A couple of deals recently have been described as structured covered bonds – usually in a mildly pejorative way. The phrase ‘structured covered bond’ has come and gone over the last twenty years, what does it even mean?
The first time I heard the phrase was in 2003 when HBOS launched the first ever UK covered bond. As pioneers of securitisations, they had no problem self-identifying as a structured programme, and investors didn’t care less. But over time it has been used as, if not an insult, then at least a way to get a higher spread. Is that justified? I’ll discuss that in my next post. But first, let’s try to define what a structured covered bond actually is.
The obvious, but I think wrong definition, is a covered bond created entirely according to contract law with no ‘Covered bond law’, passed by parliament to define it. A covered bond law does three main things: it tells you what a covered bond should look like, it empowers you to create a bond that looks like that, and it grants supervisory powers over that bond to the competent authority. The latter is the only one that can’t be done under contract law, so should be the differentiation between structured and regular covered bonds. Let’s look at those in turn.
We know what a covered bond should look like without the law telling us. I’d like to say that the covered bond directive has defined the market’s perimeter pretty well, but the reality is that this was already done. Investors are pretty fickle, and their rules are rarely set in stone, so they don’t provide the best of definitions. But when the index providers make a ruling on eligibility and when the ECB makes a judgement on repo liquidity categorisation – both of which have happened in the past in the case of ‘structured covered bonds’ – that’s a pretty good indication that it’s a covered bond for market purposes. There is a degree of self-fulfilling to this definition – if it is eligible for the index it’s a covered bond, just as much as the other way around.
Can we make a ‘real covered bond’ without a law? Obviously, yes. The legal structuring that goes on in the majority of countries (defined as those that don’t use common law or something broadly similar) are largely a way to replicate what common law countries can achieve without a law. It is no coincidence that the securitisation market grew and thrived without a law in countries like the UK and America. In the former basically identical technology was applied to create the first, and all subsequent covered bonds. Same in countries with related legal systems like Australia and Canada.
Don’t take my word for it. The acid test is whether the legal opinion is strong enough to persuade rating agencies to give AAA ratings on the basis of the legal technology deployed, in particular the segregation of the assets. If it is, you don’t need a law. And yes, I accept that not everyone will agree with the ‘if its good enough for the rating agency’ test. Rating agencies aren’t perfect, but they are better than almost anyone at looking at a legal opinion and judging whether it is sufficiently watertight for commercial considerations.
But looking at those markets which, at inception were regarded as structured and which now generally aren’t, the main reason is the third thing that we normally get from a law: a supervisory regime.
All banks are supervised. That goes with the territory. Some are better supervised than others – banking union (remember that? we spoke about it before we spoke about capital market union) should have addressed this, but let’s be realistic. But the supervision of a bank is not the same as the supervision of a covered bond programme ‘designed to protect the interests of covered bond holders’ (in the words of the UCITS directive). In fact, they might be at odds – it isn’t difficult to envisage a scenario when the interests of the covered bond holders and of the bank supervisors are at odds (‘put more assets in the pool!’, ‘no, we need those assets for emergency liquidity repos’). With the most accommodating regulator in the world – one who might agree voluntarily to undertake covered bond supervision in the absence of a mandate to do so - this potential conflict remains.
For this reason alone I think that the absence of a specific mandate to supervise the covered bonds per se is a suitable delineation between structured and not structured covered bonds.
Whether the supervisory regime is a material difference, or even necessarily a positive for non-structured covered bonds, is something that I will discuss in the next post.
Not everyone agrees with me. Yesterday I was on a call with a respected lawyer (from a jurisdiction that adopts an on-balance sheet covered bond structure, natch) who insisted that all covered bonds that employed an SPV should be considered structured covered bonds. But that is a very self-serving definition, and bordering on parochialism.
By Richard Kemmish