The threshold effect of saving rates on the trade balance and the net non-oil export in Iran

Introduction: The continuous deficit in the non-oil trade balance in Iran has aroused the curiosity of researchers and politicians. Due to its abundant natural resources, Iran is heavily dependent on oil export revenues, which is not a real productive sector. On the other hand, the global oil market has always experienced instability and sharp fluctuations. The dependence of foreign exchange earnings in the country on oil exports has caused a drastic transfer of these instabilities to the supply of foreign currency and its market in the economy. This, in turn, has adversely affected the external sector and other macroeconomic variables. This issue has attracted the attention of policymakers to adopt policies to eliminate the trade deficit, especially without dependence on oil.
For any acceptable policy in this area, the trade balance function needs to be properly identified. Therefore, this study seeks to estimate the trade balance and the non-oil trade balance with an emphasis on saving rates and economic sanctions by using a nonlinear smooth transition regression model during the period of 1978-2017.
Methodology: In trade balance studies, the three variables of real exchange rate, domestic per capita income and foreign per capita income are often used as explanations of the trade balance. In this article, in addition to these variables, the two variables of saving rate and dummy variables for sanctions are considered as effective factors in two separate models for trade balance and non-oil trade balance.
In general, there are various reasons for the effect of savings rate on the trade balance, including the direct effects of savings rate on GDP and its indirect effects according to the theory of borrowable funds. According to this theory, with the increase of the saving rate, the supply of funds increases too, which reduces the interest rate. The reduction of interest rates leads to an increase in capital outflows, and, as a result, the demand for domestic money decreases. Thus, the depreciation of the domestic currency leads to cheaper exports and more expensive imports, which can lead to improved trade balance.
Various studies in recent years have shown that the trade balance model can follow a nonlinear process. In this regard, after examining the existence of a nonlinear relationship, the Smooth Transition Regression model is used to investigate the relationship between savings rate and the other determinants of trade balance. Therefore, the following model is estimated in the framework of the STR method:
 
 
In this equation, TB is the logarithm of Iran's trade balance, which is estimated in both total and non-oil trade balances in the framework of two separate models. Also, S is the saving rate, y is the logarithm of the domestic per capita real income, wy is the logarithm of the world per capita real income, and t is the dummy variables of sanctions (since Iran has always been sanctioned in the study period, two dummy variables of moderate and strong sanctions are used to measure the level of sanctions). In this equation,  is a transition function in the STR models, the values of which vary between zero and one according to the conditions of the economy and determine the nonlinear effect of the variables on the trade balance.
Results and Discussion: According to the results, the nonlinearity of both trade and non-oil trade balance models and the instability of the coefficients of the variables in these two models have been confirmed. When the savings rate is changed as a transition variable, the amount and sometimes the direction of the variables’ effect change too. In the first regime, the saving rate in both models had negative effects and, in the second regime, it had positive effects on trade balance. Because savings rates have often been above the threshold for many years, increasing them can lead to an improvement in the trade balance. The real effective exchange rate also had positive effect on both except that in the first regime of non-oil trade balance, but the magnitude of its effect varied in different regimes depending on the type of the trade balance. Thus, for proper policy-making with the real exchange rate tool, it is necessary to pay attention to the goals and conditions of the trade balance model at the same time.
The sanctions have had a significant effect on the total trade balance, but they have had no significant effect on the non-oil trade balance. The reason could be the oil nature of most of the sanctions imposed on Iran.
Conclusion: One of the most important results of this paper is the confirmation of nonlinear relationships in the trade balance model. Therefore, in order to avoid any error in estimating the effects of the implemented policies, it is necessary to pay attention to the change of coefficients over time and to the conditions of the economy. Increasing savings rates in most periods has led to improved trade balance. By properly directing the savings to production and removing the barriers to investment, the government can increase production and exports and thus improve the trade balance.