SAP 金融风险管理模块中Credit and Debit Value Adjustments的计算方法

SAP TRM（TReasury）中，提供了CVA/DVA（Credit and Debit Value Adjustments）的支持；所谓CVA/DVA是对金融产品记账时候，对于考虑到客户的信用对借出金融产品，比如investment进行价值调整（Credit Value Adjustments）；同时对于借入的金融产品进行价值调整，比如borrowing，这个就是(Debit Value Adjustments), 有意思的是，存在因为自身风险增大，而造成借入减计，产生profit情况。现在的会计准则，这些都会计入其他综合受益（OCI)。

 

 

SAP提供两种计算CVA/DVA调整实现方法，

一个是通过 Credit Spread Curve和Yield Curve组合成综合composite Curve，来进行金融资产with risk 的带风险的NPV计算。这里Credit Spread就是客户或者自己的PD*LPG(损失率*损失发生几率）

还有一个就是基于可预见的风险资产，和PD*LPG乘积，算出可能损失金额，和没有考虑风险的RISK Free NPV进行相减，得出带风险的NPV。

下面的是内部分析的英文

Since the financial crisis, counterparty risk has become more prominent. Notably, the IFRS13 accounting standard requires companies to value OTC derivatives including credit value adjustments (CVA) and debit value adjustments (DVA).

•The CVA accounts for the possibility of a default by the counterparty : It is subtracted from the risk-free fair value

•The DVA accounts for the possibility of a default by your own company: Its absolute value is added to the risk-free fair value. (With the convention that the DVA is negative, this corresponds to a subtraction of the DVA.)

 

The risk-free fair value and the CVA/DVA contribute to the fair value: NPV = Risk-Free NPV – CVA – DVA

 

Calculation Methods for Credit and Debit Value Adjustments

1. Difference Method (Method 1)(1708           delivery already)
                       With this method,      the system first calculates the (risk-based) NPV      using      the yield curve stored in the evaluation type. If you have also made           the settings for credit spreads, the system also      takes credit      spreads into account when calculating the NPV (in a composite           yield curve).
                       In this way, the      system calculates the risk-free NPV with the      risk-free      yield curve stored in the evaluation type.
                       The CVA or DVA is the difference      between the risk-free NPV and the      (risk-based)      NPV:
                       Risk-free NPV - NPV = CVA or DVA

1. Based on Expected Exposures (Method           2, 1711)
                       With this method,      the system calculates the risk-free NPV on the      basis      of the risk-free yield curve that is assigned in the relevant           evaluation type. The system      calculates the CVA/DVA      separately. First of all, the      expected exposures for a set of future dates      are      calculated (or entered manually), aggregated, and weighted with the           product of default probability (AWKT or PD) and loss given      default (LGD).
2.  Constant Exposure Approach ,            with which the risk-free NPV is calculated for the      evaluation date and       then for which it is assumed that      it remains constant over time (->       always the same      expected exposure for all future dates) .
3. Variable Exposure Approach , with            which the NPV is calculated for all future dates      (changing horizon) with       a constant evaluation date.      In this case, the expected exposures can vary       over      time.

 

• 

          

Where

• LGD = loss given default
• D = discount           factor
• t = time
• AWKT=      PD  = default probability
• C in           subscript = business partner
• I in           subscript = your own company
• EPE =           expected positive exposure, evaluation time risk-free NPV
• ENE=           expected negative exposure, evaluation time risk-free NPV
• EPE and CVA           are positive, Incoming payment in future
• ENE and DVA           are negative, outgoing payment in future

系统配置并不是很多，简单示意如下

 

• Gross Relative Fair Value Approach
            CVA and DVA are distributed across all transactions of the netting      group in proportion to the NPV (fair value) of the transactions.
• Net Relative Fair Value Approach
            The system proportionally distributes CVA across the transactions      with a positive NPV (fair value), and DVA across those with a negative NPV      (fair value).

来自 “ ITPUB博客 ” ，链接：http://blog.itpub.net/22259926/viewspace-2564923/，如需转载，请注明出处，否则将追究法律责任。

转载于:http://blog.itpub.net/22259926/viewspace-2564923/