The Study Of Early Warning In Spatiotemporal Endogenous Credit System

In the period of economic prosperity,the growth of credit will promote economic development.However,without the supervision and guarantee of the monetary authority,the growth of credit will lead to systemic financial risks and may even lead to a financial crisis.Secondly,when the increase of credit exceeds a ”critical point”,it may also lead to a financial crisis.In recent years,many scholars have found that the formation and development of financial crisis not only change with the change of time,but also the spatial pattern plays a crucial role in the origin and development of financial crisis.Therefore,this paper aims to study the use of early warning indicators to detect the warning signals issued by the system in the process of evolution in the spatiotemporal endogenous credit model.We use the methods of patch size distribution and various general leading indicators to detect the warning signals of regime shift in the spatiotemporal endogenous credit model.The main results obtained are as follows: with other parameters unchanged,by changing the control parameters,we find that the whole system can be changed from the state of high asset prices to the state of low asset prices through folding bifurcation.Firstly,under such conditions,for the spatiotemporal endogenous credit model of one dimension,we use spatial variance and spatial skewness to predict the above state transition.It is found that when the system moves from the place far away from the critical point to the place near the critical point,the spatial variance and spatial skewness will become steeper and steeper.In addition,it is found that the system evolution is almost unaffected by different diffused nuclei and near neighbor diffusion.Secondly,for the two-dimensional expansion of the spatiotemporal endogenous credit model,on the one hand,we used the inverse cumulative patch size distribution to predict the state transition.The results showed that the inverse cumulative patch size distribution would show a power law distribution only when the system evolved to the critical point,and the inverse cumulative patch size distribution would deviate from the power law distribution when the system was before and after the critical point.Moreover,we find that different spatial scales have little effect on the evolution of the system.On the another hand,the spatial warning index widely used in dynamic systems is used to predict the imminent critical transition.These indicators have successfully predicted critical shifts in asset price conditions from high to low.In the current situation of accelerating economic globalization,it is impossible that one of financial institutions can be independent when they face financial crisis.There are very complex connections among financial institutions.Therefore,we construct a spatiotemporal endogenous credit model,and the patch size distribution and general warning indicators have good predictive ability for the occurrence of financial crisis in this model.The results provide a certain reference for the early warning of financial crisis and will be helpful to further prevent the occurrence of financial crisis.Finally,a summary is given and we hope that our methods can promote the utilization and test of spatial warning signals in real financial systems.