بررسی مداخلات ارزی در ایران: رویکرد رگرسیون داده‌های ترکیبی با تواتر متفاوت

یکی از جنبه‌های مهم سیاست‌های ارزی، سیاست مداخله بانک مرکزی در بازار ارز است. بانک مرکزی از طریق تغییر در ذخایر خارجی خود، جهت ایجاد ثبات در بازار ارز مداخله می‌کند. هدف این پژوهش برآورد تابع واکنش مداخلات ارزی بانک مرکزی ایران است. برای این منظور از الگوی رگرسیون ترکیبی با تواتر متفاوت (میداس) که امکان می‌دهد، متغیرهای سری زمانی با تواترهای متفاوت کنار هم در یک رگرسیون قرار گیرند، استفاده شده است. بر این اساس با استفاده از داده‌های مربوط به سال‌های 1381 تا 1396 برای متغیرهای مداخلات مستقیم ارزی (سالانه)، سطح ذخایر ارزی (فصلی)، صادرات نفت و گاز (فصلی) و نوسانات نرخ ارز (فصلی) تابع واکنش مداخلات ارزی بانک مرکزی برآورد شد. نتایج نشان داد که در میان متغیرهای موردآزمون، میزان مداخله سیاست‌گذار، حساسیت بیشتری به میزان صادرات دارد و همراه با افزایش صادرات نفتی، مداخله بانک مرکزی در بازار ارز افزایش می‌یابد. همچنین، طی دوره زمانی مورد بررسی، مداخلات بانک مرکزی عمدتا از نوع مداخلات ناهمسو بوده است. به علاوه، با استفاده از سه روش وزن دهی آلمون، بتا و آلمون نمایی به بررسی عملکرد متغیرها جهت پیش‌بینی مداخلات ارزی بانک مرکزی پرداخته شد. نتایج نشان داد که بر اساس شش معیار ارزیابی، متغیر صادرات نفتی از نظر پیش‌بینی عملکرد بهتری ارائه می‌دهد. هم‌چنین، مقایسه عملکرد پیش‌بینی‌های انفرادی و ترکیب پیش‌بینی‌ها با استفاده از روش رتبه‌های میانگین مجذور خطا نشان داد ترکیب پیش‌بینی‌ها که تاثیر تمام متغیرهای مورد آزمون را منعکس می‌کند، نسبت به پیش‌بینی‌های انفرادی عملکرد بهتری ارائه نمی‌دهد.

Foreign exchange intervention in Iran: Mixed frequency data sampling (MIDAS) approach

Introduction: One of the important aspects of foreign exchange policy is the Central Bank intervention in the foreign exchange market. The Central Bank intervenes to stabilize the foreign exchange market by changing its foreign reserves. Governments prefer to keep the exchange rate stable because any sudden fluctuation can destroy the confidence of economic actors in the market and harm the financial market and the market for physical goods. In this regard, an important method of analyzing the behavior of Central Bank officials to control and manage the foreign exchange market is to estimate the reaction function and identify the factors affecting foreign exchange interventions of the bank. Knowing about the factors affecting foreign exchange interventions helps to forecast the next interventions of the Central Bank and gives economic actors a better understanding of the behavior and decisions of policymakers and their effects on financial markets and macroeconomic variables.Methodology: In this study, a Mixed Frequency Data Sampling Model (MIDAS) has been used to investigate the factors affecting the foreign exchange interventions of the Central Bank. This method deals with the highfrequency variables as independent variables next to the lowfrequency dependent variable. The frequency of the dependent variable must always be less than the frequency of the independent variable(s). Therefore, the Midas approach can use the maximum amount of information gained from highfrequency series. Better prediction is also made for the dependent variable. In this research, the data related to the years 2002 to 2018 were used to estimate the reaction function of foreign exchange interventions by the Central Bank. This was done with the variables of direct foreign exchange interventions (annual), the level of foreign exchange reserves (seasonal), oil and gas exports (seasonal), and exchange rate fluctuations (seasonal).Results and Discussion: As the results showed, among the studied variables, the rate of policymaking intervention was more sensitive to the amount of export. Also, with an increase in oil exports, the Central Bank’s intervention in the foreign exchange market increases. In the period under review, the central bank interventions were mostly leaningagainst–thewind. In addition, the three methods of weighting Almon, Beta, and Exponential Almon were used to compare the performance of the variables to predict foreign exchange interventions of the Central Bank. In order to compare the performance of individual forecasts and the combination of forecasts, the mean squared error rank method was used. As the results showed, the combination of forecasts did not provide better performance than individual forecasts. Therefore, the individual estimates of foreign exchange interventions are highly valid.Conclusion: Through examining the reaction function of the Central Bank of Iran interventions, one can observe the irregular behavior of monetary authorities in the face of different market conditions. The interventions of the Central Bank in Iran obey no specific rules, and the foreign exchange policymakers intervene in the foreign exchange market in a completely discretionary manner, without considering a specific pattern for intervention. Although this form of intervention in periods of abundant foreign exchange earnings has been able to stabilize the exchange rate to some extent, with a decline in the foreign exchange earnings and foreign reserves, the power of the monetary authorities to manage the exchange market has diminished and the exchange rate has jumped to adapt to the realities of the economy. Therefore, it seems necessary for the Central Bank to regulate the method of foreign exchange interventions. The necessary condition is to synchronize the growth of liquidity in Iran with global liquidity and prevent its incompatible growth with the principles of the domestic economy. In fact, one of the most important prerequisites for successful foreign exchange market management is to control the unbridled growth of liquidity. In this case, we can hope for the success and effectiveness of foreign exchange interventions. The management of foreign exchange resources during periods of boom and bust of oil revenues is of particular importance too. The government’s injection of surplus foreign exchange earnings into the foreign exchange reserves will enable the Central Bank to use the resources accumulated in the accounts to manage the foreign exchange market in the event of oil shocks and foreign exchange earnings decline. This policy will help to prevent exchange rate fluctuations due to oil fluctuations. Derivatives can also be used as an alternative to cash market intervention. The advantage of derivatives is that their use does not necessarily change the country’s foreign exchange reserves, nor does it overshadow the monetary policies. Therefore, the bank can change the exchange rate or prevent fluctuations by issuing derivatives without affecting the monetary basis.