1、Table of Contents,1Executive Summary2CRMS Project Review3CRMS Project Commercial Process Improvements4HAVICS System Overview 5HAVICS System Technical Components 5.1Rating Methodology Overview5.2Rating Decision Support Parameters5.3Pricing Methodology Overview5.4Pricing Decision Support Parameters5.5
2、Limits Methodology Overview5.6Limits Decision Support Parameters,PurposeThe agreed upon purpose of the limits model is to establish both a maximum prudent exposure limit and a desired exposure limit. Those two categories encompass a number of individual limits that are calculated within the model:Re
3、gulatory limits. These limits are simply the regulatory requirements generally based on bank capital and covering individual companies, groups, jumbo loans and Hanvit subsidiaries.Target Debt Limits. These are most likely applied only to individual companies. They are based on sliding rating-scale p
4、ercentages of company valuation.Total Exposure Limit. This limit incorporates information about equity and derivatives positions the Bank may have with the borrower.,Regulatory Limits MethodologyFSS regulations require lending to individual borrowers and to company groups not to exceed certain bank
5、capital amounts. All limits are based on outstanding amounts and current bank capital.LimitIndividual Company20% of CapitalGroup25% of CapitalJumbo LoansSum of all Jumbo 5 x CapitalSingle Hanvit Subsidiary10% of CapitalAll Hanvit Subsidiaries20% of Capital,Target Debt Limits MethodologyThe Target De
6、bt Limit is designed to provide a quick lending limit benchmark based on a company valuation. The target limit is applied to all outstanding debt the company has to both Hanvit and other banks. The percentage scale at right displays the amount of company value by rating that sets the limit. These pe
7、rcentages were subjectively derived based on discussion with Bank members. Ongoing maintenance would require a review of the practical implementation of these percentages.,Total Exposure Limit MethodologyThe purpose of limits, according to the regulators, is to ensure that the bank does not have too
8、 much exposure to any one borrower so that it places the bank in jeopardy. This notion of maximum prudent exposure is represented in the regulatory and target debt limits. A complimentary idea is setting of a desired exposure for the Bank. This notion of limit setting incorporates methods of portfol
9、io management. In practice at other large international banks, the desired exposure is well below that of the maximum exposures as evidenced by experience with the KMV model. The largest exposures are usually within the regulatory or maximum exposures, but they do not meet the portfolio managers des
10、ires.The method assumes that a borrowers limit corresponds to the first point at which additional credit exposure would make more than a maximum allowed contribution to portfolio risk. This maximum, marginal contribution would be set by credit policy. The approach implies that the borrowers size, ri
11、sk rating, types of credit facilities, and correlation with the banks entire portfolio will affect the limit. Thus, limits will be lower for smaller, higher risk borrowers, who post little collateral, and are highly correlated with the bank.,Total Exposure Limit Methodology (continued)The methodolog
12、y uses a computationally feasible equation for approximating the credit-portfolio-risk contribution of a borrower. The formula below, representing the general form of risk contribution, accounts for derivatives and equity in addition to loans outstanding.,RC = EDF_WT x LIED_WT x CORR_WT x EXP + EQTY
13、_EXP x EQTY_CORRVariableDefinitionRCEstimated contribution to portfolio riskEDF_WTWeight proportional to the borrowers EDFLIED_WTWeight proportional to the loss in event of default typical of the borrowers facilitiesCORR_WTWeight proportional to the borrowers correlation with the portfolioEXPNominal
14、 exposure adjusting for loan equivalency (including mark-to-market derivatives) but not the above weightsEQTY_CORRWeight proportional to the borrowers equity correlation with the portfolio by company sizeEQTY_EXPNominal exposure mark-to-market equity value but not the above weights,Total Exposure Li
