Value of climate change news: A textual analysis

Abstract

There is a growing consensus that climate change poses a financial material risk to investors. In the face of escalating climate change risks, investors are seeking strategies to safeguard their portfolios. Building on a growing literature that combines textual analysis and dynamic hedging approach, we propose a method to construct portfolios of publicly traded assets that dynamically hedge climate change risk in the Norwegian stock market.

As climate risk is not directly observable, we apply Latent Dirichlet Allocation to extract news on climate change from more than 400,000 articles published in the Norwegian newspaper Dagens Næringsliv between January 2013 and February 2022. Using these data, we develop the DN Climate Change News Index and use innovations in this index as a hedge target. The hedge portfolios are constructed using third-party environmental scores from MSCI and Sustainalytics, along with firm-level data of equities listed on the Oslo Stock Exchange.

The DN Climate Change News Index show high correlations with international counterparts, indicating its ability to capture major global climate events and assessing the intensity of climate change–related news coverage. However, despite a positive out-of-sample correlation, the mimicking portfolio approach fails to construct efficient hedge portfolios against innovations in the index. Hedge portfolios based on Sustainalytics E-scores show a 0.21 out-of-sample correlation with innovations in the DN Climate Change News Index, indicating that the index may provide a relevant signal for investors.