Modeling of monetary and credit system indicators of the Russian Federation in multidirectional scenarios of oil market dynamics

Abstract

The purpose of research – to build mathematical models that describe interrelationships between the key market indicators, significant for the Russian economy, and macroeconomic indicators of the monetary system. Materials and methods. In this study, we applied methods of model description, mostly used in control theory, meant for technical engineering, such as linear discrete transfer functions, space-state models and nonlinear Hammerstein-Wiener models. To identify these models, we used System Identification Toolbox from Matlab software package, mostly used for mechanical systems’ analysis. Based on the known input and output signals, a mathematical model was estimated. Time series of macroeconomic and market indicators for the period from January 10, 2008 to January 10, 2018 were used for identification. Results. Two prediction models were designed in this work. The first model describes a sequential transfer from the oil price and dollar- to-ruble exchange rate to the gross domestic product, then to M2 and then to loans. Dependencies between economic parameters are described by linear discrete transfer functions. There is only one difference in the second model’s general structure: the sequence of the last two transitions from the gross domestic product to loans, and then to M2. In addition, nonlinear Hammerstein-Wiener models describe last two transitions in the second model. As a result, predictions for macroeconomic indicators’ trends were given on different time horizons: three, seven and twelve years and with two differently directed scenarios of the oil market. The conclusion. Despite close values in the models accuracy estimation, they give similar results for matching scenarios, but different growth rates in general, in the forecast. Such a result in scenarios shows, that a sharp fall in oil prices has a stronger impact on given macroeconomic and market indicators, which, in its turn, shows the capability of the models to make correct trend predictions. In further studies, it is possible to move from macroeconomic indicators to their more particular components at meso- and micro levels.