Research On Pricing Of Dual-factor Weather Derivatives Based On ARIMA-LSTM Combination Mode

In recent years,abnormal weather events occur frequently in China,and due to the obvious differences in climate and industrial structure between the east and west of China and between the north and south,China is facing an increasingly more serious weather risk threat,which also puts forward higher requirements for China’s weather risk management ability.As an important tool of weather risk management,weather futures,options,swap contracts and other weather derivatives,in mature markets abroad and related industry has been widely used,but domestic weather derivatives contract development is still in its infancy,still not in the domestic various regions,industry fully effective application in the weather risk management.Under this background,fully combined with our regional weather risk complex and regional climate differences,the actual situation of different industrial structure,study foreign mature market weather derivatives development experience,explore for domestic various regions and industries of different meteorological elements have wide applicability of weather risk management tools,has important practical significance,this is the overall national security concept and accurate grasp,help build a food security lines,and for energy,tourism and other weather sensitivity to guarantee the healthy development of the industry,boost high quality development in our country.This paper first introduces the weather risk management and its main management measures,the types of weather derivatives and the common main weather indexes,and discusses the differences between weather derivatives and weather insurance.Secondly,according to the existing research and application of weather derivatives contracts at home and abroad,combined with the regional differences of climate and economy in China,we try to preliminarily develop and design the dual factor of temperature and rainfall,and derive the calculation formula of weather composite index in this paper.Thirdly,based on the meteorological data of Nanjing and Urumqi,which have obvious differences in industrial structure and climate characteristics,this paper constructs a combined model of ARIMA time series and LSTM neural network,to simulate the average daily temperature and monthly rainfall data of the two places,and then obtains the comprehensive weather index valuation of the two places.On this basis,this paper uses the differential pricing method and Monte Carlo stochastic simulation method for Nanjing,Urumqi weather index futures,options,and comparison with the actual price,proves that the contract for the weather risk management in different regions.Finally,this paper prospects the research prospect of developing weather derivatives contracts with wide applicability in China,and provides relevant suggestions for the development of weather derivatives in China.