The Effect of Exchange Rate Uncertainty on Iran's Non-Oil Trade Balance: Stochastic Volatility Model Approach

Since the advent of the current float in 1973, most countries have been concerned about the uncertainty that floating exchange rates have introduced into world markets. Most European countries have tried to avoid this uncertainty by joining the Euro zone and adopting a common currency, which has flavors of a fixed exchange-rate system. However, most other countries still have to deal with the side effects of exchange rate uncertainty since their exchange rates float.
There are several channels through which exchange rate volatility could affect the trade flows. First, if traders are risk averse, they could reduce their activities due to exchange rate uncertainty in order to avoid any loss. Second, exchange rate uncertainty could directly affect the trade volume by making prices and profits uncertain, especially in countries where forward markets do not exist such as the developing world. Even if forward markets do exist in some industrial countries, some studies indicate that forward markets are not very effective in completely eliminating exchange rate uncertainty. Third, if exchange rate volatility persists over a longer period of time, it could induce domestic producers to switch buying from foreign sources to domestic sources, reducing the volume of trade, especially traded inputs. Finally, exchange rate uncertainty could also affect direct foreign investment decisions which in turn could lower the volume of trade. To reduce the price fluctuation due to exchange rate volatility, production facilities would be located near final markets, leading to change in pattern of trade.
Considering the numerous economic changes affecting the exchange rate in the studied period (1989-2015) in Iran, like the oil price shocks, the change of commercial policies, and changing foreign exchange policies and economic sanctions, studying the effect of exchange rate fluctuations on the non-oil trade balance is essential. The relationship of exchange rate and trade balance is explained by various theoretical approaches like the elasticity approach, absorption approach, Marshall-Lerner Condition, J-Curve approach, monetary approach and the two countries imperfect substitution model approach.
Of these, the two countries imperfect substitution model of the Rose and Yellen (1989) is used in this paper to model the relationship between exchange rate and trade balance. This approach shows the nature of the relationship of real exchange rate on trade balance in both short and long run. It stipulates that depreciation of the real exchange rate improves trade balance. Besides, the model assumes that there are no perfect substitutes in the imports and exports for the locally produced goods and services.
Rose and Yellen (1989) start with a specification of the import demand equations. As in Marshallian demand analysis, the volume of imported goods demanded by the home (foreign) country is determined by real domestic (foreign) income and the relative price of imported goods. Clearly, real income has a positive impact on the volume of import demand, and the relative price of imported goods has a negative relationship This study, investigates the effect of exchange rate uncertainty on Iran’s non-oil trade balance for the period 1989-2015. Stochastic volatility model (SV) is used for calculating uncertainty index of exchange rate. Then, the effect of exchange rate uncertainty on Iran’s non-oil trade balance is estimated using Rose and Yellen (1989)’s model and Johansen and Juselius (1990)’s cointegration procedure.
An alternative to GARCH-type models is the class of stochastic volatility models, which postulate that volatility is driven by its own stochastic process. Estimation and inference of SV models are more complicated than for GARCH models. On the other hand, SV models have some advantages compared with GARCH models. For example, SV models offering a natural economic interpretation of volatility are easier to connect with continuous-time diffusion models with SV, and are often found to be more flexible in the modeling of financial returns. A variety of estimation methods have been proposed to estimate the SV models, in this paper MCMC methods are used to estimate of SV model. The results show that all variables in trade balance model are integrated of first order or in short hand are I(1) and cointegration test suggests one cointegrating vector for this variables.
Briefly, the results of estimate vector show that real exchange rate, volatility of real exchange rate and GDP of Iran have negative effect on non-oil trade balance and GDP of the world has positive effect on Non-Oil trade balance. The results of estimating trade balance in long-run show that real exchange rate has negative effect on non-oil trade balance; therefore, the Marshall-Lerner condition is not met. In addition to, J-Curve is rejected due to impulse response functions. Also, uncertainty of the real exchange rate has negative effect on non-oil trade balance; therefore, policy makers must adopt policies to help stabilize the real exchange rate.