Taiwan is an isolated island located in the south East Asia. Since Taiwan is lack of nature resources, thus, a huge part of the economy is export-oriented. To stimulate the economy to grow and activate the international trading, the Free Trading Agreement (FTA) is an activator to allow larger quantity of trading over the world. The foreign exchange rate plays the major role affecting the trade surplus in the export-oriented economic system. Hence, a stable and accurate foreign exchange rate forecasting model is important for the economic activity participants. In this paper, the event study method is used to examine 3 international trading related events including the Economic Cooperation Framework Agreement (ECFA), the Taiwan-Japan Bilateral Investment Arrangement (BIA), and the Agreement between Singapore and the Separate Customs Territory of Taiwan, Penghu, Kinmen and Matsu on Economic Partnership (ASTEP) signed between Taiwan and other participants. The foreign exchange rate forecasting models are built by the time-series methods and the computational intelligence method, namely, the Generalized Autoregressive Conditional Heteroscedasticity (GARCH), the Exponential Generalized Autoregressive Conditional Heteroscedasticity (EGARCH), and the Interactive Artificial Bee Colony (IABC), respectively. In the event study, the observation period is chosen to include 70 days for both pre/post-event. The Mean Absolutely Percentage Error (MAPE) value is used to examine the forecasting accuracy of the models. The experimental results indicate that the IABC constructed foreign exchange rate forecasting model is the most capable one to resist the impact caused by the specific events. In other words, the impact results in more significant forecasting error in the GARCH and the EGARCH models, but not in the IABC model.

This work is partially supported by the Key Project of Fujian Education Department Funds (JA15323), Fujian Provincial Science and Technology Project (2014J01218), Fujian Provincial Science and Technology Key Project (2013H0002), and the Key Project of Fujian Education Department Funds (JA13211). The authors also gratefully acknowledge the helpful comments and suggestions from the reviewers, which have improved the presentation.

Authors and Affiliations

1. College of Information Science and Engineering, Qingdao, China

   Pei-Wei Tsai, Jing Zhang & Yong-Hui Zhang

2. Fujian Provincial Key Laboratory of Big Data Mining and Applications, Fujian University of Technology, Fuzhou, China

   Pei-Wei Tsai, Jing Zhang & Yong-Hui Zhang

3. Department of International Business, National Kaohsiung University of Applied Sciences, Kaohsiung, Taiwan

   Li-Hui Yang & Jui-Fang Chang

4. Council of Indigenous Peoples, Executive Yuan, Taipei, Taiwan

   Vaci Istanda

Corresponding author

Correspondence toJui-Fang Chang.

Editors and Affiliations

1. Faculty of Computers & Information, Cairo University, Giza, Egypt

   Aboul Ella Hassanien

2. Dubai International Academic City, The British University, Dubai, United Arab Emirates

   Khaled Shaalan

3. CS Dept. Faculty of Computers and Inform, Suez Canal University CS Dept. Faculty of Computers and Inform, Ismailia, Egypt

   Tarek Gaber

4. Ahmed Orabi Square , Menouf, Egypt

   Ahmad Taher Azar

5. Faculty of Computer & Information Scienc, Ain Shams University Faculty of Computer & Information Scienc, Cairo, Egypt

   M. F. Tolba

Reprints and permissions

© 2017 Springer International Publishing AG

Policies and ethics