Hi all,

the summary of what has been achieved this week:

Item 1 (STD issues identified by Deloitte): 

The model calculations are fixed (Issues 1-9).

The results still do not match 100% because:

a)      Large number of exposures identified as undrawn(off-balance)  on Deloitte .xls calculator are mapped as on-balance (to N_RETAILBOOK.OUTSTANDINGBALANCE) in the capital manager database.  This results to overstating of RWA calculated to model as CCF=50% is not utilised for on-balance exposures.

The off-balance portions of exposures shall be mapped to N_RETAILBOOK. UNDRAWNBALANCE.

Examples of such exposures: N_RETAILBOOK.PRIMARYACCOUNTID in (63347105,63514509,63220010,63585309) – please compare the “07 Retail FSA004- my amendmentsv5 (version 1).xls” and what is in N_RETAILBOOK.OUTSTANDINGBALANCE and N_RETAILBOOK. UNDRAWNBALANCE for these exposures.

This is an ETL issue and must be fixed on Principality side. – please contact whoever is responsible for ETL.  

b)      In limited number of cases, the Deloitte results(“07 Retail FSA004- my amendmentsv5 (version 1).xls”) in column “Amended Total” are not correct. (in most cases where column “Difference” in the same .xls sheet  is >0). The fixed model calculations however match to results calculated in columns =SUM(AO540:AQ540). This has been discussed with Keith and he seemed to agree that model results were correct in these cases. Example of such exposures: PRIMARYACCOUNTID in (44181201,41671709)

Item 2 (IRB and STD discrepancies in aggregations):

The changes have been implemented to both writeback and reporting model.

The STD discrepancies have been fixed. (on-the-fly and written back results now matching)

Fix for LGD-measures implemented. (brought down the IRB RWA significantly)

Fix for IRB aggregation implemented. (on-the-fly and written back results now nearly matching –there is still difference of 0.2% - I will investigate this further)

Item 3 (Reconciling the written back and “on-the-fly” RWA calculations under the standardised approach):

The RWAs now match between Writeback and Reporting model (both On-the-fly and dtatamap results)

The Latest versions of the models with all teh changes implemented during this week are:

Writeback model: “Retail - B2CM Retail v16 (Build3) WB_JM15_full”

Reporting model: “Retail - B2CM Retail v16 (Build3)_REPJM_02_full”

(Could you please upload these model folders to our FTP site please)

Early next week-after I get the models from FTP site- I will provide detailed list of model changes implemented next week.

Please could you make sure the ETL issue is fixed, and then check the results coming out of the modified models.

As soon as you are able to re-test after the ETL is fixed, I might come back for a day or two if you identify some remaining issues or need the assistance with further testing.

Kind regards,

Jan

v 72ke je vlastne jedno a kedze z29ky trcia iba hlavne pokojne pouzi injekcne inly? potrebujes asi skor nieco rychle a presne nez sa pokusat vybrusit hlavne z plastu.

Vším co se podobá Browning M2 .50 Caliber Machine Gun.

zdravím  může mi prosím někdo poradit za jaké kulomety můžu změnit originalní na B-29 Superfortress 1/72 jestli bych třeba mohl použít  resinové  z b-25 nebo jiné děkuji