As I’m asked about the optimal choice of covered bond model quite frequently, I thought I’d share a few thoughts, the final and newest alternative: the special purpose vehicle.
Having worked on the very first SPV based covered bond structure sixteen years ago (OMG), I could be accused of bias. In addition to being the newest model, it is also the most fashionable, having been used - according to my highly unreliable calculations – in 11 of the 17 newest countries to the covered bond market (to 5 on balance sheet and 2 special bank. Yes, I know that doesn’t add up to 17. Told you my numbers were unreliable*).
The advantages of the model are clear enough but it is a bit more complex and, whilst you needed the lawyers to get it right in the ‘direct issuance’ model, you need the accountants to get it right in this model. I’ll leave it up to you to decide which is more likely.
The advantage, at least from a legal point of view, is that for most of the exercise you use existing technology rather than have to invent anything new – you sell assets to a third party rather than create new ‘on balance sheet’ structures. The third party can easily enter into the other contracts that it needs to. It’s legally straightforward.
Of course, there are legal changes necessary sometimes – particularly in those countries where the transfer of assets by sale contract between the issuer and the SPV needs a little help from statute law. But these are, at least conceptually, straightforward.
It is structurally more complex and that can put some people off, this is supposed to be a simple product after all. But structural details such as notices to pay under guarantees aren’t really that complicated compared to whatever will happen to a ringfenced pool when an issuer under the direct model fails.
No, the real problem is making sure that the legal segregation of the assets doesn’t create any kind of commensurate economic segregation. Supervisors, auditors, accountants, bank capital supervisors and internal stakeholders all need to be understand that the legal segregation needs to be ignored for bank capital, taxation, operational and accounting purposes. That’s easier in countries that follow the substance of a transaction rather than its legal form. It’s a secured borrowing in substance, treat it like one.
Having commented on the three models in these last three posts, one important point: they all work. Investors and rating agencies seem to be more or less indifferent to the choice of model and it is too easy to overstate the importance of the choice when contemplating a new covered bond jurisdiction. Easy and dangerous.
One covered bond is better or worse than another one for many possible reasons, the choice of legal model isn’t one of them. So try not to lose any sleep over it.
*OK, one country allowed two models in their new law.