Do Risk Preferences Shape the Effect of Online Trading on Trading Frequency, Volume, and Portfolio Performance?

ABSTRACT

How do investors’ risk preferences influence the relationships between investors’ online channel use intensity and both their trading behaviors and performance? This study answers this important question even as investors are increasingly relying on the Internet for their trading activities. We leverage rare and unique micro-level historical dataset from more than 7,000 investor accounts over a 44-month period between 2010 and 2013 at a large brokerage firm in China. The dataset and analyses enable us to provide new insights into how investors’ online channel use intensity and risk preferences jointly influence their trading behaviors and performance, even though some other aspects of financial markets have changed considerably over the years. The findings reveal that although online channel use intensity is associated with increased trading volume, trading frequency, and investment returns, these effects differ across investors with different risk preferences. We find that while online channel use intensity has strong positive effects on transaction frequency for both risk-seeking and risk-averse investors, it has a much lower effect on trading volume for risk-averse investors than for risk-seeking investors. We further find that risk-averse investors with higher online channel use intensity outperform investors with other risk preferences in terms of investment performance. This paper contributes to the emerging literature at the intersection of information systems and behavioral finance by revealing the moderating role of risk preferences in the relationships between investors’ online trading channel use intensity and both their trading behaviors and outcomes. We discuss the implications for research and practice.

KEYWORDS:

• Stock market
• online trading
• portfolio performance
• return on investment
• risk preferences
• trading behavior
• FinTech

Acknowledgments

We thank Vladimir Zwass (Editor-in-Chief), anonymous associate editor and two reviewers for their constructive suggestions in the review process that significantly improved the paper. We are grateful to the participants of the Workshop on Information Systems and Economics, and the seminar participants at IIT Roorkee, Indian School of Business, and University of Florida for helpful comments. We thank Xinshu Zhao and John G. Lynch for helpful comments regarding mediation analyses. Finally, we thank Andy Seagram, Carline Toscano, and Ali Ferguson for editing, and the anonymous associate editor for many useful comments to improve the presentation and layout.

Disclosure statement

No potential conflict of interest was reported by the author(s).

SUPPLEMENTARY MATERIAL

Supplemental data for this article can be accessed online at https://doi.org/10.1080/07421222.2023.2196777

Correction Statement

This article has been corrected with minor changes. These changes do not impact the academic content of the article.

Notes

¹ We thank the anonymous AE for pointing to the value of historically relevant datasets for addressing fundamental questions.

² In our research context, these technologies were increasingly accessible by investors during our observational period. The most popular instant message software, Tencent QQ, was initially released in 1999, but it started to become widely used in the 2000s. Sina Weibo, one of the biggest social media platforms in China, was launched in 2009.

³ We summarize a list of selected studies on online trading in Online Supplemental Appendix A.

⁴ The brokerage firm did not provide a mobile app channel in our sample period.

⁵ Our index for China stock and federal bank was obtained from the RESSET database [82], which is a leading financial data provider of model testing and investment research and reliable data sources sponsored by leading China research institutions, such as Tsinghua University and Peking University. We calculated the end-of-month return of individual i based on the weighted returns of the portfolio held by her.

⁶ The sample size for trading performance is reduced due to missing data when calculating risk-adjusted returns. We report the analysis of trading behaviors with the full sample to avoid losing further observations unnecessarily, and the results for trading behaviors with the reduced sample are broadly similar to the results from the full sample.

⁷ The panel does not include months when an investor did not make a transaction because both trading frequency and trading volume are 0, and OnlineUseIntensity% is not defined when the number of transactions is 0. We obtained a smaller sample size for trading performance due to missing data when calculating risk-adjusted returns.

⁸ Because the natural log of 0 is not defined, we added 1 to trading frequency before taking the natural log.

⁹ We used the exchange rate of the Chinese yuan to the US dollar on January 8, 2013, which is 6.21 CNY to 1 USD.

¹⁰ We thank an anonymous AE and reviewers for some of these analyses.

