Estimation and inference in nonlinear nonstationary panel data models

Within this PhD research the focus was on estimation and inference method for economic panel data that contains observations
on multiple firms or individuals observed over multiple time periods. The studies investigate how the classical estimation
method performs under the presence of nonlinear economic relationships among nonstationary economic variables. We
differentiated between two types of nonlinear functions: asymptotic homogeneous functions and integrable functions. We
both prove in theory and demonstrate by simulation studies that the classical nonlinear least squares estimation perform
well and normal inference can be conducted. The theoretical findings in this Phd research could be used by empirical studies
on economic cycles, exchange rates, investment and capital market imperfections and so on.