Kantil Birim Kök Testine Dalgacık Yaklaşımı ve Türkiye İçin Satın Alma Gücü Paritesi Hipotezinin Geçerliğinin Testi (original) (raw)

This study investigated whether there would be an improvement in size and power properties by applying the wavelet transform, which presents time and frequency information together, to the quantile unit root test developed by Koenker and Xiao (2004). For this purpose, the size and power properties of the test under different distributions, filters, and wavelet transforms were investigated by Monte Carlo simulations. According to the results of Monte Carlo simulations in small samples, the best performance under discrete wavelet transform (DWT) is with D(4) filter; the best performance under maximum overlap DWT is obtained with the Haar filter. In large samples, it has been observed that the difference between filters for both transformations disappears. Results When compared to Koenker and Xiao (2004)'s quantile unit root test, it is concluded that there is no improvement in the size and power properties of the test with DWT and MODWT transformations. In the application part of the study, the validity of the purchasing power parity (PPP) hypothesis for the top 10 countries in Turkey's exports as of 2018 was examined for the period of 01:2002-12:2018. For this hypothesis, which is reviewed with the stationarity of the real exchange rate (RER), traditional unit root tests (ADF, PP, KPSS), Fan and Gencay (2010) 's wavelet unit root test based on variance ratio, Koenker and Xiao (2004) 's quantile unit root test and wavelet-based quantile unit root tests were used. Traditional unit root test results indicate that the hypothesis is generally invalid for Israel, Russia, Poland, China, and Saudi Arabia; According to the wavelet unit root test, it was concluded that the PPP hypothesis was invalid for Israel, Bulgaria, and Saudi Arabia. According to Koenker and Xiao (2004) 's quantile and wavelet-based (DWT) quantile unit root tests, similar results were obtained except for the Euro Area, USA, and United Kingdom. When the results were evaluated together, the PPP hypothesis was invalid for Israel, Russia, China, and Saudi Arabia but valid for other countries. iii ÖNSÖZ Başta danışman hocam Prof. Dr. Selahattin Güriş'e, görüşlerinden yoğun şekilde faydalandığım Prof. Dr. Burak Güriş'e ve jüri üyelerine teşekkürlerimi sunuyorum. Ayrıca hem fikirleri hem destekleriyle tezime katkıda bulunan meslektaşlarım ve dostlarım Çağın Karul, Erdal Berk ve Şaban Kızılarslan'a binlerce kez teşekkürler. Son olarak desteğini ve sabrını asla esirgemeyen aileme… Üniversite hayatımın en büyük hediyelerinden olan Betül ile Nur'a… Arkadaşlarım Ebru, Enes, Cansu, Burak ile Can'a… ve bu tezin yazım sürecinde bir saniye bile olsa yolumun kesiştiği iyi gelen herkese teşekkürler.