¹¹ We thank the AE for emphasizing the need for studies involving historical archival datasets for assessing the generalizability of theories beyond the period of the data itself. Certainly, the issues of risk preferences of investors and the extent to which they should engage in online trading continue to be of importance, even though online stock trading has changed over the years. Indeed, IS researchers often use such archival data for addressing enduring and fundamental questions [45,67,88]. Economists also do not hesitate to use data from the 1930s in Germany [42] just because of the age of the data. Scientific understanding depends on a focus on the importance and relevance of the research question rather than the currency of data itself as the sole criterion.

¹² We thank the anonymous AE for pointing to this discussion and the value of conducting studies across countries.

¹³ We find evidence for mediation via trading frequency but did not find evidence for the mediation via trading volume [96]. We thank Xinshu Zhao and John Lynch for helpful and clarifying comments about the role of R-squared in tests for mediation that are largely based on statistical significance of the product terms involved in mediation.

Additional information

Notes on contributors

Yang Pan

Yang Pan ([email protected]) is an Assistant Professor of management science at Freeman School of Business, Tulane University. She holds a PhD from the Robert H. Smith School of Business at the University of Maryland. Her research examines the impact of entrepreneurial startups on the behavior of IT/Tech firms and the shaping of people’s behavior by technology, such as Fintech and digital platforms. Dr. Pan’s research has been accepted in premier journals such as MIS Quarterly, Information Systems Research, and Production and Operations Management. Her research has been awarded the best student paper (finalist) recognition at the Academy of Management, Technology & Innovation Management Division.

Sunil Mithas

Sunil Mithas ([email protected]; corresponding author) is a World Class Scholar and Professor at the Muma College of Business, University of South Florida, where he serves as a director of Rankings and Reputation. Identified as an MSI Young Scholar by the Marketing Science Institute, Dr. Mithas is a Distinguished Fellow of the Information Systems Society of INFORMS. Previously, he was the Ralph J. Tyser Professor of Information Systems at the Robert H. Smith School of Business at the University of Maryland, where he co-directed two centers. He is a visiting professorial fellow at the UNSW Business School, Sydney, and has held visiting positions at the University of California, Davis, UNSW, University of Mannheim, and HKUST, Hong Kong. Dr. Mithas is the author of two books, and his research published in top journals and conference proceedings has won multiple best-paper awards, and was featured in practice-oriented publications such as the MIT Sloan Management Review, Management Business Review, and Bloomberg. He has consulted and conducted research with a range of organizations, including A. T. Kearney, EY, Johnson & Johnson, the Social Security Administration, U.S. Census Bureau, and the Tata Group.

J.J. Po-An Hsieh

J.J. Po-An Hsieh ([email protected]) is an Associate Professor of Information Systems at Georgia State University. His research interest includes IT use and impacts, IT-enabled innovation, data analytics, artificial intelligence, and the future of work. Dr. Hsieh has served on the editorial board of MIS Quarterly, Information Systems Research, and Journal of the Association of Information Systems, among others. He received the Annual Best Associate Editor Award from Information Systems Research and the Annual Best Journal Paper Award and the Best Conference Paper Award by the CTO division of the Academy of Management. He further received the Best Paper Award at the International Conferences on Information Systems. He was voted as the Top Professor by GSU students.

Che-Wei Liu

Che-Wei Liu ([email protected]) is an Assistant Professor of Information Systems at the Kelley School of Business, Indiana University. He completed his PhD at the University of Maryland. His research interests include healthcare IT, IT Worker, and IT adoption (FinTech & Artificial Intelligence). Dr. Liu’s research has been accepted in premier journals such as Management Science, Information Systems Research, Journal of Management Information Systems, MIS Quarterly, and Journal of Economic Behavior & Organization. He received the All-S.T.A.R. Fellowship and CIBER PhD Research award. His research has also been awarded the Best Paper runner-up awards at International Conference on Information Systems and International Conference for Smart Health.