In this thesis, I study microeconometrics and its applications in empirical questions. There are three studies, which investigate the housing vacancy rate, sticky demand, and the quantile forecasting methods for time series data. In the first study, I use a mixture density model to estimate the housing vacancy rate of 15.85% from housing-units level electricity consumption data. This high vacancy rate reflects the high volumes of investment demands in the Chinese housing market. In the second study, I relax a fundamental assumption in the literature of dynamic pricing, from regular demand to sticky demand and investigate the related economics outcomes. In the third study, I extend a novel nonparametric quantile forecasting method that is from being only applicable to i.i.d data to being more general applicable that allows for weakly dependent time series data. The simulated and empirical data illustrated the efficiency gain of my method over other existing nonparametric methods.
