Author(s)

Jing Yuan

Corresponding Author

Jing Yuan

Abstract

Economic forecasting has always been an important part of government, enterprises and individuals' decision-making. However, the deviation between the mainstream economic forecasting model and the actual economic performance has always been an important research field. The purpose of this study is to explore the deviations between different types of economic forecasting models and actual economic performance, and to analyze the possible reasons for these deviations. Based on the data of prosperity survey, industrial production and benefit statistics, and considering the inherent periodicity of the economic environment, this paper establishes three economic index forecasting models, namely IS-LM model, CAPM and BP, to forecast and analyze the sales revenue index of industrial products in the fourth quarter of 2020 in J province, and then uses multiple linear regression method to fit several forecasting models to obtain the final combined forecasting model. It is found that all kinds of economic forecasting models show certain forecasting bias in different situations. These deviations are partly due to the limitations of the model itself, including the assumptions, simplification and parameter estimation of the model. In addition, the model may also show shortcomings in predicting external shocks and emergencies, which may lead to deviations. Through empirical research, the combined forecasting model has achieved satisfactory results.

Keywords

economic forecasting; deviation; Combination forecasting