Original title: Breaking the pooling bias: a hybrid unsupervised and causal framework for corporate sustainability
A single model can hide differences between companies. The researchers group firms with similar sustainability patterns using K-Means, a method that sorts items into clusters. They then study the drivers within each group and use SHAPA method for explaining how much different factors contribute to a model's result., a tool for explaining how factors contribute to results. The analysis finds different sustainability pathways, including diminishing returns from environmental policies among top-performing firms.
Imagine making dinner for a group, with everyone following the same recipe. Some people need less salt, while others need more time in the oven, so one set of instructions does not work equally well for everyone. In this paper, grouping firms by similar sustainability patterns is like noticing which diners need which adjustments, then checking what changes help each group.
Regulators and corporate leaders may be guided poorly by one rule for every firm. This study finds that sustainability drivers differ between groups, suggesting that effective policies may need to fit the firms they target.
The abstract reports patterns from a global corporate panel, but does not describe the panel's time span, the measures used, or how the results were validated. It also does not establish that the identified patterns apply to every firm or that localized policies cause better outcomes.
Ask anything about this paper. Answers come from its abstract, and say so when the abstract doesn't cover your question.
Breaking the pooling bias: a hybrid unsupervised and causal framework for corporate sustainability — Abdul Kadar Muhammad Masum, Md. Abul Kalam Azad, Najmus Saadat — Finance Research Open — doi:10.1016/j.finr.2026.100185
Abstract licence: CC BY
Explanation generated by Synthesis from the abstract — it is not the original paper.