Endogenous structural breaks and unit root tests : an application to Singapore

Abstract

The seminal work on unit roots and macroeconomic variables was that of Nelson and Plosser (1982). They find that most macroeconomic variables follow a random walk process. The implications of their empirical findings are that existing theoretical relationships between macroeconomic variables are spurious and current macroeconomic models are misspecified. Perron (1994) tested for unit roots after allowing for structural breaks and found evidence disputing their claims. Perron empirically showed that the inclusion of structural breaks in the unit root tests will cause the macroeconomic variables to follow a deterministic trend, instead of following a random walk process.