In my previous post I asked what a structured covered bond actually is. But how should it be valued?
A good starting point is to define the upper and lower bands of the bond’s value. In theory a covered bond should fall somewhere in between a senior unsecured bond from the same issuer and a ‘regular’ covered bond, surely?
The spread on a senior unsecured bond from the same issuer should be the upper boundary of the structured covered bond’s spread. There are just possibly some reasons why this would not be the case – the structured covered bond might be less liquid or have an off market coupon, but these are no more than normal variation in value between bonds from the same issuer and would have to assume that no investor gives any value to the structural elements that is greater than these minor inconveniences.
That assumes that you are comparing like for like in terms of eligibility for bail-in. According to the bank recovery and resolution directive, secured bonds are exempt from bail in, as are those covered bonds that are guaranteed but not secured, those in Italy or Holland for example. But a structured covered bond that is guaranteed (as most of them are) and therefore isn’t explicitly secured might not be exempt – so the correct upper boundary should be senior un-preferred debt of the same issuer.
That established, what of the other boundary? Could a structured covered bond ever be better than a traditional one? In theory: yes, in practice: it’s difficult to see.
All covered bond investors have huge theoretical exposure to one of the most over-valued asset classes in Europe: residential property. That’s fine because of all of the other mitigants in place to protect the credit. But what if house prices crashed at the same time that bank credit started to fall – they are pretty closely correlated after all - and what if, at the same time, a populist government started interfering heavily in the mortgage market in a way that severely damages creditor’s rights?
At a time like that, or in anticipation of a time like that, diversification into a less sensitive and over-valued asset class might start to look attractive. Loans to SMEs being the structured covered bond market’s go to example alternative asset class. Enough to overcome the negative features of a structured covered bond? Possibly. But probably not.
Having established the upper and lower value boundaries (and the cases when they might not actually be boundaries at all), is it right to think of the relative value of a structured covered bond in terms of the continuum between the boundary valuations? That’s certainly the approach that has been adopted in the case of the recent Deutsche bank covered bonds (about one third of the distance between the boundaries) and entire articles were written about this in the case of the 2013 Commerzbank transaction (which neatly ended up almost exactly half way between the two extremes).
But a structured covered bond is a covered bond – with imperfections – not a hybrid between a covered and a senior bond. So I would rather define the value in terms of either, a covered bond, with additional spread for the areas in which it might be worse, or a senior unsecured bond given credit for the areas in which it might be better, both subject to the constraint of the boundaries that we started with, but not defined by the distance between them.
So, for example, take the value of a traditional covered bond, and calculate how much the worse regulatory treatment of a structured covered bond would be valued by the marginal investor. Then add in factors for other variables: different rating transition probability (could be higher, could be lower, depending on the structure), different investor perceptions of the underlying assets, different treatment in repo, liquidity facilities, different eligibility for purchase programmes, etc.
Perhaps the biggest differentiating factor is the one I mentioned in my previous post – the lack of a bank supervisory regime dedicated to the interests of covered bond holders. Theoretically this should be taken into account in the rating assigned – the rating agencies will give it some value in the case of a traditional covered bond, if the structured covered bond wants to get to the same rating they need to give something else of equal value instead – realistically, more collateral. Is the value assigned to the supervisory regime by the rating agency correct? Would you rather take a bond with 5% over-collateralisation and covered bond specific supervision, or one with 10% o/c and just normal bank supervision? Impossible to generalise, but precisely the sort of question that needs to be asked when valuing a covered bond, or, for that matter, a supervisory regime.
Or you could approach the problem from the other direction. Take a senior unsecured bond and give value for the higher credit rating achieved – relatively easy to quantify based on generic credit rating/spread curves.
Relative value is an art, not a science. All three approaches to valuing a structured covered bond could equally be valid. That’s what makes it so difficult.
By Richard Kemmish