Does Algorithmic Trading Affect Analyst Research Production?
35 Pages Posted: 4 Jun 2020
Date Written: May 10, 2020
We document that Algorithmic Traders (ATs) reduce analysts’ stock coverage and the number of analyst research reports. This evidence reflects that ATs pre-empt trades on new information, which reduces non-AT investment-driven demand for analyst research. Consistently, the effects we document dominate (i) when analysts produce new information rather than simply disseminate public information; (ii) when there are more institutional investors demanding research for trading purposes rather than for monitoring purposes; (iii) for stock recommendations rather than earnings forecasts (as the latter also serve a monitoring purpose); (iv) when stock liquidity is higher, as ATs require high liquidity for profitable trades; (v) for smaller stocks where the analyst’s reputational cost of dropping coverage is lower. We address endogeneity using firm-fixed effects, regressions in changes, and using a natural experiment based on the tick-size pilot programme. Overall, our results suggest that an unintended consequence of algorithmic trading is lower analyst research production.
Keywords: algorithmic trading; analyst coverage; earnings forecasts; stock recommendations
JEL Classification: D53; G12; G14; M41
Suggested Citation: Suggested Citation