Модели сценарного прогнозирования валютного курса рубля

In modern days, predictive models are widespread. They allow determining the future values of variables of interest based on previous dynamics, for instance, analysis of patterns or signals. An obvious field for their application is the financial market, which allows you to make a profit thanks to information about future movements of assets. One of the most accessible investment assets is currency due to the implementation of the instrument for buying currency through commercial banks, as well as a rather low minimum threshold for monetary investments. Thus, the profit is directly related to the model used for forecasting. The article proposes a new approach to modeling the exchange rate, which is aimed at private investors with a medium and long planning horizon. The approach is based on modeling the main accounts of the balance of payments and on the assumption that imports are a constant share of the country's gross domestic product. Moreover, the model takes into account both the real and the financial sector of the economy, reflected in the balance of payments, oil prices, and the price index. To confirm the optimality of the approach, a more standard for the literature vector autoregression model was built, which, in the framework of this study, uses exchange rates of other oil-exporting countries and oil price. Based on in-sample comparisons and the root mean squared error, the new model is more accurate. According to scenario forecasting, expected cheap oil, predicted low investment income, and capital outflow from Russia are the key factors behind the appreciation of the exchange rate. Implementation of the policy helps to reduce the rate of devaluation, however, at the cost of a decrease in economic growth. Worsening of the investment climate for a short period or the standard policy of the central bank for managing reserves with the implemented restrictive policy leads to an increase in the exchange rate to a level without such a monetary policy. Simulations show that the exchange rate is more sensitive to negative events than positive ones.