About estimates of the parameters of the GDP model based on the superposition of technological waves by N. D. Kondratiev

This publication is a continuation of the article (Andrukovich P. F., 2018), in which to builda model of GDP dynamics based on the superposition of long waves by N. D. Kondratiev it was proposed . This model, called the DV-trend, was further applied to describe the dynamics of US GDP for the period from 1790 to 2018 in (Andrukovich P. F., 2020b). It was noted that when estimating the parameters of the DV-trend, computational problems arise related to the structure of the model itself and with theoretical restrictions on the type of long waves of N. D. Kondratiev. In this article, these problems are discussed based on the analysis of a series of calculations for estimating the parameters of models built to describe the dynamics of GDP in the United States and great Britain. It is shown that the usual statistical methods for estimating parameters, even when calibrating a significant part of them, give values that do not correspond to the theoretical concepts of the form of long Kondratiev waves. In this regard, to obtain the correct values of the DV-trend parameters, it is proposed to use the expert method of their estimation (Andrukovich P. F., 2020a). The results of calculations to estimating the parameters of the DV-trend using this method when constructing models of GDP in the UK and the USA are discussed . It is shown that the model parameters obtained in this way correspond to the main provisions of the theory of long waves, both in terms of the shape of the technological waves themselves and the structure of their aggregate as a whole.