论文部分内容阅读
本文通过建立一个宏观经济变量间的复合模型，模拟经济增长、货币供应、物价上涨以及通货膨胀之间的数量关系，为经济发展和货币供应政策提供模拟试验。该模型的特点是在经济计量模型中嵌入人工神经网络模型而形成复合模型，该模型共含６０个内生变量、９年外生变量、３９个随机方程，３９个人工神经网络模型。人工神经网络模型是在经济计量模型的基础上建立并分别进行训练。复合模型的联立动态模拟结果表明其精度比单纯的经济计量模型精度要高。研究结果指出通货膨胀一方面受到货币供应的影响，是多发货币的结果；另一方面又反作用于货币供应。通货膨胀对货币供应的影响皆呈负效应，而且对Ｍ２的影响较对Ｍ１更强些。货币供应的收入弹性和价格弹性都有预料的符号，而且Ｍ２的弹性分别大于Ｍ１的相应弹性，这些结果与亚洲其它６国得到的相应结论是十分一致的。
In this paper, by simulating the quantitative relationship between economic growth, money supply, price inflation and inflation, we set up a composite model of macroeconomic variables and provide simulation tests for economic development and money supply policy. The model is characterized by embedding artificial neural network models into econometric models to form a composite model. The model contains 60 endogenous variables, 9 exogenous variables, 39 random equations and 39 artificial neural network models. Artificial neural network model is built on the basis of econometric model and trained separately. Simultaneous dynamic simulation of the composite model shows that the accuracy is higher than the mere econometric model. The result of the research shows that inflation, on the one hand, is affected by the money supply and is the result of multiple currencies; on the other hand, it also acts against the money supply. The effects of inflation on the money supply all have negative effects, and the impact on M2 is stronger than that on M1. The income elasticity of money supply and price elasticity are both expected symbols, and the elasticity of M2 is greater than the corresponding elasticity of M1 respectively. These results are in good agreement with the corresponding conclusions reached by the other six Asian countries.