In a note on Friday, credit strategists at Citi noted the widening gap between the CDS spreads of European financials and their sovereigns, and went so far with it as to use the D-word: decoupling. Decoupler #1 being Italy, driven in no small part by the increasingly desperate need of banks to hedge against the sovereign, i.e. “silo trading”.
In more logical rational normal times, sovereigns act as a floor for their banks. It’s due to the tendency of governments to come to the rescue of lenders that are too-big-to-fail, or “systemically important financial institutions”, SIFIs (to use the latest and most fashionable acronym).
It has, however, become increasingly clear that sovereigns need to put their own oxygen mask on first and get their budgetary situation in order, before assisting others. And in the meantime, banks are being subjected to more stringent capital requirements. From Citi, regarding banks:
In particular, the market has taken comfort from the prospect of potential capital injections, increased capital ratios, and the prospect of guarantees on new bond issuance. But as yet, nothing is decided.
The series that Citi use to illustrate the decoupling are the Markit iTraxx SovX Western Europe and the iTraxx Europe Senior Financials. Graphing the two from the start of the SovX’s existence:
The tricky thing with using the Markit iTraxx Europe Senior Financials though is that it contains insurers too, which isn’t exactly what one wants to look at. Here are the exact constituents of the index and how they were doing on Monday morning:
On the one hand, Deutsche Bank and UBS, but on the other Aviva and Munich Re.
FT Alphaville Credit GeekBox
We enjoy reading The Economist on the weekend, but we do wonder what’s going on with charts like these:
The spreads in the chart seem like they are those of the Markit iTraxx Senior Financials, but why compare that (inclusive insurers) to an average of six large US banks? Why not do an average of European banks too? Or just say it’s the iTraxx Senior Financials? Again, we ♥ The Economist (and we ♥ footnotes, and want to believe that these are compatible ♥s).
The point that the Citi strategists are trying to get at is that the decoupling trend between sovereigns and financials, and even corporates, is more pronounced for some than others. Italy being a case in point. The below graph is still including insurers in order to keep to the iTraxx Europe universe.
And why is the decoupling so pronounced for Italy?
Back over to Citi (emphasis ours):
This seems counterintuitive due to the high quantities of Italian sovereigns on the balance sheet of Italian financials. We reckon this is more due to ‘silo trading’. CVA desks want to hedge their Italian exposure but are less concerned by Financials and corporate CDS. Equally, US hedge funds wanting to take a view on Europe are probably not as interested in the spread between sovereign CDS and financials. As a result, we’ve seen some temporary disconnect between the two.
So even though Italian banks hold a non-trivial (cough) quantity of Italian sovereign bonds, their CDS spreads aren’t being pushed up as much because of the more substantial demand by all banks to hedge against the sovereign alone. There’s also a lack of interest from hedge funds whose activities may otherwise have served to compress the differential.
The situation is even more extreme when one strips out financials and looks at all other Italian corporates:
Which leads the Citi analysts to conclude that corporate spreads are due some widening.
The main focus for now, however, is on how strong technical factors are for Italian sovereign CDS, where the “technical” in question is the demand from banks’ CVA desks, which manage the counterparty risk that the sovereign poses to them.
Bond holdings are a factor, but Italy is quite a special case, in a way, because of the large derivatives portfolio that it has with banks. As a sovereign, it doesn’t post collateral, which you can imagine makes banks feel rather vulnerable (to put it mildly). Portugal and Ireland do post collateral, but that’s a fairly recent innovation.
Over the course of a number of years, Italy issued a lot of CCT bonds that are variable rate. At the end of 2009, “General Government debt” was €1,763bn of which €235bn was variable rate. By August 2011, the same figures were €1,900bn and €187bn as the variable rate debt runs off and isn’t replaced.
At some point or another Italy decided it didn’t much like paying out variable rates and entered into interest rate swaps with banks. Under those swaps, Italy paid the banks a fixed rate and received a floating (i.e. variable) rate which it then passed onto the bond holders. Hence Italy was left paying the fixed rate its bank counterparties, thereby effectively turning its variable rate debt into fixed rate debt.
As rates have fallen over the last few years, the variable-rate amount that banks have to pay Italy has fallen, and the fixed rate that the swaps are locked at is a lot more than the current market rate, making the banks in-the-money to Italy. That, in turn, leads to increased levels of anxiety on the CVA desks at banks that are tasked with managing counterparty risk.
Italy also has issued debt in foreign currencies and swapped that back into euros. As the euro has depreciated, the sovereign is out of the money on those swaps too.
The interest rate swaps and forex swaps certainly aren’t the sum total of all of Italy’s derivatives positions, but overall the positions see many, if not most, of the sovereign’s counterparties to be in-the-money, i.e. the banks are showing gains on the swaps while Italy’s accounts should show a loss, albeit a paper one.
And so every time Italy starts to look like a riskier counterparty, the more the CVA desks buy CDS (or short Italian bonds for that matter), which in turn pushes spreads up further, which makes them look more risky and so on. In fairness, maybe this shouldn’t be called a “technical”. Perhaps an ugly fact of life (UFoL)?
Whatever it is, it leads one to wonder to what degree Italian CDS spreads are influencing bond yields. There are many factors at play here though, and this is just one.
And is the move wider being exacerbated by naked shorts? The ban, after all, doesn’t come into effect until the summer of 2012.
Many place the tipping point for Ireland and Portugal at the day that clearinghouse margin requirements for the countries’ bonds went up. CDS spreads are one of the factors that LCH Clearnet Ltd’s RepoClear service takes into consideration when deciding on those collateral haircuts, as well as the yields over an AAA benchmark index.
We leave you to contemplate the current state of play:
Related links:
Are Italian Yields Once Again At The Tipping Point? – The Big Picture
Allies intensify pressure on Berlusconi - FT
Sovereign CDS posterchildren – FT Alphaville