An Introduction to Time Series Models

Time series econometrics is a branch of econometrics that has undergone many developments over the last 40 years. This chapter offers an introduction to time series models. After laying down a number of definitions, it focuses on the essential concept of stationarity. It presents the Dickey-Fuller unit root test for testing the non-stationary nature of a time series. The chapter then exposes the basic models of time series – the autoregressive moving-average models (ARMA models) – and the related Box and Jenkins methodology. A multivariate extension is proposed through the presentation of VAR (vector autoregressive) models. Finally, the concepts of non-stationary time series econometrics are presented by studying the notions of cointegration and error-correction models. Several empirical applications illustrate all the notions.