Applying DEA Model To Internal Credit Rating System Of ACLEDA Bank PLC.(Cambodia)

This study evaluating the current internal credit rating system of ACLEDA Bankused scoring method and examining the credit rating systems based on Data EnvelopmentAnalysis (DEA) to suggest the supplement, complement or alternative credit rating systemfor ACLEDA Bank to help it better manage its credit risk.The sample we used in this study contains80firms from different sectors. We findthat all of sample firms are correctly predicted as non-default firms by the current creditrating. Although the existing models are useful for credit rating, new models have toemerge in response to ever-changing business practices. The current rating model is basedmostly on credit office since it need to assign weight to criteria, so that the result may varyfrom one credit officer to another credit officer, thus it maybe not meet the internal-basedapproach under Basel II in term of Objectivity-the rating result should be reproducible byvarious subjects i.e. with the same information different analysts should come to as similara rating result as possible. It might be risky to the bank that everything depends on onecredit officer. This study therefore aims at introducing the application of data envelopmentanalysis as the supplement, complement or alternative credit-scoring model. Unliketraditional credit-scoring building on a formula where weights to a set of criteria areassigned subjectively, DEA will automatically generate the relative weights for analysis.However, incorporating DEA demands additional considerations, which are discussed inthis study. As opposed to well-known methods such as multiple discriminate analysis,logistic regression analysis, and neural networks, which require ex-ante information ofgood/bad classification, this approach only needs ex-post information of the observed setof input and output data of the objects of interest (client firms) to calculate their respectivecredit scores. The credit ratings are derived through a customized credit rating model thatconsiders both quantitative and qualitative components. The credit ratings score areranging from1to5. A credit rating score of1means excellent credit quality, a creditrating score of5means very poor credit quality where the corporate entity is highly likelyto default. Investment grade corporate receive credit ratings score between in the range of0to2.36. The credit managers assigned the internal credit ratings to each corporate andthe internal credit ratings represent the bank’s view of the corporate credit quality. This paper also obtained the available financial information for each corporate containingBalance sheet, Income statement and the Ratios. The full sample results show significantimplements in real output potential of useful DEA model. It is found that the efficiencywas more or less evenly distributed with a decreasing trend as the input or outputdecreased, but this was not the case for the rating. We found that most of the firms areefficient except4firms that not efficient but get the efficiency scores very near to1showing that it is similar to the current ratings giving by ACLEDA that most firms arerated high quality and good quality. In summary, simple efficiency and credit rating havesome elements in common and DEA can play an important role in credit assessment as asupplementary tool to credit rating models to help meet the Basel II requirements.