Time series econometrics deals with time series data that poses challenges due to non-stationarity. There are three types of stochastic processes - stationary, purely random, and non-stationary. Random walk models including random walk with and without drift are examples of non-stationary processes. A unit root stochastic process refers to non-stationary time series. Time series can be either trend stationary or difference stationary. Failing to account for non-stationarity can result in spurious regressions with high R-squared but no meaningful relationship between variables.