Further we showed that an exhaustive testing approach was sine qua non for taking a correct
account of the stochastic properties of the Turkish inflation data. First, although the implemented
stationarity and unit-root tests (ADF, KPSS, CH) and fractional integrations tests (GPH, LWE) point to
the stationary behavior of the Turkish inflation series from 2003 to 2009, the HEGY test demonstrates
that seasonal unit roots are present and thus stochastic seasonality is dominant in Turkish inflation.
Second, a structural break analysis (CUSUM, Chow test, QLR) indicates that the Turkish inflation data
exhibits a series of structural breaks with the latest being in mid 2003. In consequence to this, we
propose that the emphasis in any time-series modeling strategy should always be on rigorous testing
along the lines of a thorough test-test-test approach.