Drei Aufsätze zu hochfrequenter Finanzmarktökonometrie und individuellem Handelsverhalten

In recent years high-frequency finance has become one of the most active research fields in finance and economics. The wide-spread availability of high-frequency datasets has particularly spurred research within this field and has, in turn, given birth to the rapidly expanding bridge between finance and econometrics, high-frequency financial econometrics. Developments in this field have addressed topics such as risk and liquidity measurement, market design, market microstructure and the general behavior of financial agents by applying quantitative methodologies. Advances in computing technology during the last two decades have given further momentum to these research efforts - not to be underestimated, as the detailed, precise and efficient collection of large high-frequency datasets as well as the development of computationally intensive estimation procedures lie at the forefront of current research in financial econometrics.Trading activity datasets which describe all trading actions in sets of assets for individual investors over a given time period are becoming increasingly available. These are more complex than standard high-frequency and limit order book datasets, allowing for the detailed analysis of individual trading behavior on the micro level. These datasets represent the limit of market microstructure information which can be made available to the financial econometrician. The analysis of such trading activity datasets requires advanced econometrics techniques able to account for their micro panel structure, with individual observations on different types of activities for several trading instruments being irregular in time.This dissertation advocates and contributes to the development of advanced econometric techniques for the characterization of individual trading behavior on the basis of complex trading activity datasets. Herein, three stand-alone papers address different aspects of individual trading behavior. They are all based on an unique trading activity dataset of foreign exchange market activity provided by OANDA FXTrade, an electronic trading platform and market maker in the foreign exchange market