News intensity and asset returns: the case of currency volatility

Ilanit Avioz, Haim Kedar-Levy, Crina Pungulescu

Research output: Contribution to journalArticlepeer-review


There are both theoretical reasons and empirical evidence for financial markets rewarding investors who put effort into acquiring relevant information. This article shows how a systematic approach of encoding text, ‘semantic fingerprinting’ can be applied to a set of news headlines from The Wall Street Journal to measure the ‘news intensity’ − the volume of relevant news − pertaining to three major currency indices: dollar, pound and euro. In a dataset that spans two decades, we find a persistently positive link between the ‘news intensity’ and the volatility of currency returns, that becomes significantly stronger in times of recession: ‘bad news’ tends to translate into higher volatility.

Original languageEnglish
JournalApplied Economics Letters
StateAccepted/In press - 1 Jan 2024


  • currency indices
  • natural language processing
  • News
  • volatility

ASJC Scopus subject areas

  • Economics and Econometrics


Dive into the research topics of 'News intensity and asset returns: the case of currency volatility'. Together they form a unique fingerprint.

Cite this