Vector model of autoregression of indicators of industrial activity of a construction enterprise

The article analyzes the existing economic-mathematical models: correlation-regression analysis, production functions, systems of econometric equations; their general form, calculation formulas are shown, their strengths and weaknesses are revealed, a vector model of autoregression of the main indices of the production activity of the construction enterprise is proposed (labor productivity, product profitability, mechanical strength (technical level of construction machines and equipment), relative strength of the management team, timeliness of implementation works, discreteness of resource use, product cost, product quality) on the basis of the VAR model construction. As a basis for constructing a VAR-model of autoregressive indicators of the production activity of a construction enterprise, the authors suggest using a system of three interrelated equations. The advantages and disadvantages of the vector model of autoregression are presented, as well as the results of estimating the coefficients in the VAR model. The resulting values of the coefficients were analyzed using Granger's causality test, based on the analysis of the cause-effect relationship between time series. The article defines the impulse response function that describes the response of a dynamic series in response to some external shocks. The graphs of the responses of the main resultant indicators of the activity of the construction enterprise are constructed. The hypotheses put forward in the article are checked based on the use of the F-test and the LM-test. The authors analyze in detail the results of calculations and convincingly prove the relevance of the methodology proposed in the article.