A team of researchers from Bloomberg Quant Research and Amazon Web Services claim they can estimate corporate greenhouse gas emissions with relative precision.
In a recent paper, the researchers claimed that, by training a machine learning model on disclosed greenhouse gas (GHG) emissions, they are able to estimate the emissions of other companies globally who do not disclose this information.
In April this year, the European Commission adopted a new package of climate impact measures, hoping it would improve the flow of capital towards sustainability initiatives.
However, without precise insight into greenhouse gas emissions, investors will struggle to remain compliant. And according to the paper’s authors, only a “small portion” of companies voluntarily disclose their emissions.
To train the machine learning model, the researchers used more than 1,000 features and more than 24,000 rows of disclosed emissions from 3,960 companies, provided in the past two decades. Various areas were covered, including carbon emissions and the use of various resources, human rights and diversity and inclusion, management structure, executive compensation, employee relations, just to name a few.
“In this paper, we show that our model provides accurate estimates of corporate GHG emissions to investors such that they are able to align their investments with the regulatory measures and achieve net-zero goals,” the report concludes.
- These are the best free software for small businesses available now