How Can Network Structure and Information Add Economic Values to FinTech? Two Studies on Co-attention Based Networks from the Design Science Perspective
Financial technologies (FinTech) are recasting the relationships between users and firms. Today, investors are increasingly adopting digital platforms to acquire and process investment-related information, participate in online discussions, and interact with other users. Meanwhile, users also leave their "digital footprints" which convey clues in reflecting their correlated demands of different firms and assets. Particularly, we are interested in users’ online co-searches of related firms and assets. Upon this, we construct networks to reflect users' co-attention relationships among assets and firms. Our objective is to show whether the structural characteristics and the information content in networks are of economic values and impact, and how we can use them for network-based inference. In line with the design science paradigm, we will introduce two designed artifacts in the context of co-attention networks. First, we develop a composite metric to reflect the overall information flow associated with an individual node in a network. We evaluate the effectiveness of this metric by applying it to predict stock returns. The results indicate that our metric has better predictive performance than alternative measures. In the second study, we design a novel market segmentation approach from the demand side by applying network embedding and clustering techniques. Our approach can yield economically meaningful segments which are different from traditional industries.
Dr. Wuyue Shangguan is a postdoctoral researcher in the Department of Management Science. Before joining Xiamen University, she gained her doctoral degree in management science and engineering from Zhejiang University. She was also a visiting Ph.D. in McCombs School of Business, The University of Texas at Austin in 2017 and 2018. Her research is focused on social media analytics, social and information networks, FinTech and digital investment platforms. She uses a variety of methods in her research such as econometrical analysis, causal inference, data mining and machine learning. Her research has published in SSCI journals and leading international conference proceedings.