Information in Asset Market Experiment with Bot
Speaker: Enrica Carbone (University of Campania)
This seminar is hosted by Professor John Hey - firstname.lastname@example.org
Abstract: This paper proposes an experimental asset market in which human traders interact with a robot trader, with the aim of evaluating the effect on transactions price behavior, of random arrival of
good news or bad news on fundamentals.
The experimental design considers a continuous double-auction market based on Smith at al. (SSW, 1988), in which human agents exchange securities manually, but they interact with a robot trader, which uses an algorithm to exchange securities between markets.
The experiment aims to reproduce the main strategy adopted by high-frequency algorithmic traders in real markets to obtain profits, that is, an arbitrage strategy between markets. For this reason we resumed the structure of Angerer, Neugebauer and Shachat (2019), implementing an arbitrage strategy but evaluating the impact of market news, in presence of the robot traders, on the assets price discovery.
Based on the experimental literature on learning mechanisms in financial markets, for example Khunen (2015), we expect asymmetries in transactions price behavior, in presence of good news compared to the presence of bad news on fundamentals. In particular, we expect, as in Weber and Welfens (working paper 2007), a price drift and an underreaction of the transaction price. We expect this effect to be mitigated by the presence of the arbitrage robot traders.
The experimental design provides the trading of two related assets in two separate markets. The presence of two related assets, is justified by the need to create arbitrage opportunities for robot traders. There is a type X asset that pays stochastic dividends and an asset Y that pays stochastic dividends equal to X + 50 units.
Co-authors: Tibor Neugebauer (Luxembourg) and Angelo Ventrone (Salerno)