Current Proceedings on Technology

Yazarlar: Cuneyt Sevim, Ozkan Bali, Serkan Gumus, Erkam Guresen

Konular:-

Anahtar Kelimeler:Early warning system,Currency crisis,Radial basis function networks,Support vector machines,Multi-layer perceptron,Logistic regression

Özet: People or firms, who can predict crises, can turn chaotic environment of crises into an environment with many opportunities. On the other hand, people or firms, who cannot predict crises, can face the negative effects of crises such as unemployment and getting poor. Especially for firms, having an early warning system for currency crises will enable them to plan raw material requirements, production or service capacity, stock levels, marketing strategies and cash requirements. With a good plan negative effects of crises can be minimized. The main purpose of this paper is to develop an early warning system to predict the currency crisis. In this study, a data set covering the period of January 1992 - December 2011 of Turkish economy is used, and an early warning system has been developed with Artificial Neural Network (ANN) models such as Multi-Layer Perceptron (MLP) and Radial Basis Function Networks (RBFN), Support Vector Machine (SVM) and Logistic Regression Models. Financial Pressure Index (FPI) is calculated as a result of percentage changes in dollar exchange rate, gross foreign exchange reserves of the Central Bank, and overnight interest rate. Financial Pressure Index is used as the dependent (crises) variable, while thirty-two macro-economic indicators are used as the independent variables. Three models, which are tested in Turkish case, have given clear signals anticipating 1994 and 2001 crises 12 months before. In addition, all three of the model predictions of a currency crisis, according to Turkey's economy, are not expected (ceteris paribus) until the end of 2012.

---