The Research And Application Of Time Series Analysis In The Economic Investment

The foreign direct investment (FDI) has gradually become the main form ofinternational capital flow with the increasing FDI activities since1980s. Under thisimpetus, the economic technology of less developed countries has been greatly enhanced.The foreign direct investment activities played a vital role in the development of theglobalization and the execution of China’s reform and opening-up strategy, China hasbecome a large country which makes appropriate use of FDI. Therefore, it is of muchsignificance that we can make a reasonable and effective decision based on the study of theforeign direct investment in the new period. Time series analysis is an important branch ofthe mathematical statistics and it is widely used in the economic field. To provide thegovernment and the investors with reliable reference for developing a reasonable policy,time series analysis can establish an accurate model and prediction for China’s foreigndirect investment. In this thesis, the author will establish a time series model on China’sForeign Direct Investment and predict the trend of FDI in the next several years. The maincontent includes:1. It studies the modeling method of China’s foreign direct investment in detail. Wecan obtain the dynamic structure of China’s FDI trends well by the analysis of actual data.According to one of the characteristics that the financial time series data of FDI issusceptible to noise interference, it studies the wavelet analysis as well as the impact ofelimination of noise interference on predicting. The simulating experiment of China’sforeign direct investment verifies that the improved algorithm is superior to the basic timeseries analysis method in calculation accuracy by MATLAB and Eviews6.0.2. Modeling for China’s foreign direct investment, it integrates the advantages oflinear and nonlinear time series model which can overcome the drawback of onlydescription of self-correlation and ignorance of heteroskedasticity during previous researchfor China’s foreign direct investment. This model can explain the dynamic structure ofChina’s FDI trends more scientifically and thoroughly. 3. The combination model of the intervention analysis and time series analysis hasbeen studied in this thesis to make the model more practical. In actual life, China’s foreigndirect investment and other examples could be often influenced by the emergencies whichwill exert great effects on time series analysis. Consequently, it is necessary to remove theinfluence of intervention element of these data. This thesis mainly aims at establishing theintervention model against our country foreign direct investment data intervention effectcaused by economic crisis as well as time series model on the data following elimination ofthe interference by Eviews6.0. Based on the comparison of the predicted results, it verifiesthat the intervention analysis can reduce the prediction errors and improve the accuracy ofpredictions of our country foreign direct investment.