The covered bond directive has a surprisingly large number of paragraphs that make normal people question the intent. A couple of people now have asked me what is meant in article 10 ‘composition of the cover pool’. The answer is: whatever you want it to mean.
Article 10 – one of the shortest and vaguest in the entire directive - tells member states to devise some rules for banks issuing bonds with primary assets with different ‘structural features, lifetime or risk profiles’. Huh?
To understand this article you need to go back to the EBA’s kneejerk reaction expressed in 2014 that they really don’t like mixed asset pools. The separation of different assets classes into different cover pools was a ‘general principle of best practice’. As they wrote this they recognised that countries with a long tradition of mixed asset pools would object and the ‘general principle’ didn’t actually make it into the best practice that they ended up with.
Traditionally ‘mixed asset pools’ are a French thing – mortgages and public sector assets in the same pool. But it is more complex than that. Mixed assets could also mean a mixture of any of the broad asset definitions in the various paragraphs of article 129 of the CRR – in particular paragraph e and f – residential and commercial mortgages. Securitisation people are often amazed that a mixed commercial and residential pool is even allowed given how distinct the asset classes are in their market.
For some reason no-one cares about mixing the two paragraphs that deal with different types of public sector exposure. Me neither.
Going back to 2014 and the EBA. They suggested that the limits that some countries apply to certain assets (typically a cap on commercial mortgage exposure) “by construction impose consistency” – which is clearly just wrong. A 100% residential mortgage bond that suddenly switches to a capped 30% commercial exposure is hardly exhibiting a consistent risk profile.
In acknowledgement that the lobby for mixed asset cover pools would be too strong and that hard limits on pool composition were too difficult, the EBA delegated the task to national supervisors who were told to ensure that the “composition of the cover pool does not materially change for reasons other than amortisation”. Even that reference to amortisation is confusing given that the pools are revolving.
This best practice requirement on national supervisors is the seed of the article 10 that we have today. But looking at the wording again: ‘structural features, lifetime or risk profiles’ goes far beyond the distinction between commercial or residential mortgages. It could, for example, be interpreted as limits on geographical diversity, or currency exposure (in those countries that allow foreign currency mortgages), or even risk scores.
Or instead of limits, requirements that the composition remains constant – the EBA spoke about risk remaining constant “over the life of the bond”. But as there are very few programmes that have just the one bond outstanding at any time this effectively means over the life of the programme, which effectively means, for ever.
Investor protection by rules on cover pool composition sounds like a great idea but in the total absence of more specific guidelines it is really difficult to know how to interpret it in practice.