Analysis and Forecasting of Cash Circulation in the Volga Federal District

Since the state of cash circulation is one of the key characteristics of the economic sphere, it is necessary to ensure a sufficient number of banknotes in the cash desks of Bank of Russia divisions to meet the demand of credit institutions in each region. At the same time, it is necessary to optimize cash flow in terms of reducing costs for excess production, storage and transportation of cash. To achieve this goal, the work constructed forecast models that allow us to estimate the expected value of the emission result in the context of denominations and regions of the Volga Federal District. The development of predictive models was based on multivariate regression based on the ARIMA model, as well as a neural network model with TFT architecture. The combination of forecasting tools used made it possible to identify the most accurate model for practical use in business processes, as well as to assess the feasibility of including additional macroeconomic indicators characterizing the business activity of each region in the forecast model.