Robo-Advisory and Decision Inertia - Experimental Studies of Human Behaviour in Economic Decision-Making

Abstract

Investing in the stock market is a complicated and risky undertaking for private households. In particular, private investors face numerous decisions: for instance, whether to invest in stocks or bonds, buy passively or actively managed investment products, or try something new like Bitcoin. They must decide where they can get independent financial advice, and whether this advice is trustworthy. As a consequence, information systems researchers design and build financial decision support systems. Robo-advisors are such decision support systems aiming to provide independent advice, and support private households in investment decisions and wealth management. This thesis evaluates robo-advisors, their design and use and thus their ability to support financial decision-making. Addressing this research need, my thesis is organized in three parts (part I-III ) consisting of four quantitative experimental studies, two qualitative friendly-user-studies, and one qualitative interview study. In Part I, Chapter 3 examines how robo-advisors can be designed for inexperienced investors. In particular, I derive design recommendations for the development of robo-advisor solutions and evaluate them in a three-cycle design sciences process. Requirements related to the clusters ease of interaction, work efficiency, information processing and cognitive load are identified as key elements for robo-advisory design. In Part II, Chapter 4 focuses on an important bias in economic decision-making - decision inertia, the tendency to repeat a decision regardless of the consequences. As a result, a decision-maker can make repeated suboptimal investments. To understand this bias more deeply, I investigate decision inertia in a general experimental setting and identify motivational and cognitive drivers of this phenomenon. Thus, I relied on behavioural, on self-reported, and on bio-physiological measures in three laboratory studies. In Part III, Chapter 5 specifies the findings from Part II to find and evaluate strategies to reduce decision inertia in financial decision support systems. For that purpose, I investigate two nudges (design features) to reduce inertia in investment decisions. My results suggest that defaults and warning messages can help participants to overcome decision inertia. Furthermore, the results illustrate that designers have to be careful not to push decision-makers into the decision inertia bias by accident. In summary, this thesis gives design recommendations for practitioners and scholars building robo-advisors. The insights can help to develop robo-advisors, and to increase advisor quality by considering decision inertia in the system design phase and consequently, it illustrates how to counteract this malicious decision bias for private investors.