Real Estate Corporate Credit Risk Assessment Study

Recent years, as estate industry is becoming warmer and warmer,many estate companies loan from the commercial bank or gaindeveloping capital from financial and other non-financial institutions.This paper researches the credit risk of the estate company from the angleof the institution issuing loan like commercial bank.This paper is divided into six chapters. The first chapter introducesthe background, researching methods and thinking path, researchingstatus quo; the second chapter, correlative theoretical overview, expatiatesthe principle and steps of the mathematic methods used in this paper, thethird chapter is the construction of the evaluation index system; the fourthchapter sets up Logit model in the premise of assuming that financinginformation can reflect the credit of a company, it passed test ofmulti-statistic index and reached the effect that expected rate exceeds85%.;in order to show the scientific and applicable character of the indexused in the foregoing text, the fifth chapter tests a typical company, andcompares the characteristic and applicable environment between Logisticregression method and fuzzy evaluating method;the sixth chapter is theconclusion and a vista.To the non-financing qualification index, this paper adopts signatureanalyzing method and determines proportion by experts' marking andalternative optimum method. Considering the different discriminatorypower, this paper cuts the weaker index off by single factor varianceanalysis when constructs Logit model in order to increase discriminatorypower and save the quantity of the work. Considering the characteristic ofvariety index, this paper deals with the positive, negative and moderateindex differently and standardizes the index. In order to avoidmulticollinearity, this paper uses principal component analysis to processbeforehand. At last, this paper construct regression model using Logisticregression method and test whether the model is right by using wald test,-2LL test, Cox&SnellR~2 and Nagelderke R~2 statistic tests and test theapplicability of the whole index system by using multi-level fuzzyevaluation. The test method can solve the problem of various index, difficult to determine the relative importance and that evaluators faceuncertainty and fuzzy, as well as construct the various complicatedfactors to multi-level index system. Based on it, this paper determines theproportion among the index and makes fuzzy factors quantifying.