15、mit Methodology (continued)The marginal risk contribution is computed from the change in risk contribution with respect to indebtednessMRC = DRC/DDHere MRC denotes the marginal risk contribution, D change, and D total indebtedness of the borrower. A borrowers credit limit is determined by finding th
16、e point at which MRC reaches a ceiling of maximum marginal risk contribution (MMRC) set by policy.Limit = EXP at which DRC/DD= MMRC,Table of Contents,1Executive Summary2CRMS Project Review3CRMS Project Commercial Process Improvements4HAVICS System Overview 5HAVICS System Technical Components 5.1Rati
17、ng Methodology Overview5.2Rating Decision Support Parameters5.3Pricing Methodology Overview5.4Pricing Decision Support Parameters5.5Limits Methodology Overview5.6Limits Decision Support Parameters,Maximum Marginal Risk Contribution (MMRC)To make the total exposure limit approach operational, a thres
18、hold for the MRC has to be established. In a full, portfolio risk-return analysis, the threshold would conceivably be set depending on the return associated with a borrowers loans. Since this limit setting method does not incorporate pricing, a reasonable alternative would set the MMRC from regulato
19、ry limits. In this case, an extremely large borrower is selected with the highest risk grade in an industry with a low correlation with the banks overall portfolio. Using this method it is assumed that that borrowers entire indebtedness is with the bank. MRC value corresponding to the regulatory lim
20、it on credit exposure to a single borrower is then solved for. That MRC would represent the MMRC threshold to be used in determining limits for all borrowers.Several initial experiments with the various borrowers produced roughly an MMRC value of .0075. It is expected that this value will be refined
21、 as the total exposure limit becomes more fully accepted within the Bank. Ongoing maintenance of this factor will require both informal loan officer reviews and a more rigorous analysis using various large borrowers at the regulatory limit.,Concentration WeightsThe total exposure limit algorithm req
22、uires weights that are proportional to the borrowers correlation with the portfolio. The initial limits model will have three concentration levels, high, medium and low. These three concentrations are most likely sufficient coverage of correlation information. Any smaller segmentation of the concent
23、rations may place too great a burden on the limits model operator since an overly large amount of research should not be necessary to operate the model.The initial model weights were determined mostly subjectively with some support from trial and error iterations of the model. When a sufficient amou
24、nt of data has been gathered, it is recommended an actual correlation analysis be run on a representative portfolio to determine reasonable ranges for the concentration and weight parameters.,Equity Volatility FactorsIn order to place equity amounts held by the Bank on a more equivalent basis with l
25、oans and derivatives, a weight proportional to the borrowers equity correlation with the portfolio by company size is required input. Initial volatility assumptions were determined similarly to the way concentration weights were determined, mostly subjectively with some support from model tests, tho
26、ugh some knowledge of US equity markets helped bracket the magnitude. A study of correlation between equity volatility and company size should be carried out soon. Such a study, requiring historical equity data, would become the preferred method for maintaining and validating these factors.,pOXLp7v0
27、djZKylHSJr3WxBmHK6NJ2GhiBeFZ7R4I30kA1DkaGhn3XtKknBYCUDxqA7FHYi2CHhI92tgKQcWA3PtGZ7R4I30kA1DkaGhn3XtKknBYCUDxqA7FHYi2CHhI92tgKQcWA3PtGshLs50cLmTWN60eo8Wgqv7XAv2OHUm32WGeaUwYDIAWGMeR4I30kA1DkaGhn3XtKknBYCUDxqA7FHYi2CHhI92tgKQcWA3PtGZ7R4I30kA1DkaGtgKQcWA3PtGZ7R4I30kA1DkaGhn3XtKknBYCUDxqA7FHYi2CHhI92tgK
28、QcWA3PtGshLs50cLmTWN60eo8Wgqv7XAv2OHUm32WGeaUwYDIAWGMeR4I30kA1DkaGhn3XtKknBYCUDxqA7FHYi2CHhI92tgKQcWA3PtGZ7R4I30kA1DkaGhn3XtKknBYCUDxqA7FHYi2CHhI92tgKQcWA3PtGshLs50cLmTWN60eo8Wgqv7XAv2OHUm32WGeaUwYDIAWGMes02GshLs50cLmTWN60eo8Wgqv7XAv2OHUm32WGeaUwYDIAWGMes02dLPqafkFGlzcvv2YiRQYHbhR8AI1LKULh3xvjDzkEAMGr8xbwF1bH1oIM30E7xp,
