Research

Every investor can see what a company is worth. Few can see what it delivers. Markets price a company's value and spend heavily on the data behind it. Nobody prices the other side: the tonnes of greenhouse gas a company's products will avoid, the lives its vaccines will save, the clean power it will add. Benemetry builds that missing side.

AI made it possible to research these outcomes company by company. Deterministic code turns the research into one figure: euros of market value for each unit of real-world result. So an investor can ask a question the sustainability data of today cannot answer. How much real-world outcome stands behind each euro of market value?

How we measure

• By product line. We measure what a company's products do in the world, never what the company says about its own impact.
• In physical units, across sectors. A tonne of greenhouse gas avoided is the same tonne in every sector, so one figure can travel from one sector to another.
• Reproducible. Every figure is computed in deterministic code: the same inputs give the same answer every time. For our research we publish the data and the code, so your own analyst can rerun a result.

The paper

"Measuring Carbon Avoidance Efficiency: Six Orders of Magnitude in Climate Equities" (2026), J. Pazos. A working paper on SSRN. DOI 10.2139/ssrn.4623238.

Replication data and code

The paper's data and code are on Zenodo under a Creative Commons Attribution 4.0 (CC BY 4.0) license. DOI 10.5281/zenodo.19675432. The deposit holds:

• the full set of 34 physics proxy models, one for each sector
• anonymized inputs for 187 firms
• panel data for 2024Q1–2025Q4
• a Python calculator in deterministic code. It gives back every Climate Efficiency (CER) value in the paper.

Author

Jay Pazos, ORCID 0000-0002-9943-7805.

The themes and climate sectors we measure today are on What we offer.