我国通货膨胀率的波动与组合预测研究

发布时间：2018-12-14 21:10

【摘要】：通货膨胀水平一直是社会关注的焦点,它既关系到普通民众的自身财富,又关乎社会经济发展与稳定。改革开放以来,我国通货膨胀率波动比较明显,大部分时期处于通胀和通缩交替发生阶段。2014年以来,随着我国经济进入新常态,全球经济增长乏力,国际大宗商品价格大幅下跌,我国和全球经济又面临陷入通货紧缩的风险。通货膨胀率作为政府宏观调控最为重要的参照指标之一,其未来走势将对我国宏观政策产生重要影响,因此,对通胀率的预测就显得尤为重要。我国最为常用的衡量通胀指标为居民消费价格指数,即CPI。本文在前人的基础上,通过组合预报方法对我国CPI的走势进行预测,以提高预测方法的准确性。本文首先对我国改革开放30多年来六次通胀周期进行分析,得出六次通胀周期的成因,根据我国通胀原因的不同,将通胀阶段分为加入WTO之前和之后两个时期,并得出新时期影响我国通胀的因素。其次,我们利用小波分解的方法对1990年以来的CPI序列进行分解,得到一层近似数据和五层细节数据,分别对六组数据进行时间序列建模并预测,重构得到小波分解的时间序列预测值。再次我们结合对我国通胀影响因素分析,选取具有先行效应的国际原油、铁矿石、大豆等大宗商品价格变动率,制造业采购经理人指数(PMI)、非制造业采购经理人指数等经济先行指数,M2增长率、工业增加值率、通胀预期等进行滞后项回归建立CPI多变量滞后回归预测模型,并用多元模型和小波模型结合组合预报得到较为精确的预测结果。通过本文研究,利用小波方法和多元回归方法所得到的预测结果均比较好,而考虑到短期物价波动更容易受到外部影响,本文比较建议采用多元方法并结合小波方法进行组合预报。
[Abstract]:Inflation has always been the focus of social concern, it is not only related to the wealth of ordinary people, but also related to social and economic development and stability. Since the reform and opening up, the fluctuation of inflation rate in China has been obvious, and in most of the period, inflation and deflation have occurred alternately. Since 2014, as China's economy has entered the new normal, the global economic growth has been sluggish. The international commodity price falls sharply, our country and the global economy again face the risk of falling into deflation. As one of the most important reference indexes of government macro-control, inflation rate in the future will have an important impact on China's macro-policy. Therefore, it is particularly important to predict the inflation rate. The most commonly used measure of inflation in China is the consumer price index (CPI.). In this paper, the trend of CPI in China is forecasted by combined forecasting method on the basis of predecessors, in order to improve the accuracy of forecasting method. In this paper, the author first analyzes the six inflation cycles in the past 30 years of reform and opening up in China, and concludes the causes of the six inflation cycles. According to the different causes of inflation in China, the inflation stage is divided into two periods: before and after China's entry into WTO. And draws the new time influence our country inflation factor. Secondly, we use wavelet decomposition to decompose the CPI series since 1990, get one layer approximate data and five layer detail data, and model and predict the six groups of data in time series respectively. The time series prediction value of wavelet decomposition is obtained by reconstruction. Thirdly, we choose the international crude oil, iron ore, soybean and other commodities price volatility rate, manufacturing purchasing managers' index (PMI),). Non-manufacturing purchasing managers' index, M2 growth rate, industrial value added rate, inflation expectation and so on are used for lag regression to establish CPI multivariable lag regression model. The multivariate model and wavelet model are combined with the combined prediction to obtain more accurate prediction results. Through the research of this paper, the prediction results obtained by wavelet method and multivariate regression method are good, but considering that short-term price fluctuation is more susceptible to external influence, In this paper, multivariate method combined with wavelet method is suggested for combination prediction.
【学位授予单位】：首都经济贸易大学
【学位级别】：硕士
【学位授予年份】：2016
【分类号】：F822.5

【相似文献】