Development of structural fracture unit root tests: An application with macroeconomic data

Unit root tests with a structural break are among the topics that have been emphasized in time series analysis in recent years. This is because, according to unit root tests without a structural break, series usually contain unit root. However, unit root tests with structural break show that the reason for nonstationarity in the series may be a structural break. In the light of the developments in the literature, the fact that time series analysis is generally based on the assumption of linearity may cause the results of unit root tests to be erroneous. For this reason, unit root tests have been developed that can be used in Case the series is not linear. Accordingly, unit root tests based on the analysis of breaks in a time series with threshold autoregressive models are used. These unit root tests are called nonlinear unbroken unit root tests or Fourier unit root tests. Another important issue in the literature is the use of tests that determine whether the series are linear or not. As a result of these tests, unit root tests suitable for the series are determined and the results obtained are more accurate and consistent. In this study, unit root tests are examined in linear and non-linear forms within the framework of unbroken and break unit root tests in order to reveal the development of structural break unit root tests. Accordingly, in the first part, after explaining the concepts of stationarity and structural break within the scope of time series, unit root tests are presented with and without linear breaks. In the second part of the study, nonlinear unit root tests are theoretically explained as unbreakable and with break. In the third part of the study, the theoretical issues discussed in the first and second parts are applied to three macroeconomic series, and the results are interpreted both in the context of time series and economically.