| China is the leading country of pig production and pork consumption in the world. Stable development of the pig industry is very significant to promote the development of the animal husbandry, agriculture, also occupies a very important position in the national economy as a whole. In the period of2000-2011, three were sharp fluctuation of pig price in China, which forced up CPI. Growth of CPI,78%is due to the price of food, while pork price rose to59.8%(Jiang Hongyun). Thus, pig production fluctuation not only affects the macroeconomic operation, but also affects the social stability and economic development.This paper first introduces and analyzes China’s pig production conditions briefly. The number of pig raising households reduced greatly, but backyard farms (households) still accounted for an absolute majority, about more than95%. China’s pig scale is only about30%in2003, after several years of pig production structure adjustment, but pig scale has been close to70%in2010. Although the degree of the scale has been greatly improved, China’s pig breeding level, compared with international breeding level, there is still much room for improvement.The research contents of this paper are as follows:(1) Study on pig production fluctuation in China mainly adopted the research method of single index in the past. This paper makes a breakthrough in study method of single index, adopting three indexes, namely, amount of slaughter, amount of livestock on hand and amount of breeding sow livestock on hand. Three indexes are not isolated, and have the intrinsic relation. Thus, it can more accurately reflect the actual situation of pig production.(2) Using the H-P filter method (Hodrick-Prescott Filter), decomposing the volatility components of the three indicators, and studying its fluctuation cycle. Fluctuation of amount of livestock on hand and fluctuation of amount of breeding sow livestock on hand have the same fluctuations:the bottom time of the amount of slaughter and livestock on hand is basically identical. Using the Grainger causality analysis, we find that fluctuation of slaughter is Grainger causal relationship of fluctuation of livestock on hand and fluctuation of breeding sows livestock on hand; fluctuation of livestock on hand and fluctuation of breeding sows livestock on hand are Grainger causal relationship of fluctuation of the fluctuation of slaughter, and determining the influence direction and the influence degree among indexes.(3) This paper analyzes the production cost structure and benefits structure. Using Granger causality test, we find that fluctuation of the main cost, fluctuation of slaughter, fluctuation of livestock on hand, fluctuation of breeding sow livestock on hand have the following relationship: fluctuation of backyard and scale fine feed fees and fluctuation of slaughter have mutual Grainger causal relationship; fluctuation of backyard and piglet fees scale and fluctuation of slaughter have mutual Grainger causal relationship. Fluctuation of backyard death damages and fluctuation of slaughter have only a one-way Grainger causal relationship. In dynamic panel data model, fluctuations of backyard fine feed fees and backyard death loss fees have the most obvious influence on the amount of slaughter. Fluctuation of backyard and scale fine feed costs, fluctuation of backyard piglet fees, fluctuation of scale death loss fees, fluctuation of scale employment costs and amount of livestock on hand have mutual Grainger causal relationship; fluctuation of scale piglet fees, fluctuation of backyard medical treatment fees, fluctuation of scale death loss fees and fluctuation of amount of livestock have only one-way Grainger causal relationship. In the dynamic model, backyard and scale fine feed costs and backyard death loss fees have obvious effect on amount of livestock on hand. Fluctuation of backyard and scale piglet fees, fluctuation of scale death loss fees and amount of breeding sows livestock on hand have mutual Grainger causal relation; fluctuation of backyard and scale health and epidemic prevention costs, fluctuation of backyard death loss costs, fluctuation of scale medical and epidemic prevention costs and fluctuation of breeding sows have one-way Grainger causal relation. In the dynamic model of amount of breeding sows livestock on hand, backyard piglet fees, medical and epidemic prevention costs and scale piglet fees have obvious influence on the amount of breeding sows livestock on hand. By analyzing the effect of fluctuation of pig production yield on fluctuation of pig production, we find that backyard and scale pig yields have not Grainger causal relation with the amount of pig slaughter; fluctuation of backyard and scale pig yields have two-way Grainger causal relation with the amount of livestock on hand; fluctuation of backyard and scale yields have mutual Grainger causal relation with the amount of breeding sows livestock on hand. In the three dynamic models only returns to scale has a significant effect on the amount of breeding sows livestock.(4) Public health events have little obvious impact China’s pig production fluctuation from the long view, while there is short-term effect, and the effect is localized. Effect of the epidemic on the pig production is larger, and the time of impact is very long. Influence of national policy on pig production has no still clear conclusion, and different scholars disagree on pig production policy. Using the system dynamics, we simulate and analyze some factors such as pig diseases and pig production policy that can’t be quantized. The simulation results reveal that, under the conditions of pig production without delay, pig supply can be very good to adapt to the change of market demand, not cause sharp fluctuations in the amount of pig slaughter. This shows that, the process delay of pig production is one of the causes of the fluctuations of the amount of pig slaughter. Process delays of pig production make the amount of pig slaughter become frequent. The occurrence of major epidemic is also cause of pig production fluctuation. The occurrence of major epidemic changes the trend of breeding sows livestock on hand. Fluctuation frequency of pig livestock on hand does not change, but the amplitude of fluctuation has changed. Using the system dynamics model, we simulate the subsidies to breeding sows and frozen meat reserve policy which can’t be quantized.(5) Compared with the effect of BP artificial neural network method (BP artificial neural networks (ANN)) and support vector machine method (support vector machine (SVM)) on amount of pig slaughter, amount of pig livestock on hand and amount of breeding sows livestock on hand, we determine to adopt the method of support vector machines to establish pig slaughter prediction model, while the BP manual neural network method is useful for prediction of amount of pig livestock on hand and amount of breeding sows on hand. We predict the trend of amount of pig slaughter, amount of pig livestock on hand and amount of breeding sows livestock on hand in2013. Based on the above research, with the basic framework of pig production risk early-warning theory, this paper forecasts the risk of pig in2013and in2014by using quantitative warning method.The innovation points of this research:1. In the method, using the Grainger causality test model, analyzing pig material and service costs, artificial costs, the Grainger causal relationship between pig production yield and pig production fluctuation, the direction and amplitude of fluctuation among, using dynamic panel data model for empirical analysis, using system dynamics simulation analysis for the factors which can’t be quantized.2. In content, breaking through the measure of pig production fluctuation with the single index, adopting the three indicators such as amount of pig slaughter, amount of pig livestock on hand, amount of breeding sows livestock on hand to measure the fluctuation of pig production in China. Based on support vector machine model production and BP artificial neural network model, we predict China’s pig production, and conduct the early warning system research to the fluctuation of pig production based on the forecast.This paper presents policy proposal of pig production fluctuation regulation based on correlation analysis. From strengthening the construction of animal husbandry information network platform, perfecting and improving epidemic prevention system, strengthening the construction of feed industry, moderately improving the scale degree of backyard, issuing pig futures in a real time and strengthening the coordination of the macro-control measures of government departments at all levels, this paper puts forward some policy suggestions in six aspects. |
PDF Full Text Request