Exchange Rates
Currency substitution

The phenomenon of currency substitution, whereby the residents of one country hold foreign as well as domestic currencies in their portfolios, has assumed growing importance in recent years. Currency substitution, it has been argued, is a critical factor in the volatile behaviour of exchange rates. It has also been shown to have radical implications for monetary policy, undermining the basis for targeting domestic monetary growth and giving more emphasis to intervention in the foreign exchange market.

In Discussion Paper No. 107, Research Fellow Michael Artis and Shaziye Gazioglu use simulation techniques to study currency substitution in the context of a model of two interdependent economies. In Artis and Gazioglu's model inflation is governed by the interaction of domestic wages, set in overlapping wage contracts in which excess demand pressure is transmitted by 'Phillips'-style effects, and overseas inflation, which is transmitted by mark-up pricing. Each government issues a single financial asset, its 'currency', which is held by its own private sector and by the private sector and government of the other country. Each sector's demand for the two assets (domestic and foreign currency) is specified in detail in the model. The supply of each currency is determined by the fiscal deficit of the government concerned and by the domestic currency counterpart of increases in that government's holdings of foreign currency. Aggregate demand is modelled in a conventional way in Artis and Gazioglu's model. Private sector real wealth is an important determinant of demand, and this implies, via the accounting identities, an exact link between the budget deficit and the deficit on the current account of the balance of payments. The model therefore has a 'Reaganomics' flavour: a rise in the budget deficit which leads to a currency appreciation also leads to a rise in the balance of payments deficit. Despite the assumption of a floating exchange rate in the model, domestic inflation ultimately depends not only on domestic but also on foreign financial policy because of the linkages between the two countries. Even in the long run flexible exchange rates do not permit monetary interdependence, given the assumptions of the model.

Economic theory, as well as recent experience, underlines the importance of the volatility of exchange rate responses to shocks. Artis and Gazioglu simulate 'high currency substitution' variants of their basic model in order to investigate how the volatility of the exchange rate depends on the degree of currency substitution. Their results indicate that the degree of exchange rate volatility (the amount of under- or overshooting) depends on the nature of the shocks imposed on the model. The influence of currency substitution is also sensitive to the nature of the shocks in their simulations. In some cases a higher degree of currency substitution actually leads to a dampening of exchange rate volatility, large enough to convert overshooting into undershooting.
Artis and Gazioglu outline how their basic model could be extended to include a second traded asset in each country (a bond). The model could also be used to study the 'switching in' of an exchange rate stabilization rule in place of the floating rate regime. The parameters of the inflation processes could also be varied, to study the effects of differing degrees of price 'sluggishness' and real wage rigidity.


Currency Substitution in a Two-Asset Two-Country Model: A Simulation Approach
M J Artis and S Gazioglu

Discussion Paper No. 107, May 1986 (IM)