Combination Forecast On Consumption Of Livestock Products In China

Livestock products was an essential protein resource for dwellers in our country and to produce thesethings closely concerned with the national economy and people’s livelihood. However, frequentfluctuations of their prices in recent years had become increasingly stronger. For example, since 2003 China had underwent three entire hog cycles; each of them lasted for four years on average, while theprice increased by 50% at most. Belonging to “normal” business cycle, hog cycle resulted fromdelaying adjustment of supply to demand; Meanwhile, it generated inefficient allocation of productiveresources and restricted healthy development of husbandry industry. In accordance with microeconomictheory, fluctuations in the market of livestock products were caused by gaps between supply anddemand. Then research on consumption of livestock products gradually became a hot issue on accountof given supply.This paper generally reviewed researches on consumption of livestock products, discussed thevariations and influential factors, and introduced the basic principles of adopted models; thencalculated annual demand for livestock products according to data provided by FAOSTAT, thusacquiring the samples for modeling. On this basis, the paper conducted a specific empirical analysis ondomestic consumer demand for pork according to the research idea of combination forecast, discussedthe merits and faults of the chosen models according to empirical experience, and then forecast Chinesepork consumption in short run. In addition, this paper also studied the demand for other livestockproducts(including beef, mutton, poultry, eggs and milk) and examined the validity of combinationforecast. Given these conditions, the significances of research on domestic demand for livestockproducts were listed as follows. Firstly, it would help form reasonable industrial expectation andconsequently avoid blindly increasing or reducing production owing to short-term price changes.Secondly, monitor the changes of consumer demand, study the reasons for price changes and providenecessary evidences for decision-making.In accordance with the research, the paper gained conclusions as follows. Primarily, factors thatinfluenced consumption of livestock products in China mainly included income, price, urbanizationlevel and holidays. Secondly, genetic algorithm could solve parameters for support vector machine andachieve better goal. Thirdly, combination forecast could reduce the risk of single-model forecast andimprove forecast precision; thus it’s necessary to adopt combination forecast to solve relevant problems.Forthly, the forecast precisions were higher than that of ARMA model. Lastly, Chinese porkconsumption would annually increase by 2.33% from 2014 to 2016.In reference to previous experience, this paper proposed an original research idea, obtained someremarkable results and contribute several suggestions in relation to agricultural monitoring and earlywarning. Firstly, refined agricultural statistical work and established professional database foragricultural research. Secondly, built up necessary indexes for agricultural monitoring and improved themechanism of information publishing. Lastly, transform the routine regulatory measures and increasemethods for risk management.