Summary of WP/92/71
“The Implications of Cross-Border Monetary Aggregation” by Jeroen J. M. Kremers and Timothy D. Lane
Some recent studies suggest the possibility of estimating a stable aggregate demand-for-money relationship for the group of countries participating in the exchange rate mechanism of the European Monetary System (EMS), which would facilitate the setting of policy targets for a European Central Bank.
This paper examines, within a theoretical framework, the implications of using data aggregated across countries to study money demand. Two sources of bias in estimates are considered: aggregation bias occurs to the extent that different countries in the group have different money-demand relationships, while specification bias occurs to the extent that there are omitted variables, errors in measurement of the explanatory variables, or other specification errors.
In the EMS, the possibility of currency substitution--and of international portfolio substitution more generally--may lead to specification bias in single-country money-demand estimates. Currency substitution means that residents of each country hold more than one country’s money, so that the demand for each country’s money depends on the incomes and interest rates of other countries as well as its own. International portfolio substitution means that each country’s residents consider foreign assets among the alternatives to holding money, so that a properly specified money-demand equation would include foreign as well as domestic asset yields as opportunity-cost variables.
The paper demonstrates how these specification errors may bias single-country money-demand estimates, especially by giving the impression of unduly slow adjustment of money balances toward their desired levels. The way in which cross-border aggregation may reduce specification bias at the cost of introducing some aggregation bias is also examined. While single-country money-demand estimates ignore the effects of currency substitution, cross-country aggregate estimates internalize this effect. Moreover, the analysis suggests that such estimates may support the view that currency substitution may be an important consideration in the stages leading up to economic and monetary union (EMU).