Revisiting Purchasing Power Parity for India using threshold cointegration and nonlinear unit root test

This study examines the Purchasing Power Parity (PPP) hypothesis in case of India for her five major trading partners over the period of 1991M<Subscript>1</Subscript>–2009M<Subscript>2</Subscript>. The study used the DF-GLS unit root test and threshold autoregressive (TAR) model as well as momentum-TAR (M-TAR) models for empirical analysis. However, we relied on TAR and MTAR models based cointegration tests to draw conclusions because of their superiority to traditional cointegration techniques as these models have limit cycles, amplitude dependent frequencies, and jump phenomena. These models are capable of producing asymmetric limit cycles and are suitable for time series data. Our empirical exercise reveals that PPP hypothesis does not exist for all major trading partners in case of India. This reveals that intermediate goods face high barriers to trade in this sampled countries. This supports the argument that Indian government has not been able to strike out the proper balance between flexibility and stability between real bilateral exchange rates and thus unable to maintaining confidence in the domestic currency that has been evident from the recent fall of rupee in relation to the US dollar. Copyright Springer Science+Business Media New York 2014