With the collapse of the Bretton Woods System in 1971, a new era has started in exchange rate predictions. In this era prediction of exchange rates, which is crucial in economic decision processes, became a major research topic. In this study alternative methods are explored to pre‐dict exchange values through Artificial Neural Networks. For this aim prediction performances of three different Artificial Neural Network models are examined; first model that is con‐structed by using the Lagged Values in time series models, second model that is constructed by using the variables of “Monetary Model”, which is a structural model to predict exchange rates, the last model that is constructed by using the “Purchasing Power Parity Model”, which is another structural model to predict exchange rates. In the study, exchange values of US Dollar and Euro are predicted against the Turkish Lira. Results of the study show that Artificial Neural Networks model, which was constructed by using the lagged values of exchange value variable, has the best prediction performance. Artificial Neural Network models that are constructed by using the variables of Purchasing Power Parity Model and the variables of Monetary Model are ranked as the second and the third in the predic‐tion performance ranking respectively.
Alan : Sosyal, Beşeri ve İdari Bilimler
Dergi Türü : Ulusal
Benzer Makaleler | Yazar | # |
---|
Makale | Yazar | # |
---|