01 / Language
Make text measurable.
Language models turn disclosure and sentiment into data.
Wenlan (Tony) Xie
Finding signal in the noise.
Economics research on language and financial markets.
01 / Words
The most important signals arrive as words.
One review is an opinion. Thousands of them start to look like a forecast.
02 / Measurement
A language model reads every one.
Each review becomes a sentiment score, measured the same way at any scale.
03 / Evidence
From words to evidence.
Then the real test: does that sentiment show up in returns?
Read the paperReading Between the ReelsIJFE 2026- Noise
- Words
- Scores
- Evidence
01 / About
Wenlan Xie.
Economics at the University of Chicago.
I study how language shapes disclosure, investor attention, and financial markets.
- NowUniversity of Chicago · Economics
- M.ComUniversity of Sydney · Data Analytics
- B.A.University of Canterbury
02 / Research
Language into evidence.
Studying how information shapes decisions.
02 / Method
Separate signal from noise.
Econometrics and simulation test what the patterns mean.
03 / Markets
Understand decisions.
Corporate disclosure, investor attention, and financial outcomes.
03 / Publications
Selected work.
-
2026
Integrating Monetary and Environmental Goals: Exploring the Impact of Green Monetary Policy on Corporate Environmental Disclosure
Accounting Forum
-
2026
Reading Between the Reels: AI-Powered Analysis on Movie Review Sentiment as a Predictor of Market Returns
International Journal of Finance and Economics