Order flow, information and trading behaviour in foreign exchange and equity markets
It has been shown in the literature that under asymmetric information, trading process itself is a part of pricing mechanism and order flow is the vehicle of information transmission and has a profound impact on prices. This thesis is composed of three major closely connected parts on order flow economics: (1) exchange rate determination and inter-market order flow effect, (2) market conditions and order flow impact and (3) limit order execution and microstructure factors. The first part of this thesis empirically investigates the price impact of order flow in four major currency markets and the results show that order flow has strong impact on exchange rates in all four markets and over various sampling frequencies. In a new result, inter-market effect is discovered where exchange rate movements in one market can be explained by the order flow in other relevant markets. In terms of forecasting ability, the order flow model out-performs random walk model that has so far beaten all macro-based exchange rate models. The second part addresses the dependence structure between flow and price change in the FX markets and finds that flow-return relationship is not linear as assumed in the previous literature. Order flow tends to be more informative and has larger impact on prices when market spreads are large, volume is low or volatility is high. These results cannot be fully explained by existing micro structure models. The last part of thesis studies how limit order execution probability is affected by microstructure factors. Using the tick data from the London Stock Exchange, it is demonstrated that price aggressiveness, spread and potential market pressure have significant impacts on the limit order execution.