Market Conditions, Order Flow and Exchange Rates Determination

In comparison to macroeconomic models of nominal exchange rates, the market microstructure approach performs better in explaining exchange rate changes over short time horizons. The microstructure approach implies that information in order flow drives the dynamic processes of price evolution. This paper studies the informativeness of order flow under different market conditions in the foreign exchange market. We find that order flow tends to be more informative when the market experiences large bid-ask spreads, high volatility or low trading volumes. We also identify nonlinearities in the relationship between order flow and foreign exchange rate changes and find that these nonlinearities can be captured by the Interaction model and the Logistic Smooth Transition Regression (LSTR) model