Company emissions estimates for private markets

About this episode

Private equity has grown to $8.2 trillion in assets and can make up to 20% of a multi-asset portfolio. Yet most private companies have no obligation to report emissions, which leaves investors guessing at the footprint of their holdings. That gap matters more as disclosure rules in California, the UK, Australia, Singapore and New Zealand start to cover private companies.

This episode covers our research into a better way to fill it. We built machine learning models that estimate Scope 1, 2 and 3 emissions for any private company from as little as one financial metric, such as revenue or total assets.

What you’ll hear

Most investors estimate private company emissions with industry emission factors. These give every company in a sector the same intensity per dollar of revenue, so two steelmakers with very different processes end up with the same footprint.

Our models learn from reported emissions instead. Tested against reported data, they were up to seven times more accurate than industry factors. For companies with revenue between $100 million and $1 billion, industry factors overstated emissions by 10 to 100 times.

The episode also covers which financial inputs work best. Revenue gave the strongest results, with total assets a workable fallback.

Read the full research

The white paper sets out the full method, including how we measure uncertainty for each company.

Read the white paper