Nowadays financial news is no longer only available via traditional printed media, but also through real-time online sources, such as news websites and social media. The increased availability of financial news and investors’ ease of access to it, can impact market prices. Indeed, an increasing body of academic literature documents that news items from different sources affect investor sentiment and thus asset prices. Deriving sentiment indicators from news items requires the use of text mining algorithms to analyze textual news items and summarize them in numerical sentiment scores. Various data providers now offer such data sets of pre-processed news items.
The goal of this research project is to investigate news-based sentiment indicators and their ability to predict the returns of individual stocks and corporate bonds. Robeco Investment Research has access to rich historical databases that enable back-testing and evaluating of investment strategies. You will conduct a literature study, implement various news-based sentiment indicators and determine their predictive power for the returns of stocks and bonds.
The project covers the entire quant model development cycle: analyzing the data, programming the back-tests, analyzing the results, discussing results with researchers and portfolio managers, writing a research report and giving a presentation. As with all Super Quant internships, the assignment will be supervised by an experienced empirical researcher of Robeco’s Investment Research department. Creative, analytic and programming skills are essential to successfully complete the project.
Allen, McAleer, Singh, 2015, “Daily market news sentiment and stock prices”, working paper
Da, Engelberg, Gao, 2011, “In search of attention”, Journal of Finance
Hafez, Xie, 2016, “News beta: Factoring sentiment risk into quant models”, Journal of Investing
Tetlock, Saar-Tsechansky, Macskassy, 2008, “More than words: Quantifying language to measure firms’ fundamentals”, Journal of Finance