What price sovereign risk within a bank bond? BIS answer here:
That’s a chart from a fresh paper on sovereign effects on bank funding, released by the Bank for International Settlements on Monday. If you’ve followed the eurozone’s vulnerability to the sovereign-bank loop effect on FT Alphaville, quite a bit will be familiar. There’s much here about transmission channels through central bank funding and credit ratings for example. There’s also another nice chart of the fortunes of government-guaranteed bank bonds across jurisdictions since 2008:
But we think the paper is at its most interesting with its views of banks trying to hedge this sovereign risk. Generally, it’s not an optimist. For example, here’s one of the clearest explanations so far on how derivatives hedging can have a perverse effect on risk premia, via credit value adjustment (CVA) activity:
The impact on banks is exacerbated by the fact that sovereigns (and other highly rated entities) often use unilateral credit support annexes (CSAs), meaning that they do not post collateral to offset mark-to-market losses on derivatives, but will receive collateral on their mark-to-market gains. This negatively affects banks in two ways. First, banks’ mark-to-market claims on sovereigns are uncollateralised, increasing their CVA risk. Second, if banks hedge their derivatives positions with sovereigns using offsetting trades with other entities that are covered by bilateral CSAs, then banks can face additional funding strains as they need to post collateral in one transaction without receiving any reciprocal collateral in the corresponding hedge transaction. Banks sometimes hedge themselves against sovereign risk by buying CDS protection or short-selling government bonds, but depending on the liquidity in these markets, this can push up sovereign risk premia and cause further CVA losses…
(Note that the paper concludes that regulators should embrace two-way CSAs for sovereigns)
There’s a further intriguing tidbit on a similar effect within the iTraxx Financial index, something a bit less discussed:
According to market commentaries, investors’ hedging strategies may have contributed to tightening the link between sovereign risk and bank funding conditions. They assert that the iTraxx Financial Index is being used as a proxy hedge for sovereign risk due to the lower liquidity in the SovX index and individual sovereign CDS contracts. And once the iTraxx Financial Index moves, its constituents (CDS of individual banks) also move, as inconsistencies between the index and the single names are arbitraged away.
A rigorous test of this hypothesis is beyond the scope of this report. However, since late 2009, sovereign and financial CDS indices have co-moved closely (Graph 21). Moreover, since early 2008, the credit risk of individual euro area countries – for which the spread between each country’s 10-year bond and the German bund is a proxy – has shown an increasing correlation with the iTraxx Financial Index, and the level of correlation (for both weak and strong countries) has at times been quite high. The single-name CDS included in the iTraxx Financial Index have also been co-moving closely with the index itself, suggesting that banks’ idiosyncratic factors are playing only a small role in their price dynamics.
All very interesting. And topical.
Say if there was a banking system which didn’t tend to issue as many bonds itself as compared to international peers… but did purchase lots and lots of its sovereign’s debt, and will naturally seek both assurances against marked day-to-day volatility and capacity to hedge exposures when volatility does materialise…
…well, then Italy would be a prime, huge test case, wouldn’t it?
Related links:
What keeps an EU finance minister awake at night? – FT Alphaville
The Janus-headed salvation – ECB paper, 2009
Bagehot, bailouts and banks – the entwining continues – FT Alphaville