Analysis Of The Optimal Reinsurance Retention Models

The reason that the reinsurance market is called the insurance market "the safety valve", "the regulation", occupies the important status in a country financial system, because it during the safeguard insurance business quickly good development's, is playing the vital role to the national finance security. With China's participation in the process of gradually accelerating economic globalization, especially since the accession to WTO, the cancellation of the original statutory reinsurance and the recent "Chinese reinsurance market development planning," the enactment, the Chinese reinsurance market in the development speed and scale of the degree of marketization and competitive situation and other aspects of significant changes have occurred, is facing new opportunities and challenges.This study is based on portfolio theory and information entropy theory as the theoretical basis, the portfolio theory is the central issue of how to choose the investment decision-making risks and benefits.Make reinsurance as a kind of investment behavior, solve the minimal risk and expected return whole the biggest goal programming problems, and double selecting optimal reinsurance retention。Based on the information entropy to measure risk and risk premium for variance said that the main defects is actual income variance of an average revenue from the wave fluctuations, the actual income, said the uncertainty, regardless of the actual income is higher than average earnings or below the average income.While information and entropy is complementary relation between the entropy, and know the information is less, uncertainty, Conversely, the entropy, know more information, uncertainty and smaller. Establish mean-variance logarithm-entropy model can better for optimal reinsurance.Using information entropy of disproportional reinsurance to solve the optimal problem, this is the largest privately owned the innovation. Also put forward by mean-variance logarithm-entropy model separately predict the optimal proportion and scale reinsurance, compared to leave after a set of optimal reinsurance strategy。By application of statistical data model, an example is the article the following conclusions:Frist:Through the inspection, the mean-variance logarithm-entropy model than mean-variance model is more advantageous than the variance of information entropy, the more suitable for risk measurement. Specifically is lower yields and more in the lower risk.Second:Disproportional reinsurance and proportion reinsurance as two different types of reinsurance means, the reinsurance strategy is different. Which means reinsurance, considering proportion reinsurance and disproportional reinsurance optimal value, comparative analysis and realistic significance to realize the Pareto optimality.