Latest News:
Amid all the downgrades to European sovereigns, the ECB recently and pointedly asked all rating agencies (paraphrase) “how can a structured finance bond in Portugal have a debt rating higher than the Portuguese sovereign?” The agencies jumped to the task of creating and publishing their responses.
Moody’s, S&P, and Fitch responded uniformly that the top rating of a securitisation can be as many as six notches higher than the sovereign! These rating agency reports go on for many pages but provide minimal hard insight. One can summarise the lot of them as “sovereign risk is important, we consider it carefully, hence we impose this truly onerous cap of six notches.” What this means is that a securitisation embedded entirely within a country of sovereign rating A- can have a top tranche rated triple-A (without even a clear statement that the sovereign risk requires more credit enhancement). The Moody’s report is best because it discusses most of the important sovereign risks: the obligors of the SPV assets will perform much worse than projected when the sovereign is in default; the financial institutions that act as servicer, collection account holder, and swap counterparty will fail with the sovereign; and enforcement of contracts and property rights may weaken with political turmoil.
But nobody asks obvious questions. If an RMBS bond rating is six notches higher than the sovereign, then the rating agency considers it highly likely when the sovereign defaults that the bond will survive and make all payments with no disruption. What analysis provides this level of certainty? Why is it not far more defensible to say that securitisations strongly embedded within a sovereign will have significant risk of disruption for numerous reasons upon sovereign default? Such reasoning would cap such securitisation ratings at the sovereign level or very near this level.
What do Creditflux readers believe? Is it reasonable to have securitisation ratings well above the sovereign? Should Greek RMBS be investment-grade while the Greek sovereign is single-B?
This article has been submitted by a Creditflux reader as a suggested point for discussion. Mr Wraith describes himself as a wizened rating agency insider traumatised (by) history. Please add your comment below, and send your own suggestions for similar points for discussion to editorial@creditflux.com.


It is recommended that you do not log out if you regularly access Creditflux on this computer.
Once you have logged out you will need to re-register by entering your email address and receiving an email from us to gain access.
Click here if you are sure you want to log out.

Already a registered user? Click here to login.

This article is only available
to Creditflux subscribers.
Already a subscriber? Click here.
As a part of your trial subscription
you will receive:


Bookmarking this article will save it in your membership area for your reference at a later date. You can bookmark as many articles as you like.
To access your membership area click here or on 'Manage My Account' located in the top right hand corner of any page. You must be logged into the site to use this feature.
For help, please contact us on
+44(0) 20 7253 9510.
In some ways it makes no sense as if you have a 6 notch ceiling, then you are saying if the sovereign defaults, then the securitisation is automatically downgraded to B+ (if my maths is right and this is 6 notches above D). However at that stage, it would probably be known whether the underlying RMBS were going to keep paying or not as the nature of the sovereign default would be known (bond holders take a haircut on their principal and everything continues as normal or moratorium with currency controls with no payments on domestic assets to those outside the country). Also on simple maths if a B+ rating implies a 25% chance of loss in 7 years and Greece is already B+, then this implies on the 6 notch ceiling there is a 25% chance of Greece defaulting and the RMBS going to B+, which would imply that the RMBS has a 25% chance of finding itself in a state where it has a 25% chance of defaulting itself as implied by its B+ rating. This means on this logic it has a 25%*25% = 6.25% chance of defaulting today on this risk alone so it should be rated no higher than BBB- (4 notches above the sovereign)
I would tend to agree with the rating agencies, although I don't know if 6 notches is the correct distinction. I would suspect the notching should be deal specific. Moody's has the strongest argument because they define their ratings as expected loss opinions (vs. default risk opinions). An expected loss philosophy acknowledges there may be delays in payment (defaults) but the rating speaks to both default and loss given default. So the sovereign failure may delay a structured deal within that sovereign but the rating speaks to the expectation the recovery will be very high.
I would say - rating agencies aren't right about anything they do. Why would you expect to be them right in this case?
Yes, I agree with the rating agencies, in principle. A similar case happens with covered bonds - the ratings of the covered bonds can be several notches higher than the rating of the senior unsecured debt of the same Issuers. And correspondingly the market requires for the covered bonds a yield far lower than the one required for senior unsecured bonds. Ditto in RMBS, spread required by the market on Greek RMBS is lower than spread on Greek government bonds.