Некоторые аспекты построения и использования динамических стохастических моделей общего равновесия (DSGE)

The article deals with practical aspects of constructing dynamic stochastic general equilibrium models (DSGE). A basic model of the real business cycle theory (RBC) is considered. The basic RBC model and DSGE models are compared, their similarities and differences are analyzed. By the example of the Russian economy it is shown that these classes of models are sensitive to smoothing parameters. In other words, the procedure of identifying the potential (equilibrium) GDP is crucial for the subsequent analysis of the economy's properties. It is shown that Russia's GDP can be adequately modeled by extracting the long-term component by the Hodrick-Prescott filter, using the AR model of order 4 and taking into account world oil prices. An algorithm for constructing a DSGE model suitable for use in university courses of macroeconomics and mathematical modeling, simple methods of deriving the dynamic IS-curve, New Keynesian Phillips curve and Taylor equation are analyzed. The DSGE model constructed is found to have 2 equilibrium states. The problems arising when forecasting based on models with the rational expectation operator are considered. Different approaches to solving equations with rational expectations are described: the Blanchard method, through reduction to adaptive expectations, through determining rational expectations. The conditions for stability of the model considered and for cyclical fluctuations are derived. A DSGE model parameters are calibrated for the economy of modern Russia. The effectiveness of monetary policy is analyzed, including anti-inflationary and stimulating policies. The social welfare function and the optimal parameters of monetary policy minimizing deviations from target inflation and GDP deviations from the long-term level are calculated. The article can be useful for specialists in economic-mathematical modeling and macroeconomic forecasting.