The removal of capital controls among European Community countries, due to occur by 1 July 1990, and the ongoing deregulation of financial markets is likely to have far-reaching effects on money demand in different countries and on the management of the EMS. Since almost all forms of regulation to limit free currency diversification would be inconsistent with the objective of liberalizing financial markets, we should assume that European residents will be able to hold cheque accounts and cash in the currencies of their choice. In order to understand the consequences of liberalization in this respect, it is important to identify the determinants of money demand in integrated economies, the role and nature of currency substitution, and the processes through which `vehicle' or `dominant' currencies emerge. These problems were discussed at a CEPR workshop on `Money Demand and Currency Substitution', held in London on 21 October. The workshop was organized by Alberto Giovannini (Columbia University and CEPR) and financial support was provided by the Ford and Alfred P Sloan Foundations, as part of their support for the Centre's International Macroeconomics programme.
The first paper of the workshop, `Currency Substitution and Vehicle Currencies: Tests of Alternative Hypotheses', was presented by Stephen Thomas (University of Southampton) and Michael Wickens (University of Southampton and CEPR). Thomas and Wickens reviewed various theories of currency substitution before testing the empirical support for each. These theories have very different implications for the determinants of money demand. In the `pure substitution' theory, domestic and foreign currencies are held in order to provide money services, and relative demands are proportional to their relative cost, i.e. the nominal interest rate differential. A variant of this theory assumes that holdings of money earn a zero nominal rate of return and hence a negative real rate. This approach suggests that relative money demands depend not only on interest differentials but also on inflation differentials or (supposing either instantaneous purchasing power parity or uncovered interest parity) the expected rate of depreciation of the exchange rate. In portfolio theories of money demand, in contrast, the opportunity cost of holding money is the interest forgone in not holding bonds.
Thomas and Wickens also developed a third theory that emphasized the role of `vehicle' currencies such as the US dollar, which is used to finance many transactions that do not involve the United States. Dollars are also used in the world pricing of many commodities, such as oil, and to denominate foreign debt. The demand for a reserve or vehicle currency must also therefore take account of the volume of international activity and the total value of debt denominated in that currency. The Deutschmark or the ECU could also assume such a role in the future.
In their empirical work, Thomas and Wickens explored the demand for M1 in the United States, West Germany and Japan and the demand for dollar balances held by non-US residents. The tests involved estimating the significance of variables representing the various alternative explanations of money demand. The most important result of the authors' regressions was the statistical significance of variables intended to capture portfolio and transactions demand effects: foreign rates of interest, the volume of exports and OECD output. The empirical significance of currency substitution lay in the role of currencies as providers of transactions services to foreigners, they concluded.
Nobuhiro Kiyotaki (University of Wisconsin-Madison) presented his paper on `A Contribution to the Pure Theory of Money', written with Randall Wright (University of Pennsylvania). Kiyotaki and Wright discussed the foundations of the transactions demand for currency and, in particular, the conditions under which individuals are willing to forgo real resources in order to hold `fiat money'. Fiat money is an intrinsically useless asset used as a means of exchange, which is neither based on any useful commodity (e.g. gold or silver coin) nor convertible into any commodity-based money nor based on the credibility of an issuing authority, such as a central bank.
They analysed a general equilibrium model incorporating Jevons's `double coincidence of wants': that is, under a barter system, a transaction can only take place when the set of goods or commodities one individual is willing to exchange is exactly what is required by the other party. The existence in Kiyotaki and Wright's model of a large number of differentiated products and of variations in individuals' tastes for those products ensured that pure barter was difficult, so money was a useful medium of exchange. Although some barter always coexists with monetary exchange, an individual holding money acquires a desired commodity faster than an individual holding some other good. The authors proved that the existence of equilibria in which money was held was robust even if fiat money was a less than ideal asset, as when it offers a lower rate of return than that available on other assets, or a less than ideal medium of exchange, as when it incurs high transactions costs, storage costs or taxes on the use of money. Even with these properties, money can continue to circulate and play a role in facilitating trade and improving welfare.
In `Would Competition Between Currencies Stabilize Exchange Rates and Prices?', Michael Woodford (University of Chicago) investigated the likely consequences of increased currency substitutability in a multiple currency, cash-in-advance model similar to that used by Lucas and Stokey to consider substitution between cash and credit. Household utility functions were defined over bundles of different types of goods in such a way that transactions using different currencies were treated as involving distinct types of goods. The degree of substitutability between these distinct commodities in household utility functions thus represented the degree of substitutability between the various currencies themselves. Woodford could then represent increasing substitutability of currencies, due to institutional or regulatory changes, as a change in the utility functions of households while holding constant other parameters in those functions.

Analysis of the model suggested that, contrary to widespread opinion, the degree of currency substitutability had little effect on equilibrium price levels and exchange rates except in the limiting case of perfect substitutability. But, by increasing the likelihood of exchange rate and price indeterminacy and the related problem of the existence of `sunspot' equilibria, increased currency substitutability could make management of fixed exchange rates more difficult.
Woodford argued that this cast considerable doubt on the wisdom of evolutionary approaches towards fixed exchange rates in Europe, based on the removal of all restrictions on intra-EC currency substitution and competition, as advocated by the UK Treasury. Movement towards markedly greater currency substitutability would be likely to increase the scope for speculative instabilities in exchange rates and corresponding fluctuations in price levels. In so far as currency competition resulted in lower rates of monetary growth, as the UK government argues, this in itself could make exchange rates and prices more unstable, owing to the greater scope for `sunspot' equilibria when monetary policies are contractionary. Rather than resulting in a natural, unplanned evolution towards fixed exchange rates, as the UK Treasury argues, too great a degree of currency substitutability is likely to make a system of fixed exchange rates harder and perhaps impossible to manage. In so far as Stage One of the Delors process towards EMU leads to a substantial increase in currency substitution, the adoption of a single currency may actually become necessary in order to preserve even the degree of monetary stability achieved under the current EMS.
Niels Thygesen (Kobenhavn Universitet) concluded the workshop by discussing policy-makers' perceptions of the problems that may be created by increased substitution between European currencies. While governments and central banks are seriously concerned about the disruptive effects of currency substitution and the consequent problems for the management of monetary aggregates, most of the work of the Delors Committee, of which Thygesen had been a member, had concentrated on the process of Economic and Monetary Union and appropriate policies by member countries to sustain it. Research on currency substitution would become extremely useful to European policy-makers as the various stages in the process of monetary integration are more precisely defined.