current performance of companies^45.46. Therefore, we use return on assets, change in current sales (compared to past sales which have declined by one quarter) and current dividends per share. Third, stock prices depend on the ability of companies to repay their debts⁴⁷which is reflected in the cash ratio (total assets plus total cash holdings divided by current liabilities) and the debt ratio (total amount of long-term debt in liabilities divided by total assets). Finally, we use a one-quarter lagged stock price as an additional covariate to account for the potential rigidity in stock price behavior⁴⁸.

Table 3 shows descriptive statistics of the covariates listed above. All covariates are in millions of dollars unless stated as a percentage (%), except for dividends per share in dollars.