In this paper, we empirically investigate whether multilateral adjustment to large U.S. external imbalances can help explain movements in the bilateral exchange rates of three commodity currencies – the Australian, Canadian and New Zealand (ACNZ) dollars. To examine the relationship between exchange rates and multilateral adjustment, we develop a new regimeswitching model that augments a standard Markov-switching framework with a threshold variable. This enables us to model the exchange rate dynamics of our commodity currencies in the context of two regimes: one in which multilateral adjustment to large U.S. external imbalances is an important factor driving the commodity currencies and the second in which there are no significant U.S. external imbalances and hence multilateral adjustment is not a factor. We compare the performance of this model, both in and out-of-sample, to several other alternative models. In addition to developing this new model, another distinguishing feature of our paper is that we estimate all of our models using a Bayesian approach. We opt for a Bayesian approach in this context because it provides a simpler and more intuitive means of evaluating and comparing our different non-nested models. Moreover, it is relatively straightforward using a Bayesian approach to evaluate the importance of nonlinearities in the relationship between exchange rates and multilateral adjustment. Our findings suggest that during periods of large U.S. imbalances, fiscal and external, an exchange rate model for the ACNZ dollars should allow for multilateral adjustment effects. Moreover, we also find evidence to suggest that the adjustment of exchange rates to multilateral adjustment factors is best modelled as a non-linear process.

Also published as:

Multilateral Adjustment, Regime Switching and Real Exchange Rate Dynamics
The North American Journal of Economics and Finance (1062-9408)
January 2014. Vol. 27, pp. 68-87