Introduction

Climate Action 100+ (CA100+) was launched in December 2017 as the world’s largest investor coalition on climate change, representing what has been described as the “biggest shareholder action plan ever”. At its peak in 2023, the initiative was supported by over 700 investors with a combined 68 trillion US dollars in assets under management (AUM). However, following the departure of several large United States asset managers in 2024, such as BlackRock US and Goldman Sachs Asset Management, the current combined AUM is estimated at 47 trillion US dollars. The coalition engages with approximately 170 focus companies, which are identified as the world’s largest corporate greenhouse gas emitters, to ensure they take necessary action on climate change.

Hypotheses development

The research tests three primary hypotheses regarding the effectiveness of investor coalitions. The first (H1) posits that inclusion in the CA100+ focus list improves climate action relative to comparable companies. The second (H2) suggests that the effectiveness of the coalition increases over time as its influence and monitoring tools, such as the Net Zero Company Benchmark, expand. The third (H3) argues that companies are rational actors likely to prioritise low-cost climate actions, such as disclosure, over high-cost measures like substantial emission reductions. This creates a potential discrepancy between environmental reporting and actual performance.

The CA100+ company selection process

The coalition’s focus group consists of two main subsets. The “CA100 companies” comprise the 100 largest publicly listed corporate emitters, selected based on reported and estimated emissions from the CDP database. This selection represents an exogenous shock as companies could not self-select. In June 2018, an additional 61 companies were added as “Plus companies”, described as “transition enablers”. Unlike the initial list, the Plus companies were selected based on strategic importance and regional balance, involving prior investor knowledge which introduces potential selection bias.

Data and descriptive statistics

The study employs multidimensional metrics to assess climate action. For disclosure, it uses the ClimateBERT-TCFD model to analyse sentence-level content in annual reports across 402 companies. For actual performance, it utilises the Sectoral Decarbonisation Approach (SDA) through the Transition Pathway Initiative (TPI) to calculate sector-specific carbon intensities for 218 companies across six high-emitting sectors: airlines, automotives, cement, electricity, steel, and oil and gas. Additionally, new primary data was collected to evaluate the ambition of forward-looking carbon emission reduction targets for the years 2025, 2035, and 2050.

Research design

The study employs a Difference-in-Differences (DiD) analysis to determine the causal impact of CA100+. By using company fixed effects, the model controls for time-invariant characteristics like pollution levels and firm size. The primary assumption is that treated and untreated companies followed parallel trends prior to the initiative’s launch. A non-staggered two-way fixed effects (TWFE) model is used as the baseline, complemented by a staggered DiD specification to account for heterogeneous and dynamic treatment effects over time.

Results

The findings indicate limited effectiveness of the coalition across several metrics. There was no statistically significant impact on companies’ climate-related disclosure according to TCFD recommendations, with the DiD coefficient standing at 0.14. Similarly, no impact was observed on historical carbon intensities between 2014 and 2022. However, a significant treatment effect was found for medium- and long-term target setting. Closer examination reveals this was driven exclusively by Plus companies, where the negative treatment effect (indicating more ambitious targets) was significant at the 1% level for 2035 (-0.55) and 2050 (-1.06). No impact was observed for short-term targets (2025).

Robustness checks

Several robustness checks were conducted to validate these findings. These included using alternative proxies for disclosure, such as CDP responses, and integrating country-level data from the Climate Change Performance Index (CCPI) to control for varying national regulatory environments. The results remained robust when restricting the sample to North American companies and when repeating the analysis within a single sector, electricity utilities, which provided the largest sub-sample size.

Conclusion

Overall, the study finds only partial support for the effectiveness of investor coalitions. While CA100+ appears to influence long-term decarbonisation targets, the lack of short-term milestones raises concerns about credibility and the potential for “backloading” emission reduction efforts. The research suggests that investors may have selected Plus companies based on their existing propensity to set targets, highlighting the importance of investor selectivity. The study concludes that collaborative engagement should not replace direct policy measures, such as carbon pricing, to achieve global climate targets.