Testing for Structural Changes: A Cointegration Approach

Abstract

This study investigates whether exogenous shocks and/or economic reform programs such as the financial liberalization program of the early '80s resulted in changes in the dynamic equilibrium relationships among money, output, prices, and in some cases, interest rates. Cointegrating vectors for these variables were estimated over a full-sample period using quarterly data from 1981.1 to 1992.4 as well as a sub-sample period from 1986.1-1992.4. Various hypotheses regarding the neutrality of money on prices or output, the proportionality of the effect of money on prices or output, and the stationarity of velocity were tested using restrictions on the cointegrating vectors. If a regime change did occur after the exogenous schocks and/or the implementation of the financial liberalization program, the sub-period results would differ from the sub-sample results. The results obtained show that when M1 is issued, the results obtained for the post-shocks period are quite different from those for the entire sample period. Only the hypothesis regarding the absence of any effect from money to prices is rejected using a sub-period sample, whereas all the different hypotheses are rejected for the full-sample period regardless of the type of monetary aggregate used or whether the interest rate is included or not.