| Over the last decade, U.S. business has become more involved in the global marketplace. This trend results from many factors including a slowdown in the growth rate of U.S. markets, the emergence of affluent new foreign markets for American products, and the availability of funding to develop the least developed world. This latter market, and to a lesser extent new markets in the developed world, are markets where the U.S. can apply its comparative advantages. To develop many of these markets, U.S. firms must enter into partnerships with host nation corporate partners.;Operating in this new global environment is not without risk. One major risk area is foreign currency exchange. Incurring costs in one currency and obtaining revenue in another exposes the firm to risk in the movement of one currency with respect to another. This movement can, in some cases, enrich the firm, but is more likely to cost the firm in terms of revenue, particularly if payment is made in host nation currency and there is a period of time between initiation of the business transaction and payment.;There are, of course, financial instruments designed to hedge the risk of this currency exposure, but these instruments add to the cost of doing business internationally.;This dissertation investigates a computer simulation of the foreign currency market and uses a model to predict the movement of foreign exchange in advance using contemporary data available to the general public. The model used to make the predictions is a Simulated Artificial Neural Network (SANN). SANNs, a development of artificial intelligence, is now being applied in many fields, including business, in which timely and accurate forecasting is required.;This research examines five currencies with respect to their value compared to the U.S. dollar. The network will be trained, tested and operated with data obtained from the U.S. Federal Reserve and the International Monetary Fund (IMF). End-of-month data for the period of January 1980 through December 1990 are used.;The objective of this study is to determine if Simulated Artificial Neural Networks can predict the movement of foreign currency exchange using publicly available data. |
