Forecasting and Constructing Confidence Intervals Using Neural Networks─A case study of the US exchange rate

• Main Authors: Ku, Chun-Ming, 古峻明
• Other Authors: 李天行
• Format: Others
• Language: zh-TW
• Published: 1999
• Online Access: http://ndltd.ncl.edu.tw/handle/02638818554373192423

碩士 === 輔仁大學 === 應用統計學研究所 === 87 === The development of Taiwan economy, as being an island in the Asian Pacific, heavily depends on international trade. And hence foreign currency exchange rates stand in tight relation to the profit margin of traders. From the traders’ point of view, fixed or stable exchange rates will reduce the risk in making remittances. Historical Data demonstrates that NTD rapidly appreciated from 40:1 to 25:1 against USD since the Central Bank opened the fixed exchange rate marked in 1987. The exchange rate of USD tended to stay stable after the market became more and more mature. However, as the result of the financial crises originated from Southeast Asia, the exchange rate varied violently for the past twenty mouths and import/export traders suffered heavy losses. Forecasting The purpose of this research is to present a novel semi parametric prediction system for the US exchange rate. The prediction methods incorporated into the system consist of a neural network model that estimates the trend, as well as a Box-Jenkins prediction of the residual series. In terms of the adaptability of the Box-Jenkins methodology, the prediction intervals of the system can be successfully constructed. To demonstrate the effectiveness of our proposed method, the USD exchange rate from July 1997 to February 1999 was evaluated using a neural network model. Analytic results demonstrate that the proposed method outperforms the tradition parametric methods.