Select Page

Using AI To Assess the Impact of News on Markets


By Pierre Haren and Eric Jensen, Causality Link

In a world of information overload, how can investment managers cut through the noise and use news media to their advantage? The answers start with assessing the impact that news has on share prices— and it requires tapping into the power of AI.

Using the media to inform investment decisions is not a new concept, of course. But determining the true impact of news stories – from articles that leave room for a range of different interpretations, contain several different angles and fit into any number of industry or macroeconomic trends – has historically been difficult, limiting their utility.

Until now, that is. In a recent study in collaboration with AI research platform Causality Link and the Toulouse School of Economics, French asset manager Amundi found that when incorporated into a long-short strategy, company-specific fundamental news can produce positive excess returns up to the day after publication. This report is a key milestone for anyone seeking to learn where and how investment intelligence can be gleaned from the media.

The key is to analyze this information programmatically and at scale. Using the power of AI, Causality Link ingests thousands of news articles and other written content each day, extracts key performance indicators (KPIs) and links them to their respective companies within a vast database, along with the direction (positive or negative) and tense (past, present or future) of the news. Across thousands of documents, these “causal links” can be aggregated into powerful news sentiment signals.

Amundi’s study revolved around how, when and to what extent these signals are reflected in share prices. For each trading day, the firm constructed two portfolios. One consisted of stocks with the most positive news signals for that day, and the other shorted stocks with the most negative signals for that day. This process was carried out repeatedly to assess a range of criteria beyond positive or negative direction alone, including volume of coverage, timing of news publication, category and tense of the news, company size and more.

The result: these signal-driven portfolios outperformed the market significantly, averaging an overall excess return of +1.3% per day. These positive effects persisted into the following day, though average returns were significantly lower at +0.04%. That’s a powerful endorsement of using news media sentiment to drive investment decisions – and the findings become even more revealing as you drill deeper. Let’s explore three of the most compelling.

Focus on the Future

The study found that companies’ share prices react more strongly to news about their future than to revelations about their present and past achievements. This follows logically. Many market participants will assume that the present and past are already “priced in,” while future predictions invite more opportunity for speculation. In addition, the study found that news on the near-term future will impact prices more than that relating to the long-term future.

ESG Exaggeration

While share prices moved significantly based on news relating to company finances and fundamentals, their reaction to news on environmental, social and governance (ESG) issues was negligible. In fact, negative ESG news had a slightly positive effect on prices, suggesting a lack of causation. This makes sense because even most ESG-focused investments are financial in nature. Nobody invests to lose money – any ESG investment worth its salt has the potential to be profitable.

This article was previously published in RealClear Markets. To read the rest of the article go to Using AI To Assess the Impact of News on Markets | RealClearMarkets


Pierre Haren and Eric Jensen co-founded Causality Link, an AI-powered research platform that helps investors make better decisions faster.