Abstract: 

The high volume and time sensitivity of news and social media stories requires automated processing to quickly extract actionable information, while the unstructured nature of textual information presents challenges that are comfortably addressed through machine learning techniques. This talk will cover the following topics:
♦Extracting actionable information in the high volume, time-sensitive environment of news stories and social media content using machine learning
♦ Quantitative techniques for assigning aggregated sentiment scores and other derived metrics (e.g., sentiment dispersion)
♦ Demonstrating that using sentiment scores in your trading strategy ultimately helps achieve higher risk-adjusted returns
♦ Illustrating variation in sensitivity of sentiment with respect to industry sector, market cap, trading volume, etc.
♦ Uncovering topic codes that are more relevant for return prediction, as well as those which lead to noisy sentiment extraction and a weaker predictive signal

Bio: 

Dr. Arun Verma joined the Bloomberg Quantitative Research group in 2003. Prior to that, he earned his Ph.D from Cornell University in the areas of computer science & applied mathematics. At Bloomberg, Mr. Verma’s work initially focused on Stochastic Volatility Models for Derivatives & Exotics pricing. More recently, he has enjoyed working at the intersection of diverse areas such as data science (with structured & unstructured data), innovative quantitative models across all asset classes & using machine learning methods to help reveal embedded signals in financial data.