Open Access. Powered by Scholars. Published by Universities.®

Digital Commons Network

Open Access. Powered by Scholars. Published by Universities.®

Finance

Research Collection Lee Kong Chian School Of Business

Canonical correlation analyses

Publication Year

Articles 1 - 2 of 2

Full-Text Articles in Entire DC Network

How Commonality Persists? (Through Investors' Sentiment And Attention), Chyng Wen Tee, Raja Velu, Zhaoque Zhou Dec 2023

How Commonality Persists? (Through Investors' Sentiment And Attention), Chyng Wen Tee, Raja Velu, Zhaoque Zhou

Research Collection Lee Kong Chian School Of Business

Studies on commonality generally attribute the variation in asset returns to the variation in order flows. In this research study, we show that order flows do not predict asset returns, rather their relationship have been static over time. Thus we model both returns and the order flows as endogenous variables, and use investors' sentiment and attention as exogenous factors via a reduced-rank regression. We provide empirical evidence to demonstrate that cross-sectional commonality in attention (sentiment) is linearly (nonlinearly) associated with both returns and order flows at the intraday level, while the sentiment and attention measures themselvesexhibit a nonlinear mutual relationship, …


Why Commonality Persists?, Chyng Wen Tee, Raja Velu, Zhaoque Zhou Apr 2022

Why Commonality Persists?, Chyng Wen Tee, Raja Velu, Zhaoque Zhou

Research Collection Lee Kong Chian School Of Business

We show that order flows do not exhibit predictive power on asset returns, and their relationships have been static over time. We use a reduced-rank regression formulation to model both returns and the order flows as endogenous variables, and use investors' sentiment and attention as exogenous factors. We provide empiricalevidence to demonstrate that cross-sectional commonality in attention (sentiment) is linearly (nonlinearly) associated with both returns and order flows at the intraday level, while the sentiment and attentionmeasures themselves exhibit a nonlinear mutual relationship, thus revealing the multi-dimensional aspect of the commonality relationship.