Investor biases in financial markets

University of Technology Sydney. Faculty of Business.Behavioral biases are a key determinant of investors’ wealth outcomes in financial markets. However, there is still much to learn about the nature of these biases. This thesis studies different types of financial markets to understand the individual- and aggregate-level effects of specific biases and how these biases are influenced by the environment. The first study examines how biases drive people to participate in pump-and-dump manipulation games in cryptocurrency markets. Cryptocurrency manipulators publicly announce the intended manipulations and invite others to join them. In a simple framework, we show that these pumps are inherently negative-sum games for non-manipulators, as manipulators have an advantage. So, why do people participate in these schemes? Rational agents do not participate unless they have a skill or speed advantage. However, overconfident agents and gamblers participate, even without any advantage. We find strong empirical support for both mechanisms. Pumps generate extreme price and volume distortions, and cause large wealth transfers between participants. The second study asks whether investor loss-making tendencies are influenced by the trading environment, particularly the trend towards frictionless access to markets. We hypothesize that adding back certain “trading frictions” in markets can make investors think harder and mitigate losses arising from impulsive trading. Using laboratory experiments, we examine how investor performance is impacted by various frictions: transaction costs, time delays in placing orders, and tasks requiring cognitive effort. High costs and time delays have no effect or harm performance, whereas cognitive tasks benefit participants who are most prone to underperforming. We conclude that frictions can yield performance benefits if they help inattentive investors consider nonsalient fundamental information they might otherwise neglect. The third study investigates the aggregate effects of attention to fundamental information in stock markets. To isolate attention to information, we propose a new measure that involves classifying investors as inattentive when they do not cancel or update their pending orders after material overnight news events. We find strong underreactions and price drifts after material corporate announcements for stocks that receive less attention. These drifts cannot be explained by other mispricing effects, such as mispricing after earnings announcements and for stocks favored by retail investors. In summary, this thesis contributes by investigating how biases drive participation in loss-making schemes in financial markets, how certain elements in the market architecture can reduce investor losses by directing their attention to fundamental information, and the aggregate effects of inattention to fundamental information in stock markets