BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Textual Analysis in Asset Pricing.
DTSTART:20200911T160000
DTEND:20200911T173000
DTSTAMP:20260408T071415Z
UID:56a9e63305f0accd63ffd0ae9c10b2e6824a159180b1978432e872d8
CATEGORIES:Conferences - Seminars
DESCRIPTION:Bryan KELLY\, Yale\nBryan Kelly will present a combination of 
 the following two papers:\n\n	The Structure of Economic News\n\nWe propose
  an approach to measuring the state of the economy via textual analysis of
  business news. From the full text content of 800\,000 Wall Street Journal
  articles for 1984–2017\, we estimate a topic model that summarizes busi
 ness news as easily interpretable topical themes and quantifies the propor
 tion of news attention allocated to each theme at each point in time. We t
 hen use our news attention estimates as inputs into statistical models of 
 numerical economic time series. We demonstrate that these text-based input
 s accurately track a wide range of economic activity measures and that the
 y have incremental forecasting power for macroeconomic outcomes\, above an
 d beyond standard numerical predictors. Finally\, we use our model to retr
 ieve the news-based narratives that underly “shocks” in numerical econ
 omic data.\nhttps://papers.ssrn.com/sol3/papers.cfm?abstract_id=3446225\n\
 n\n	Predicting Returns with Text Data\n\nWe introduce a new text-mining me
 thodology that extracts sentiment information from news articles to predic
 t asset returns. Unlike more common sentiment scores used for stock return
  prediction (e.g.\, those sold by commercial vendors or built with diction
 ary-based methods)\, our supervised learning framework constructs a sentim
 ent score that is specifically adapted to the problem of return prediction
 . Our method proceeds in three steps: 1) isolating a list of sentiment ter
 ms via predictive screening\, 2) assigning sentiment weights to these word
 s via topic modeling\, and 3) aggregating terms into an article-level sent
 iment score via penalized likelihood. We derive theoretical guarantees on 
 the accuracy of estimates from our model with minimal assumptions. In our 
 empirical analysis\, we text-mine one of the most actively monitored strea
 ms of news articles in the financial system — the Dow Jones Newswires 
 — and show that our supervised sentiment model excels at extracting retu
 rn-predictive signals in this context.\nhttps://papers.ssrn.com/sol3/paper
 s.cfm?abstract_id=3489226\n\n 
LOCATION:Zoom
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR