Study On Quantification And Assessment Of Spatial Basis Risk In Weather Index Insurance

In practice,the traditional agricultural insurance based on "individual" has many problems,such as difficult surveies and loss determination,high costs of claim,serious moral hazard and adverse selection.Weather index insurance is not based on the actual losses,but based on the objective data observed by meteorological stations.It is especially suitable for China's small-scale and scattered plots.Although weather index insurance has many advantages,there is inevitably basis risk.In view of the fact that the implementation of weather index insurance in China is mainly spatial basis risk,this thesis mainly carries out the theoretical research on the definition,impact mechanism,quantification and assessment of spatial basis risk,and explores a quantitative method of spatial basis risk based on empirical analysis,looking forward to filling the gaps in domestic research on spatial basis risk of weather index insurance and providing reliable reference for spatial basis risk management in the future.In terms of theoretical research,first,this thesis clearly defines the concept of spatial basis risk and fully realizes that the spatial basis risk in weather index insurance is caused by the spatial heterogeneity of meteorological elements.Second,this thesis explains in detail the influence mechanism of spatial basis risk on weather index insurance,and introduces ambiguity aversion theory and compound lottery model to illustrate that the fuzziness of individuals' perception of spatial basis risk affects the promotion of weather index insurance.Then this thesis innovatively puts forward a quantitative method of spatial basis risk,referring to the existing researches on basis risk in futures market and regional yield insurance,and taking the difference between simulated meteorological yield loss in county-level regions and actual meteorological yield loss in towns as the spatial basis of weather index insurance,wherein the meteorological yield loss in county-level regions needs to be simulated from the relationship model between meteorological anomaly and yield loss.In the aspect of empirical research,the research object is the soybean in Jiaxiang country of Shandong province,through analyzing the production risk,historical disasters and soybean growth period data,9 weather indexes are established,and the optimal nonlinear regression model between relative meteorological yield loss and abnormal weather indexes is established.On the above relationship model,the difference between the simulated meteorological yield loss at the county level and the actual meteorological yield loss at the town level is calculated to obtain the spatial basis.And finally the probability density function and distribution function of the spatial basis risk are fitted.This thesis tries to accurately simulate the relationship between meteorology and yield from the aspects of ensuring key growth period,establishing comprehensive weather indexes,setting index thresholds,establishing non-linear regression model,multiple fitting optimization model and so on,so the meteorological yield loss at county level is close to the simulated loss of the model.The results of the model show that the influence of climate on yield is mainly manifested in the rainstorm disaster in the three leaves-branching period,flowering-pod setting period,drought disaster in the grain-bulging period and low temperature disaster in the mature period,while the high temperature index is not significant.The quantitative results show that the spatial basis of towns in Jiaxiang ranges from 0 to 0.57.Historically,the risk of spatial basis is relatively low.In most years,the spatial basis is below 0.30,while in a few years,the basis exceeds 0.30.The probability density function of spatial basis risk is Weibull distribution,Wakeby distribution and Gen.Extreme Value distribution,of which Wakeby distribution is the most.From the distribution function,it can be estimated that the probability of the spatial basis less than 0.10 is 80%,and the probability greater than 0.30 is about 10%.The quantitative results have certain practical significance.First,the feasibility of weather index insurance can be checked in advance.If the spatial basis is generally small in history,it indicates that the regional systematic risk is strong and the heterogeneity risk faced by individuals is small,so it is suitable to carry out weather index insurance.The second is to manage the spatial basis risk in weather index insurance afterwards.If the spatial basis of a town is too large and exceeds a certain range,it needs to be subsidized obliquely.For Jiaxiang,if the spatial basis is less than 0.3,it can be considered as a conventional spatial basis risk.For these cases it can be ignored in actual management,while the towns with spatial basis exceeding 0.30 need to be compensated emphatically.