Evaluating Corporate Loans via a Fuzzy MLMCDM Approach

摘要: A defuzzification based fuzzy MLMCDM model is proposed for the evaluation of commercial loans, where importance weights criteria in structure and ratings alternatives versus subjective are assessed linguistic values represented by trapezoidal numbers. These numbers defuzzified through ranking approach center area (COA) before applying to order avoid problem multiplying more than two The COA a number has following three situations: • If abg > bdhg. lies between b, i.e., left side core(B). achg < cdh. c d, right Situation other (a) (b). Obviously averaged aggregated going from lowest-level parent finally can be obtained. advantage considering both qualitative quantitative criteria. Moreover, it also allows deal with hierarchical may have sub-criteria these sub-sub so on. numerical example demonstrated feasibility model. Furthermore, Monte Carlo simulation conducted gain insight about behavior sensitivity analysis shows that four most sensitive ROE (f1121=11.3%), collateral (f134=11.2%), competitiveness (f131=7.1%) management’s experience (f121=7.1%). an eight-level risk or bucketing system suggested using 12.5, 25, 34.5, 50, 62.5, 75, 87.5 100 percentiles distribution. As result eight buckets defined facilitate interpretation scores final classified into buckets. decision on whether issue loan not then made index potential borrower category which falls. Other corporate loans evaluation, applied management problems. Yet some limitations should taken consideration real application as basis further research. score dependent being considered bigger larger scores. Different approaches lead slightly different results. definitions well corresponding different, results different. Further research try demonstrate case study, company situations. consistency reliability contrasted historical information. computerized code could available improvements adaptations.