A Study On Correlation Between Web-query-based-CCI And Macro Economics

Researches on web search queries which benefit from the penetration and prosperity of the Internet,is a data mining technology along with the development of big data.Data from the web search queries was first applied on epidemiological study,then,many scholars applied this technology to forecast unemployment rate,movie box office,merchandise sales and many other aspects.These all fully proved that there is a correlation between the web search queries and economic behaviors.Regarding the web search queries as the reflection of consumer demand on the Internet,we construct Web-query-based Consumer Confidence Index,explore the connection between macroeconomics and the index and try to demonstrate the advantages like timely and high accuracy of prediction of the WCCI.Considering all the reasons above,WCCI can be the supplement of economic climate indexes.Our study mainly concludes following aspects:Firstly,we construct Web-query-based Consumer Confidence Index.The Consumer Confidence Index is a leading indicator to predict the future economic trends of consumption.In China,CCI is monitored and published by the National Bureau of Statistics,and there is a one month and a half lag of the publishing.The internet users whose demand and willingness can be reflected through web search data are producers and consumers who constitute the market.The changes of their focus can represent the quantitative fluctuation and price changes of commodities in the market,which is reflected by the differences of the searching and browsing activities on the Internet.Therefore,there is connection between web search queries and economic activities of market participants.The Web-query-based CCI we construct is a real-time changing variable that can cast the external impacts on the web-based consumer confidence,according to which we can apply it on some timeliness issues.After illustrating the compilation of the existing consumer confidence index in China,we discuss the possibility that whether web search queries can be the supplement of CCI or not.The index we built concludes the categories of commodities which consumers are more likely to get their information on the Internet,the data source is extensive and the acquisition of indicators is timing.In general we demonstrate that the reflection of different commodities' categories on the Internet differ from each other.And their cycles and fluctuations are also different.The process of constructing WCCI illustrates the correlation among web search queries,consumer confidence and consumer demand.Secondly,it is of great significance of monetary policy,fiscal policy and other macroeconomic policies to macroeconomic stability and development for a country.Both government and scholars all concern about the effect of macroeconomic policies and the factors influence them.Generally speaking,monetary policy aims at inflation and price stability while fiscal policy addresses depression and demand issues.In the process of macroeconomic policies transmitting to microeconomic individuals,confidence is a key point.It will influence the effect of macroeconomic policies whether the confidence is high or not.Concerning the “New Normal” condition of China's economic development,we may provide a new perspective of policy making and policy implement through web search queries.We apply MSVAR models on studying the relationship between WCCI and macroeconomic policies and draw following conclusions:There is significant regime switching pattern in the dynamic transformation of web-based consumer confidence and macroeconomic policy series.Web-based consumer confidence in China can be divided into high confidence regime and low confidence regime,each of the regime is in accordance with the policy regimes of loose and tight.The connection among web-based consumer confidence and different macroeconomic policies vary owing to the characters of different policies.The impact of WCCI on macroeconomic policies follows a pattern that the impact is heavy if the confidence is full and when the confidence is weak the impact will be light.The relevance between macroeconomic policies and web-based consumer confidence is demonstrated.Thirdly,price fluctuation,as an important part of macroeconomic fluctuations,is related to daily life of everyone.And WCCI is a kind of online reflection of consumer demand which may have interaction with price fluctuation.In this part of our study,we regard CPI as the representation of price fluctuation and weigh web-based consumer confidence through WCCI,trying to recognize the endogenous transformation mechanism in the series we study.The results can vividly show that there are significant structural mutations in both series,and the volatility of WCCI is stronger than that of CPI.In the long run,there is a higher possibility for CPI to grow in a relatively low speed than that for WCCI.As for WCCI,there may be a positive correlation between the volatility and the growth,which means the higher the growth rate of WCCI,the stronger the volatility may be.The results of MSAR model is more accurate than that of ordinary linear models,which indicates the tight connection between price fluctuation and web-based consumer confidence.The identification process of the two time series endogenous transfermation mechanism provides a new perspective for us to demonstrate the interaction between network consumer confidence and price fluctuation.The analysis facilitates us to predict the economic inflexion and to conduct inflation.Fourthly,economic growth is also an important part of macroeconomic fluctuations,which consists of investment,consumption and net exports.The DCC-GARCH models and SVAR models facility us to explore the relationship between web-based consumer confidence and economic growth which is represented by industry added value(IP).The dynamic characteristics of the two series are measured through ADF test,the Granger causality test and the cointegration test.According to the results,WCCI can Granger cause the change of IP.The significant dynamic correlation which fluctuates heavily is suggested by the DCC-GARCH models.It is also can be shown that WCCI can illustrate the variation of IP about 20%.Web search query indicates the change of consumer confidence,the search for specific keywords can be seen as a reflection of consumer demand.The information that consumers acquire on the Internet will influence their consumer demand and purchase decisions,which reflects the aggregate demand at the macro level.In this way,the fluctuation of focus may have an impact on aggregate supply,then the entire macro-economic will fluctuate.In terms of supply side,economic growth influences production activities which transforms the deposable income of consumers,as a result,the expectance of consumer will change.Web-based consumer confidence and economic growth interact with each other actively.In the perspective of household disposable income we analysis the dynamic correlation coefficient and the impact on economic growth from web-based consumer confidence.We suppose that the coefficient is affected by macro-economic climate.If the economy is in prosperity,the "squeeze effect" from increasing consumption to invest is relatively light or offset,the growth of economy is high while in a depression,the "squeeze effect" is obvious and the growth of economy is low.This hypothesis can be demonstrated by the comparison graph of the dynamic correlation coefficient and the macroscopic warning index series.Last but not least,it is of great importance to analyze factors that are relevant to economic growth.If we grasp the rule of relevant factors changes,we can have a better understanding of economic development trends,formulate macro control policies,and allocate market resources rationally.GDP,as an indicator of reflecting macroeconomics,have scholar significance to measure economic growth.GDP growth,the cumulative value and other data relevant to GDP are published quarterly by China's National Bureau of Statistics.Usually there is a lag of a half months or so.Commpared with other economic indicators,frequency and lag differences may lead to the misunderstanding for individuals to judge the condition.When we do researches with traditional linear models,frequency of the series in the models should be the same.Series of different frequencies should be converted into the same.In the procedure of transformation there will be missing information.It is under heat discussion that how to apply different frequencies series in a model while keeping most effective information.It is rational to apply high frequency data on the study of correlation between web search queries and macro-economic variables.We use the MIDAS model to study the relationship between the web-based consumer confidence and economic growth.A VAR model containing lag variable of GDP is built up as a base line model,then we construct two different polynomial MIDAS models.After comparing the results of the three models we can draw such conclusions: the performance of the MIDAS models to fit and forecast GDP are better than that of base line if the lag variable of GDP is added.The advantages of MIDAS models are obviously shown in demonstrating the correlation between web-based consumer confidence and economic growth.The problem of information lost during frequency converting is solved via MIDAS models.Most information from data is maintained in this way,which suggests the important role WCCI plays in forecasting economic growth.According to the relevant theories,we construct the WCCI using web search queries with the help of the web search queris.Then we apply this index on studying the correlation among web-based consumer confidence,price fluctuation,macroeconomic policies and economic growth.The empirical results of all models show that WCCI performs very well in reflecting the consumer confidence in China,a close relation exists among the index and the macroeconomic variables.Meanwhile,as a representation of confidence,WCC also illustrate the significant role that confidence plays in various economic activities,which can be a supplement of a macroeconomic indicator to provide new approach to judge the economic condition and to predict future.