No Bubbles in the City. An Experiment with Professional Traders

How to cite

Kendall, Ryan; Guarino, Antonio; Cipriani, Marco. 2019. "No Bubbles in the City. An Experiment with Professional Traders." AEA RCT Registry. AEARCTR-0003853. https://doi.org/10.1257/rct.3853-2.1

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Ryan Kendall* , Antonio Guarino , Marco Cipriani

* Primary investigator

Registry ID
AEARCTR-0003853
DOI
https://doi.org/10.1257/rct.3853-2.1
First registered
Last updated
Status
In development
Trial dates
2019-02-04 – 2019-11-01
Countries
Private
Registry record
http://www.socialscienceregistry.org/trials/3853

Abstract

We examine how professional financial traders behave in two classic laboratory market environments. First, traders trade an asset with a declining fundamental value (Smith, Suchanek, & Williams, 1988) while receiving private information about the asset's value. Second, traders make choices in a ``guessing game'' with peers (Nagel, 1995) as well as within our novel ``individual guessing game'' framework. We also collect data on cognitive factors (IQ and CRT), behavioral factors (risk, confidence, and digit-ratio), and personality traits (e.g. Big 5). This is compared against data from undergraduate students. In accordance with previous studies, undergraduates overprice the asset (a ``bubble'') and, in the guessing game, make choices consistent with different levels of strategic reasoning. In contrast, traders trade at prices that closely reflect the asset's declining fundamental value and make choices that suggest higher levels of strategic reasoning than the levels of the students. Professional traders are also better than undergraduates in aggregating private information through trading activity. In addition, choices from our individual guessing game can decompose a subject's own level of strategic reasoning from their belief about the population's level of reasoning. Our data show that traders anticipate that their opponents will have higher levels of strategic reasoning than do that of students.

Experimental design

We are studying behavior in market experiments across different groups: Professional financial traders and undergraduate subjects. We are also studying for differences in individual-level behavior across groups (e.g. IQ).

Design details

Randomization method
Randomization is always carried out by computer programs.
Randomization unit
We will be looking at, both, individual level choices (e.g., scores on tests and personal earnings) and group level choices (e.g., average price per period and comparison across population groups).
Clusters
We have 56 subjects in each of the two groups (for 112 total subjects).
Observations
Each subject makes make choices in the experiment.
Arms
Our treatment variable is the two different population groups.
Primary outcomes
We have two primary outcomes we are interested in. First, do the traders behave differently (better) in the market experiments than do the students? Second, based on a battery of individual-level tests conducted after the market experiments, which cognitive or non-cognitive trait best predicts performance in the experimental market.

Cited by

No citation data available for this registration yet.

Keywords