The Power of Text: How News Sentiment Influences Financial Markets
IMF Blog, December 16, 2019
Source details
- Canonical URL
- The Power of Text: How News Sentiment Influences Financial Markets
Other formats
Bibliographic details
- Authors: Damien Puy
- Published: December 16, 2019
Overview and motivation
- Author: Damien Puy
- Date: December 16, 2019
- Research question: How do investors react to news, particularly given (1) innovations in information technologies that increase the reach and speed of financial and economic news, and (2) the opening of financial markets—especially in emerging markets—to foreign participants?
- Key technologies used: natural language processing, text mining, big data.
Data and methodology
- Dataset: more than 4 million articles published by Reuters worldwide between 1991 and 2015 covering economic, financial, corporate, and political news.
- Text-mining approach: assessed tone of each article as bullish or bearish by identifying positive terms (examples: “gains,” “recovery,” “confidence”) and negative terms (examples: “crisis,” “losses,” “decline”).
- Constructed measure: a daily news-based sentiment index for both advanced and emerging markets.
- Defined “foreign news” as news involving multiple countries and their interrelationships; “local news” involves a single country.
Key findings
- Sudden changes in news sentiment have a significant impact on asset prices worldwide.
- Media tone is a very good proxy for investor sentiment.
- Foreign news (and foreign investors) play a larger role than local news (and local investors) in driving local asset prices.
- Optimism in global news sentiment generates a strong and permanent impact on asset prices around the world.
- Optimism in local news produces a more muted and only temporary effect on asset prices.
- The news-based “global news sentiment index” mirrors other measures of global risk aversion, such as the CBOE Volatility Index (VIX), which represents market expectations of volatility over the coming 30 days and is often called the “fear index.”
- The news-based index is a better predictor of future movements in international asset prices than the VIX.
Implications for practice and surveillance
- Monitoring news tone in real time is an effective way to capture sudden changes in investor sentiment that may not be captured by market-based indicators.
- Big data and text-mining techniques can provide actionable inputs for economic and financial research and for institutional daily work, including financial surveillance.
Source: The Power of Text: How News Sentiment Influences Financial Markets — Damien Puy, December 16, 2019.
