Skip to main content
Response to consultation on RTS on the calculation of Kirb in accordance with the purchased receivables approach
Go backQuestion 1: Do you agree with the requirement that a rating system shall be exclusively used for securitised exposures that the institution does not service, i.e. for the exposures that are in the scope of these draft RTS?
On balance, we agree this it is reasonable given the need to take explicit account of underwriting standards.Question 2: Should an exception be introduced for certain corporate exposures (e.g. large corporate exposures that the institution may rate using the corporate rating system it uses to rate corporate clients)? Should such exception be limited to the estimation of PD? If yes, what alternative would you propose for LGD estimation?
We do not regard the bank's underwriting standards to be a significant determinant of the default or loss in default behavior of large corporate borrowers. Hence, it does not appear important to preclude use of the rating system that the bank already has implemented for such borrowers.Question 3: Do you agree with the fact that, unlike traditional securitisations, synthetic securitisations cannot meet the general conditions set out in this article and in particular the requirements on indirect control and ownership of the securitised exposures by the institution calculating KIRB?
The issues of control and ownership are important for receivables purchased from non-financial counter-parties. We understand the need to reinterpret these rules to be consistent with the nature of securitisation exposures. But we do not see why they should preclude use of a PuRA model and hence the SEC-IRBA for synthetic deals.Question 4: Do you consider that a more detailed definition of proxy data is necessary? If yes, please provide a suitable definition.
We believe that the implications of the inversion of the usual data hierarchy should be more explicitly spelt out. This point is explained in our attached paper.Question 5: Do you consider that the provisions set out in the draft RTS are workable if applied to securitisations of non-performing exposures?
We think they are broadly workable.Question 6: Do you have any other comments on the draft RTS?
We have written a paper explaining how we believe the RTS should be extended to cover several issues relevant for empirical modellers implementing PuRA models.Name of organisation
Risk Control Limited
