Private equity reached a record-high of $8.2 trillion in 2023. That makes it a significant asset class for investors worldwide. It also makes it a significant polluter.

Private Equity represents as much as 20 per cent of a multi-asset portfolio. But globally its emissions footprint of around 3 billion tonnes is more than a quarter of the public equity footprint of ~11.4 billion tonnes.

Emissions reporting requirements for investors and other funders are ramping up globally in 2025. Our latest research demonstrates an emissions modelling approach which overcomes the issue of data scarcity for unlisted companies, which makes it so difficult for private equity investors to accurately assess their climate risk.

The work also suggests that the market's reliance on sector-based averages may be leading to systematic errors in emissions estimates. This has important implications for investment decisions, risk assessment, and climate strategy development that rely on these estimates.

We find the Emmi machine learning approach outperforms the traditional ‘industry factors’ estimation approaches by as much as seven times. It is also more consistent and reliable across a range of industries, especially in sectors with complex operational patterns and diverse emission sources.