Research On The Influencing Factors And Prediction Of Hog Price Fluctuation In China

In recent years,the price of live pigs has been affected by many factors such as the internal and external environment,with high and low price fluctuations and sharp fluctuations in production and supply.At the same time,exogenous factors,mainly pig epidemic,have blossomed at multiple points around the world.The mix of factors has not only had a direct impact on residents,farmers and the economy,but has also posed major challenges to China’s social stability and sustainable development.The hog industry is an important part of China’s agriculture and one of the most important sources of meat for urban and rural residents in China,and the study of the influencing factors and forecasting of hog prices is of great practical significance to China’s social stability and sustainable development.Therefore,it is necessary to strengthen the price forecasting of live pig market to stabilize the market price and ensure the sustainable development of live pig industry.In order to explore the best forecasting model and optimal results of hog prices,this study first decomposes the time series of hog prices based on spider web theory and equilibrium price theory,and obtains the seasonal characteristics,random factor fluctuation characteristics,long-term trend characteristics and cyclical trend characteristics of hog price fluctuations in China.At the same time,in order to extract the main influencing factors of hog price fluctuation,a Pearson correlation coefficient matrix heat map was used to analyze the covariance among the influencing factors and the degree of influence on hog prices,and the main influencing factors were extracted from the 16 characteristic variables to form a multi-factor forecasting data set for hog prices.On the basis of mining the fluctuation pattern of pork price and extracting the main influencing factors,four single forecasting models,namely Random Forest(RF),Multi-Layer Perceptron(MLP),Support Vector Regression(SVR)and Autoregressive Integrated Moving Average(ARIMAX),and a combined ARIMAX-RF model are respectively applied to fit the data in order to find the best model for pork price forecasting analysis and its optimum.The results are used to make scientific forecasts of pork prices.The main results of this paper are as follows:(1)From the trend of hog price fluctuation in China,it is found that China has experienced five complete cyclical fluctuations between 2003 and 2022,and the length of its cycle is roughly 3～4 years.Overall,it has a long-term nature and is currently in the slow upward phase of a new pig cycle.From the perspective of seasonal factors,the seasonal fluctuations of hog prices are large,and there is a difference between low season and high season.In particular,the price of hogs can fluctuate drastically under the influence of special events,with irregular fluctuation characteristics.(2)In order to extract the main factors affecting the fluctuation of hog price,17 index factors were selected from three major categories,namely supply,demand and epidemic factors,and Pearson correlation coefficient analysis was applied to arrive at the following four factors: piglet price,pork price(boneless meat),hog to grain ratio and hog breeding cost as the main influencing factors.(3)Through the comparison of forecasting models,we found that the combined ARIMAX-RF model outperformed the other models,with the coefficient of determination,root mean square error,root mean square error,absolute mean error and mean absolute percentage error of 0.916,3.778,1.944,1.535 and 0.081,respectively,indicating that the combined model can compensate for the differences between the single models compared with the single model,thus improving the forecasting accuracy.The combined model is able to compensate for the variability between single models,thus improving the prediction accuracy and stability.