Uninformed but Predictable: Corporate Trading and Price Discovery in Over-the-counter FX Markets
71 Pages Posted: 29 Apr 2022 Last revised: 13 Jun 2022
Date Written: June 2022
We characterize the demand function of multi-national corporations in foreign exchange (FX) markets and assess its impact on asset prices. Our findings suggest that corporate order flow does not consistently predict future returns, it is highly auto-correlated and, in contrast to pure noise, it strongly responds to realized asset returns. Further, we find compelling evidence of a positive impact of corporations’ net order imbalance on the post-execution volatility of currency returns. Using settlement breaks to instrument for quasi-exogenous shocks to liquidity-motivated trading, we show that a one standard-deviation drop in traded volume decreases return volatility by 38%. This effect dissipates within a trading day. We document that order imbalance also impacts dealers’ price spreads to corporations, but observable currency and trade characteristics explain less than 1% of the variation. Based on these empirical facts, we propose a dynamic model of price determination between a dealer and an uninformed agent in a setting with asymmetric information, multi-period relationships and search frictions.
Keywords: Foreign Exchange, Market Microstructure, Information, Volatility
JEL Classification: F31, G12, G15
Suggested Citation: Suggested Citation