Cardiff University | Prifysgol Caerdydd ORCA
Online Research @ Cardiff 
WelshClear Cookie - decide language by browser settings

An extreme firm-specific news sentiment asymmetry based trading strategy

Song, Qian, Liu, Anqi, Yang, Steve Y., Deane, Anil and Datta, Kaushik 2016. An extreme firm-specific news sentiment asymmetry based trading strategy. Presented at: 2015 IEEE Symposium on Computational Intelligence, Cape Town, South Africa, 7-10 December 2015. 2015 IEEE Symposium Series on Computational Intelligence. IEEE, p. 898. 10.1109/SSCI.2015.132

[img]
Preview
PDF - Accepted Post-Print Version
Download (813kB) | Preview

Abstract

News sentiment has been empirically observed to have impact on financial market returns. In this study, we investigate firm-specific news from the Thomson Reuters News Analytics data from 2003 to 2014 and propose an optimal trading strategy based on a sentiment shock score and a sentiment trend score which measure extreme positive and negative sentiment levels for individual stocks. The intuition behind this approach is that the impact of events that generate extreme investor sentiment changes tends to have long and lasting effects to market movement and hence provides better prediction to market returns. We document that there exists an optimal signal region for both indicators. And we also show extreme positive sentiment provides better a signal than extreme negative sentiment, which presents an asymmetric market behavior in terms of news sentiment impact. The back test results show that extreme positive sentiment generates robust and superior trading signals in all market conditions, and its risk-adjusted returns significantly outperform the S&P 500 index over the same time period.

Item Type: Conference or Workshop Item (Paper)
Date Type: Publication
Status: Published
Schools: Mathematics
Publisher: IEEE
ISBN: 9781479975600
Date of First Compliant Deposit: 26 February 2018
Date of Acceptance: 14 October 2015
Last Modified: 16 Aug 2019 12:45
URI: http://orca.cf.ac.uk/id/eprint/109447

Citation Data

Cited 2 times in Google Scholar. View in Google Scholar

Cited 3 times in Scopus. View in Scopus. Powered By Scopus® Data

Actions (repository staff only)

Edit Item Edit Item

Downloads

Downloads per month over past year

View more statistics