Retail trading
American Economic Review, Accepted. [+ bibtex]
We document a robust dynamic inconsistency in risky choice. Using a unique brokerage dataset and a series of experiments, we compare people's initial risk-taking plans to their subsequent decisions. Across settings, people accept risk as part of a ``loss-exit'' strategy---planning to continue taking risk after gains and stopping after losses. Actual behavior deviates from initial strategies by cutting gains early and chasing losses. More people accept risk when offered a commitment to their initial strategy. Our results help reconcile seemingly contradictory findings on risk-taking in static versus dynamic contexts. We explore implications for theory and welfare.
“Biased by Choice: How Financial Constraints Can Reduce Financial Mistakes,” (w/ A. Imas, 2022)
Review of Financial Studies, 35(4): 1643-1681. [+ bibtex]
RFS Rising Scholar Award
We show that constraints can improve financial decision-making by disciplining behavioral biases. In financial markets, restrictions on leverage limit traders’ ability to borrow to open new positions. We demonstrate that regulation that restricts the provision of leverage to retail traders improves trading performance. By increasing the opportunity cost of postponing the realization of losses, leverage constraints improve traders’ market timing and reduce their disposition effect. We replicate these findings in two distinct experimental settings, further isolating the mechanism and demonstrating generality of the results. The interaction between constraints and behavioral biases has implications for policy and choice architecture.
“Should Retail Investors’ Leverage Be Limited?” (w/ A. Simsek, 2019)
Journal of Financial Economics, 132(3): 1-21. Lead article. [+ bibtex]
Best paper finalist in Asset Pricing at SFS Cavalcade
Does the provision of leverage to retail traders improve market quality or facilitate socially inefficient speculation that enriches financial intermediaries? We evaluate the effects of 2010 regulations that cap leverage in the U.S. retail foreign exchange market. Using three unique data sets and a difference-in-differences approach, we document that the leverage constraint reduces trading volume by 23%, alleviates high-leverage traders’ losses by 40%, and reduces brokerages’ operating capital by 25%. Yet, the policy does not affect the relative bid-ask prices charged by the brokerages. These results suggest the policy improves belief-neutral social welfare without reducing market liquidity.
“Peer Pressure: Social Interaction and the Disposition Effect,” (2016)
Review of Financial Studies, 29(11): 3177-3209. [+ bibtex]
Social interaction contributes to some traders’ disposition effect. New data from an investment-specific social network linked to individual-level trading records builds evidence of this connection. To credibly estimate causal peer effects, I exploit the staggered entry of retail brokerages into partnerships with the social trading web platform and compare trader activity before and after exposure to these new social conditions. Access to the social network nearly doubles the magnitude of a trader’s disposition effect. Traders connected in the network develop correlated levels of the disposition effect, a finding that can be replicated using workhorse data from a large discount brokerage.
Inquire Europe Working Paper 2022. [+ bibtex]
“Facebook Finance: How Social Interaction Propagates Active Investing,” (w/ D. Simon, 2015)
Federal Reserve Bank of Cleveland Working Paper Series, No. 1522. [+ bibtex]
This paper shows how active investing strategies propagate through social connections in a network of retail traders, using a new database of social activity linked to individual-level trading records. A trader's good short-term performance causes them to contact others. A trader's activity increases when peers perform well and increase communication. We use the staggered entry of brokerages into partnerships with the social networking platform, which is a necessary precursor for traders to access the network, to argue these effects are causal. This pattern of communication supports active trading, even though the network reveals the low success rate of retail traders.