We strongly support the inclusion of this information for all banks. In our view, the disclosure of the split of the NPL stock in terms of past-due vintage is extremely important to reduce the current asymmetry of information. This information is essential for market participants in order to grasp the risk profile of the institution and thus strengthen market discipline. For investors, in particular, it is crucial for their assessments since time to recovery is a major factor, along with collateral, to estimate NPL value. Having access to this high quality data in advance will increase investor´s appetite and facilitate market transactions by improving planning and budget process.
In order to be effective, the first criterion to allocate exposures must be the past due vintage period. For instance, default or impaired exposures that are past due >=90 days must be reported in the relevant past due period bucket. Otherwise the information would not useless and non-comparable.
As stated in our response to question 2, the combination of time to recovery and value of the collateral are two main components that drive NPL price. For the assessment of the value of collateral, LTV ratio is the best risk indicator. However, unfortunately, recent history has proved that accuracy is directly related to recentness of appraisal. Therefore we recommend the inclusion of several buckets (e.g. 1 year, 1 to 2 years and more than 2 years) in combination with LTV buckets.
Concerning the question itself, we strongly support the inclusion of this information. Significant credit institutions with a high level of non-performing loans, for which template 7 is mandatory, must be in a position to offer as much information as possible to potential investors. Loan to value information, reported as granular as possible, is needed to assess recovery ratio. Restricting the NPL categories to solely “80% to 100%” and “over 100%” would not be sufficient to assess collateral value and perform preliminary house price sensibility analysis. Thus the differentiation of the share of loans closer to the 80% LTV mark (i.e. 60% to 80%) from the less risky ones is essential for a proper price estimation from investors and a more accurate risk profile assessment by the rest of market participants.
We welcome the EBA proposal which is a major step forward in terms of asset quality disclosure and comparability among banks and countries. We basically agree with the overall content and structure of the templates. Moreover the level of detail of the accompanying guidelines is sufficient and the only necessary addition would be the one stated in the second paragraph of our response to Question 2.
However there is a specific small addition that, in our opinion, would significantly improve the usefulness of the credit quality information of non-performing exposures (template 3): inclusion of general memo information of the share of loans and advances secured by immovable property. This would imply the inclusion of a single row between row 1 and 2 of the template.
The split between secured and unsecured (especially real estate collateralised) is fairly important for the assessment of the risk of the portfolio due to the very different LGDs and market price of the assets. Although these data (in a more detailed fashion including LTV buckets) is to be provided by significant banks in Template 7, given the degree of specialisation from investors, we believe that it should be mandatory for all banks. The burden for institutions will be limited since this information is indispensable for NPE management and thus easily accessible.