Not logged in.
Quick Search - Contribution
Contribution Details
| Type | Master's Thesis |
| Scope | Discipline-based scholarship |
| Title | High Frequency Effects of News Sentiment on FX Shocks |
| Organization Unit | |
| Authors |
|
| Supervisors |
|
| Language |
|
| Institution | University of Zurich |
| Faculty | Faculty of Business, Economics and Informatics |
| Number of Pages | 40 |
| Date | 2020 |
| Zusammenfassung | Do news releases get fully reflected in the price as soon as they are released, or does it take some time for market participants to process them? On Figure 1.1 we can see the EURUSD movement on the day when S&P downgraded Greek debt to non-investment grade, which meant Greece could default on its debts and potentially exit the Eurozone. This sparked a debate about the viability of Euro as a common currency, and consequently caused the Euro to depreciate. This depreciation was still going on even weeks after the downgrade announcement, which could indicate that the investors needed some time to comprehend the news and connect them with other experiences to decide how bad the situation in Greece truly is. Their digestion of the news in turn affected their decision to buy or sell the Euro, which ultimately materialised in a lower EURUSD exchange rate. There are many other similar events when news strongly a↵ected the exchange rate development. It would be very interesting to look at this relationship of news and price development systematically. There have been many studies (Galati and Ho (2003), Ehrmann and Fratzscher (2005), Jansen and De Haan (2005), Andersen et al. (2007), Evans and Lyons (2008)) showing the significance of the effect of macroeconomic announcements on the EURUSD exchange rate and on other assets. There have also been studies looking at sentiment of actual news articles and showing that it takes time for news articles to influence stocks, bonds and currencies (Uhl (2014), Uhl (2017), Heston and Sinha (2013), Smales (2014)). All of these studies have been done at low frequencies (daily, monthly), but to the best of our knowledge there has been no research done examining the high frequency e↵ect of news sentiment on exchange rates . How do the incoming news affect the currency rates? Does the price fully adapt instantaneously or does it take some time for the news to precipitate and for the market participants to digest it? How long is this effect? Can it be exploited? This is what we attempt to discover in this thesis. This is the first study looking into the intraday relationship between news sentiment and foreign exchange systematically . For our analysis we use an external database containing 16 years of news, altogether 100 million news articles in microsecond precision, and their sentiment scores. The sentiments were calculated by the database provider They assign low scores to negative news and high scores to positive news. We filter and process the news to generate a time series that reflects the current news sentiment towards Europe and currencies. We then show that this news sentiment time series predicts the EURUSD returns 1 hour and more into the future. This proves our hypothesis that it takes time for market participants to process the news and that the entire information surprise is not immediately impounded in the price. Our results are in line with the findings in the current literature on low frequencies (Uhl (2017), Uhl et al. (2015), Heston and Sinha (2013)). To show the relevance of the news e↵ect in practice, we furthermore develop and backtest a trading strategy that takes advantage of the results and signals we found. We focus on the EURUSD pair, since our high frequency dataset is well tagged with topic codes for those two currencies. |
| Export |
BibTeX
|