the design of an early warning system of currency crisis in iran: a logistic regression approach

the frequent occurrence of currency crises in recent years brought the early warning literature back in the researchers spotlight. in recent years, concept of an early warning system (ews) developed that should be able to identify various costly events, such as currency crises, early enough for policy makers to reduce the costs. this study attempted by using iran's economy quarterly data during the period 1988-2014 and using a model with discrete dependent variable, design an early warning system with all the required components on the iranian economy and explain it. the results of this study indicate that the designed system to greatly explain the determinants of a currency crisis in iran and has high ability in predicting this crises in the time periods studied. according to the results, currency crises in iran are due to a combination of different imbalances in the real and public sectors, external balance and the financial sector. based on these results, variables of ratio of bank loans to bank deposits, ratio of bank debt to the central bank to monetary base, inflation and industrial production growth (due to the high dependence on imports) are the largest and strongest role in increasing the probability of currency crises in iran. as well as, variables of ratio of bank deposits to liquidity, ratio of foreign exchange earnings to the central bank's foreign assets and real gdp growth are most important role in reducing the probability of a currency crisis in iran. jel classification: c25,c53, f31, g01