Portfolio adjustment in response to real changes

Recent attention has been focused on the budget deficits of the United States which rose sharply in late 1981 and continued strong for over four years. This spending was funded primarily by selling bonds to domestic and foreign citizens. International linkages provided the transmission of government consumption through goods demand, and its financing through asset demand. Three concerns immediately arise. First, foreigners exchange their money for domestic currency to buy these assets. Second, federal spending, where goods are scarce, has price implications. Together, these two influences impinge on the exchange rate. It is the combined effect which needs better definition. As is known, the exchange rate is both the price of assets and commodities. When price is an issue, supply and demand conditions are important. If there is equilibrium in the financial part of the economy, certain responses will exist between goods and the exchange rate. When real concerns comprise the second part of the economy and they are in equilibrium, there is a response generated from assets. Both components are needed to fully describe movements in the exchange rate. This dissertation deals with the first part. Some general approaches have emerged which combine goods and assets, then proceed to analyze deficit induced capital flows. Concepts, monetarist in nature, produce poor econometric results. Complex integrated models are theoretically valuable, but are inestimable, leaving confirmation unsettled. Portfolio-adjustment models, originally posited by William H. Branson, produce testable results. It is this financial statement of the economy which is joined to an equation of real demand which becomes the structural model in this dissertation. The conclusions from it can be considered a technical gain. Estimation was accomplished with 2SLS over the time interval: 1973Q2-1987Q3. The following countries were paired: U.S. and W. Germany, U.S. and Japan, and W. Germany and Japan. Other parts of the study include Chow Testing for structural change and forecasting using 2SLS and an ARIMA process