Abstract

Rankings of top performers are extensively used in competitive environments, such as financial markets, although the way they are designed and institutionalized is often criticized as biased or arbitrary at best. In this paper, we aim to understand and measure the causal effect of a stock’s placement in a ranking on investors’ attention. We conduct an at-scale randomized experiment on Finland’s most prominent financial website. It ranks firms according to their daily returns into best performers, worst performers, and representative performers. We introduce exogenous variation in the placement of stocks in the ranking, allowing us to identify the causal effect of rankings on attention. We find that a placement of a stock in the ranking results in a statistically and economically significant spike of investors’ attention. The effect is not just driven by extremely high or low returns. It is similar in magnitude for representative stocks, that have moderate returns: Investors’ attention increases by 309pp and 273 pp for stocks ranked as best and worst performers, respectively, and by 185pp for stocks ranked as representative. The causal effects of rankings on investors’ attention are unlikely to be driven by a reduction of the information search costs as in our setting all relevant information is just one click away. Such placement effects support the conjecture that even arbitrary choices for the design and institutionalization of rankings are highly influential and prone to forces of manipulation.