Currency,Banking&Asset Risk Identification And Early Warning

In 2008 the whole world was swept by the financial crisis,the most serious disaster after the "Great Depression".According to estimation by IMF,Up to 1.4 trillion dollars was cost directly from financial institution by the 2008 financial crisis,according to IMF's estimation.Sovereign debt crisis,a continuation and deterioration of the 2008 financial crisis,erupted in the European Union leading to an economic depression in Europe.In 2015,China,the second largest economy,step into a "new normal" pattern which mean that its slow growth rate was undeniable to other countries.With the development of economic integration,it's extremely unlikely for each country to avoid impact from the outside world and once a crisis breaks,a domino effect must be explosive.A crisis leads to a huge impact for each country and costs them a long time to revive,which makes it necessary to perfect the identification and early warning system for financial crisis and to identify and alert the forthcoming crisis in time in order to take a damage-control action in advance and stop its massive outbreaks.Given the statistic scope and data availability,study on index system cannot fully apply to China's national condition.With the progressive development and prosperity of our financial market,existing financial risk identification and early warning system needs perfection to meet the demands of times.Above all,there is a significant and profound point to construct a new system.First of all,this paper elaborates classic models on financial risk identification and its early warning,parses the utility and flaws of FR,KLR,STV and some other innovative models,applies MS-VAR model to describe financial risk level after some comparison and analysis.This nonlinear model can classify data set,overcome information loss when converting continuous data into discrete ones,has no need to determine crisis period beforehand.Next,we construct EMP?BFI?FBP,three financial market segments' FSI,based on previous research.While constructing FSI,we replace GDP data to avoid information loss from frequency conversion,substitute real exchange rate for that against US dollar,use month-over-month data,which makes financial risk immediately shown in FSI.Then,we build financial risk identification model based on market segments and finally determine MSIH-VAR model that outstandingly captures reality.But first we screen the data set originally from many previous researches to construct early warning system and then decide the data used in this model by Granger test.After that empirical analysis was given based on the satisfying results from this model,which test the outstanding capacity to capture financial risk and gives us key indicators resulting risk change from three financial market segments.Finally,we apply ARIMA model to predict out-of-sample values and after getting the new sample,we estimate the model again and test its out-of-sample predicament.Except the banking market segment,the predicaments underestimate real ones in the other two market segments.The conclusions are as follows:First,apart from EMP,BFI,FBP,M2 multiplier and real exchange rate overestimation are picked into the currency crisis model,real exchange rate overestimation and change rate of financial institutions' deposit are picked into the banking crisis model,change rate of both stock index and house price are picked into the asset bubble crisis model.Second,in currency crisis model,sustained growth of foreign exchange reserves is the root of risk,violate change of M2 multiplier directly and seriously affects the risk?Risk falls down when deposit-reserve ratio goes up and vice versa.In banking crisis model,loan-to-debt ratio and growth rate of credit effectively shows condition of our bank risk which is in sync with the booming in real-estate by lending.In asset bubble crisis model,sharp fluctuation in both stock index and house prices mainly contributes the risk.Third,MSIH-VAR model has a strong capacity on financial risk identification,but underestimates real ones on early warning,well indicates big events in financial market from 2000 to 2015,and reveals key factors that influence the risk of financial market segments.ARIMA model behaves not so well in prediction which is outstandingly shown in stock index and its ratio of change.In China,we have no significant economic crisis from 2000,however,a stable environment plays a key role in development of economic healthy.As a result,it brings us higher request to construct a financial risk identification and early warning system with Chinese characteristics.In the intricate world present,results we get offer a new reference for Chinese financial risk identification and early warning system.