Wolfram Computation Meets Knowledge

Wolfram Summer School


Diogo Duarte

Summer School

Class of 2015


Diogo Duarte holds a BS in applied mathematics from Universidade Federal do Rio de Janeiro and an MS in mathematical finance from Instituto Nacional de Matemática Pura e Aplicada. He is currently a finance PhD candidate at Boston University. Diogo is interested not only in topics related to financial economics, including asset pricing, macro-finance, credit risk, and banking, but also in numerical methods applied to stochastic calculus, PDEs, estimation, Monte Carlo simulation, and machine learning.

Project: Internet Sentiment and the Stock Market

In the past few years, search engines have enabled researchers to conduct analysis on topics that are currently getting people’s attention. Some recent studies showed that such data could be potentially very useful in understanding the level of economic activity. For example, an increase in the number of searches of a particular asset could be an indicator that an important event for a company is taking place in the market. Intuitively, the trade volume of a particular stock should be correlated with the interest that such a company is arousing. Consequently, one should expect that assets’ characteristics like return and volatility should also be associated with the frequency that these companies show up in search engines.

To better understand how investors’ interests in a particular company comove with stock returns and volatility, I analyzed the correlation between Google Trends searches and financial market characteristics. I performed a multivariate regression analysis to understand if the coefficient of Google search factor is significant in explaining return movements.