Structural change is an important problem in time series and affects all the inferential procedures discussed in the previous chapters. The area is so vast that it warrants a separate book. However, omitting it completely would leave an important gap in this book. We have therefore decided to cover it briefly.

This part has 3 chapters.

Chapter 13 on structural change, unit roots, and cointegration starts with a discussion of tests for structural change with known and unknown break points. We next discuss tests for unit roots under structural change. Both the classical and Bayesian approaches are covered. Next we discuss tests for structural change in cointegrated relationships.

Chapter 14 is on outliers and unit roots. We first discuss the different kinds of outliers and their effects on unit root tests. Given that outliers have serious effects on unit root tests, some outlier-robust unit root testing procedures as well as outlier-robust estimation procedures for CI systems, are discussed.

Chapter 15 is on regime switching models. These have been popularized by Hamilton who suggested using the Markov switching regression (MSR) models. This chapter discusses Hamilton's basic MSR model and several extensions of this basic model. The models in this chapter complement the outlier models discussed in the previous chapter.

One problem with the MSR model is that it models only sudden changes. In practice structural change is gradual. We, therefore, discuss model with gradual change. An important class of models with gradual change is the structural time series model introduced by Harvey.