For Better or For Worse: Algorithmic Choice in Experimental Markets

Authors: Elena N. Asparouhova, Xiaoqin Cai, Debrah Meloso, Jan Nielsen, Christine A. Parlour, Wenhao Yang

Abstract: Participants in an experimental market choose to enter private value trades manually and/or algorithmically. Each algorithm or trading robot makes or takes liquidity based on the trader’s current marginal valuation modulo a spread chosen by the trader. We evaluate experimental outcomes against both competitive equilibrium and equilibrium of the strategic game if all participants choose robots. Data from six laboratory experimental sessions support many of the theoretical findings. Most traders deploy an algorithm whenever available (the average trader deploys a robot in 82% of the rounds, and only 4% of subjects never deploy a robot), and learn to use them with experience. Compared to rounds with only manual trading, algorithms improve allocative efficiency. Realized gains from trade increase from 55% to 84%. While the allocative efficiency increases across the board, those who benefit most are the traders who perform poorly in manual trading. Our results highlight how algorithm choice can affect relative outcomes and market observables.