MANY PROBLEMS IN MACROECONOMICS, both of closed and open economies, are analyzed by means of expenditure and money demand functions. The properties of these functions play a major role in the derivation of results in these investigations. Although specifiers of such functions have in mind optimizing behavior by economic agents, explicit derivation of their properties from optimizing behavior is seldom undertaken. In many cases in which they are undertaken, it is assumed that utility depends on real money balance (e.g. [2]), an assumption which many find undesirable. There have been recently attempts to deal with macroeconomic issues by explicitly modeling the role of money in the economy thus avoiding the need to model it as an argument of the utility function-and by explicitly using optimizing behavior of economic agents (e.g. [3,5,6]). However, in those cases either the expenditure and money demand functions were characterized for steady states (e.g. [5,6]), or they were not derived because they were not required [3]. In this paper I derive an expenditure and a money demand function which arise from a problem of optimal allocation of consumption over time in which all payments are made in the form of money, there are liquidity constraints, and money is the only asset. Utility is derived only from consumption. These functions can be used to investigate macroeconomic problems as demonstrated

Three exchange-rate regimes--a float, a one-sided peg, and a cooperative peg--are evaluated and compared in terms of efficiency and welfare levels. The framework of analysis embodies country-specific monies, with the money of each country being used to transact in its commodity markets and its currency-denominated bonds. Welfare levels depend only on consumption levels. In the presence of perfect foresight all equilibrium allocations are Pareto efficient. In a floating exchange-rate regime the perfect foresight equilibrium allocation coincides with an equilibrium of a costless barter economy. The same result holds in a one-sided peg if the pegging country's exchange-rate stabilizing authority breaks even over time. In a cooperative peg regime there is a different equilibrium allocation for each combination of exchange-rate levels and monetary policies. Problems of policy coordination and conflicts in desired monetary policies are discussed.

[Three exchange-rate regimes--a float, a one-sided peg, and a cooperative peg-are evaluated and compared in terms of efficiency and welfare levels. The framework of analysis embodies country-specific monies, with the money of each country being used to transact in its commodity markets and its currency-denominated bonds. Welfare levels depend only on consumption levels. In the presence of perfect foresight all equilibrium allocations are Pareto efficient. In a floating exchange-rate regime the perfect foresight equilibrium allocation coincides with an equilibrium of a costless barter economy. The same result holds in a one-sided peg if the pegging country's exchange-rate stabilizing authority breaks even over time. In a cooperative peg regime there is a different equilibrium allocation for each combination of exchange-rate levels and monetary policies. Problems of policy coordination and conflicts in desired monetary policies are discussed.]