Model Construction And Early-warning Monitoring Research Of Economic Operation Index

In the modern market economy,any country or region’s economic development will show cyclical fluctuations,and China’s economy is also the same.Since the reform and opening up,economy also faces with repeated expansion and contraction,the economic cycle fluctuations will inevitably affect the behavior of economic agents.Therefore,it’s necessary to monitor economy and implement early-warning work.Based on the economic cycle theory and specific economic phenomena of the economic cycle fluctuation,this paper carries out the model construction and early-warning monitoring of the economic operation index.Economic include both macro and micro aspects,specifically relates to many fields of social economic,including labor market,commodity market,financial and securities market,industrial enterprises,emergencies management,etc.The research on modeling and forecasting of different economic fields is of great significance to establishment China’s economic operation monitoring and early warning system.Firstly,establishing a reasonable economic operation index model and correctly evaluating the economics’current state can provide necessary information for government to make control measures and relevant policies.Secondly,based on the prediction results,we can prospect for the economic situation in the future.This paper carries out the model analysis and forecast of economic operation index based on the prosperity index method.Introduces the economic boom and periodic fluctuation theory,traced back to the history of the development of domestic and international about economic prosperity index,introduce classification method of leading,coincident and lagging indicators in traditional economic analysis,discusses the compiling principle of diffusion index,composite index and early-warning signal diagram.Due to the economic operation of time series data tend to have the characteristics of seasonal and traditional economic analysis does not have predictive warning function,in this paper,the X-12-ARIMA seasonal adjustment method and adaptive ARIMA model is introduced into the prosperity index research,and improve the traditional prosperity analysis,results show that:（1）the data through X-12-ARIMA seasonal adjustment eliminates the influence of seasonal factors,more in line with the actual situation;（2）based on adaptive ARIMA method can automatically select the best fitting model for economic operation index,and do prediction based on the fitting model;（3）keeping the original model and the early warning signal invariant,add the prediction results into early-warning signal diagram,so that the improved early-warning signal diagram has the forecasting function.Process control technology is widely used in the production process,and here apply it into economic operation index analysis.Introduced the basic principle of traditional control chart,study the influence of self-related or outliers existence on traditional control charts;through Monte Carlo simulation analyzed the impact of autocorrelation on conventional control charts’ARL,compared two autocorrelation control chart（residual control chart and the improved EWMA control chart）.Research on the effect of outlier on the normal control chart and traditional MEWMA control chart,reviews the common single variables robust estimation method,the multi-variables robust estimation methods and robust outlier detection method;added the robust estimation method FAST-MCD into the MEWMA control chart,and the robust MEWMA control chart is constructed.Conduct autocorrelation process capability evaluation of economic operation index.Study the autocorrelation effect on process capability index C_p and C_pk;when autocorrelation exists in process,still using conventional methods to calculated the confidence interval of process capability index will produce what kinds of problem;and gives the confidence interval calculation method of these two autocorrelation process capability index.Last,propose two methods to construct economic operation index,that is,comprehensive evaluation index method and principal component composite index method.When time series data include outliers,apply robust principal component analysis method based on Fast-MCD algorithm to construct robust principal component economic operation index.In the whole study,the empirical research include seven aspects:employment prosperity index,consumer price index（CPI）,China economy coincident index,stock composite index,social total retail sales of consumer goods,bank macro safety index and macroeconomic operating index.Analyze the trend of each economic operation index,according to the actual situation of Chinese economy for verification and comparison;the constructed index can better reflect the actual situation of the current field;use the adaptive ARIMA method to realize optimal model fitting for economic operation index,on the basis of the reasonable model,predict index development in the future.