China's Foreign Direct Investment Risk Assessment Index System And Model Construction

Even though China has put forward the "Going Out" strategy back in the 1990s,the actual development of China's outward foreign direct investment has taken off since its accession to the WTO.In particular,after the economic crisis in 2008,China's outward FDI has seen a growth spurt.New measures and patterns of foreign trade and economic cooperation,represented by the "One Belt and One Road" Initiative and international capacity cooperation,mark the beginning of a new era in China's foreign trade,economic and investment cooperation.The new features have raised new requirements to institutional arrangement for investment and system building for risk prevention.Chinese and foreign scholars have conducted various researches on the risk assessment of foreign direct investment,and have achieved numerous outcomes,in terms of key risk categories,risk determinants,risk assessment methods,and comparisons among different economies.Traditional risk assessment methods are mostly limited to the static presentation of one or several statistical methods and their calculation results,with which real-time tracking of data updates and hence timely assessment results from a variety of model fittings will be hardly achieved.That is why many appraisal agencies publish mostly annual reports,but hardly quarterly or monthly assessments.Outward investment activities are behaviors involving comprehensive decision-making and require constant follow-up of the latest risk changes.When risk assessment time spans are up to one month,six months or even a year ago,the corresponding results could be less indicative due to some fundamental changes of circumstances.The advent of the big data era and the evolution of artificial intelligence have made it possible to improve and break in FDI risk assessment.Therefore,this dissertation attempts to perform a dynamic risk assessment of foreign direct investment,through innovative composition of statistical models and analytical tools.This dissertation will conduct structural,theoretical and innovative explorations in the following aspects:1.Based on a literature review on national risks,political risks,economic risks,social risks,etc.,to find out proven and commonly used risk assessment factors,and to select and apply proper factors with the modeling method in this dissertation;2.In terms of modeling selection,innovation,and application,an innovative composition will be carried out,i.e.to conduct a comprehensive analysis through PLS-PM model,Decision Tree Model,Cluster Analysis,etc.for the selection of risk assessment factors;to draw the final assessment results,based on selected factors and weight,using artificial neural network(ANN)composed of Random Forest Regressor,SVR and Bayesian Ridge.3.Using the assessment results with traditional risk analysis method by other research institutions,as a reference for comparison,this dissertationalso conducts a horizontal comparison between different results drawn,respectively,through pure artificial neural network,BP neural network,PLS-PM and composite artificial intelligence model,and identifies an innovative model and research method,based on an overview of the pros and cons of different models and algorithms.4.In consideration of risk factors,in addition to macro factors,this dissertation also attempts to explore ways regarding industrial factors at meso-level and corporate factors at micro-level.5.Regarding China's overall economic,trade and investment background,and the operational features of China's political and economic systems,this dissertationproposes some basic viewpoints to the establishment of an FDI risk prevention mechanism.In this dissertation,the author performs explorations and attempts in assessment tools and innovative methods,to provide an adequate theoretical and methodological support for further innovations and progress.