28 January 2016
IMT alumnus Gabriele Ranco, a recent graduate of the Economics, Markets and Institutions PhD program and theoretical physicist by training, led a dynamic study that analyzed web surfing patterns to make accurate predictions of stock movements. IMT Prof. Guido Caldarelli was also part of the research group. One of the outputs of this study was an article recently published in the prestigious peer-reviewed journal "PLOS ONE". The article was also just highlighted in the journal "Science News" (see link below).
Ranco and collaborators, starting from a large dataset from Yahoo!Finance (an online portal for financial news and data), combined the information coming from the sentiment conveyed by the respective number of "clicks" within a given article to forecast price returns for 100 highly capitalized US stocks at a daily and intra-day time scale. For each of the companies the team built a signed time series of the sentiment expressed in the related news which was then weighted by the number of article views. Each click action was associated with a timestamp, thus allowing for the construction of a time series at the time resolution of the minute. The team believes that their analysis is the first to take place at such intra-day granularity. Moreover, the study "illustrates the potential of 'big data' to provide new means to detect broad social patterns of belief - a problematic task in the areas ranging from public health to political polling", as Caldarelli says in his interview for "Science News".
This study, which intersects the fields of theoretic physics, economics and social behavior, is a demonstration of how the multidisciplinary and integrated research carried out at IMT School represents a novel tool to dissect complex issues. IMT School Director congratulates Dr. Ranco, a young School graduate, and Prof. Guido Caldarelli, for their successful research.
For the full entry in "Science News", which includes a link to the "PLOS ONE" article, follow the link below.