ABSTRACT
This chapter aims at shedding light on the impact of exchange rate regime on business cycle properties. In the spirit of Mundell (1961) and Fleming (1962), open macroeconomics teaches us that the propagation of real and monetary impulses depends on the exchange rate regime. The short run dynamics of macroeconomic quantities should then be affected by the degree of flexibility of the nominal exchange rate. Do flexible exchange rates imply more volatile aggregates by introducing an additional source of uncertainty or, in contrast, do they generate more stabilized fluctuations? Does the exchange rate regime affect the international comovement? This chapter provides some answers to these questions by examining business cycle properties that are affected by the shift in the exchange rate regime. This empirical study is a prerequisite to the modelling exchange rate regimes. This chapter draws on Flood and Rose (1995) and Baxter and Stockman (1989).
The latter document the consequences of the fall of the Bretton Woods System by comparing business cycle properties under fixed exchange rates (1960-73) and under floating rates (1973-86). Flood and Rose (1995) as well as Baxter and Stockman (1989) conclude that, following the fall of the Bretton Woods System, transition to floating rates leads to sharp increases in the nominal and real exchange rate volatility with no corresponding changes in the distribution ofmacroeconomic quantities. This finding adds to the “exchange rate disconnect,” that is, the discrepancy between exchange rate fluctuations and the behavior of its macroeconomic fundamentals (Obstfeld and Rogoff, 2000b). Furthermore, Baxter and Stockman (1989) do not uncover any systematic relationship between exchange rate regimes and cross-country interdependence. The originality of this chapter compared to Baxter and Stockman’s (1989) work
is twofold. First, the robustness of Baxter and Stockman’s (1989) findings is checked in this chapter since we abstract from the decade of oil shocks (1971-86). Since, Baxter and Stockman (1989) sample ends in 1985, the floating rate period coincides with major oil price changes so that one can hardly distinguish business cycle properties due to the common world disturbances from fluctuations inherent to the shift to the flexible exchange rate regime. This chapter overcomes this
difficulty by neglecting the 1971-86 period in the sample. Furthermore, we check the relevance of Baxter and Stockman’s (1989) conclusion with regard to another exchange rate regime, namely the European Monetary System (EMS). Are Flood and Rose (1995) and Baxter and Stockman’s (1989) stylized facts still relevant when one measures the impact of the EMS on business cycle behavior? Artis and Zhang (1999) gauge the consequences of membership to the EMS on comovement of industrial production indices. We extend their analysis by investigating the impact of the EMS on comovement of output, consumption and investment. Moreover, the sample consists of EMS countries as well as non-EMS countries in order to distinguish EMS-specific phenomena from general tendencies in the business cycle. More specifically, in the EMS case, the flexible exchange rate period ranges from 1971 to 1979. Comparing business cycle properties of non-EMS vs EMS countries allows to control for the impact of theseworld disturbances. Should the oil prices play a major role in the shift in business cycle properties, this shift would be observed in EMS as well as non-EMS countries. Finally, the UK, Greece andSwedenhave never beenpart of theEuropeanExchangeRate arrangement. The analysis of business properties in these countries allows to distinguish between the impact of the European economic integration and the consequences of the EMS. The second originality of this chapter lies in the statistical technique used to
characterize business cycle changes. After characterizing the short run behavior of macroeconomic aggregates across exchange rate regimes, one has to assess the statistical significance of the evolution of business cycle statistics. Due to the uncertainty on the distribution of the variables under study, we resort to bootstrap techniques. This nonparametric method allows to measure the statistical significance of changes in business cycle properties observed under pegged and floating rates. After presenting the methodology (in Section 7.2), we investigate the impact
of the Bretton Woods System and the EMS on volatility (in Section 7.3) and international comovement (Section 7.4).
