Topic Sentiment Asset Pricing with DNN Supervised Learning

54 Pages Posted: 20 Aug 2018 Last revised: 23 Sep 2018

See all articles by Hitoshi Iwasaki

Hitoshi Iwasaki

National University of Singapore (NUS) - Department of Statistics and Applied Probability

Ying Chen

National University of Singapore

Date Written: August 8, 2018

Abstract

We develop an innovative deep neural network (DNN) supervised learning approach to extracting insightful topic sentiments from analyst reports at the sentence level and incorporating this qualitative knowledge in asset pricing and portfolio construction. The topic sentiment analysis is performed on 113,043 Japanese analyst reports and the topic sentiment asset pricing model delivers superior predictive power on stock returns with adjusted R2 increasing from 1.6% (benchmark model without sentiment) to 14.0% (in-sample) and 13.4% (out-of-sample). We find that topics reflecting the subjective opinions of analysts have greater impact than topics of objective facts and justification of the quantitative measures.

Keywords: Sentiment Analysis; Asset Pricing; Portfolio construction

JEL Classification: G11, G12, C89

Suggested Citation

Iwasaki, Hitoshi and Chen, Ying, Topic Sentiment Asset Pricing with DNN Supervised Learning (August 8, 2018). Available at SSRN: https://ssrn.com/abstract=3228485 or http://dx.doi.org/10.2139/ssrn.3228485

Hitoshi Iwasaki (Contact Author)

National University of Singapore (NUS) - Department of Statistics and Applied Probability ( email )

Block S16, Level 7
6 Science Drive 2
117546
Singapore

Ying Chen

National University of Singapore ( email )

Block S16, Level 7,
6 Science Drive 2
Singapore, 117546

Register to save articles to
your library

Register

Paper statistics

Downloads
303
Abstract Views
1,348
rank
99,798
PlumX Metrics