We have long warned that the effects of a ban on 'free-market' hedging instruments could well have a negative impact on the underlying market that political leaders are 'trying' to protect (consider the fall in equity prices after the short-sale-bans) and this week brought some clarity with regard Europe's Short-Selling-Regulation (SSR) on CDS [8]. As Citi's Matt King notes: "the technical standards underlying its short selling ban reinforce the view we held previously: the ban seems likely to add to selling pressure on cash bond spreads in peripherals, even if it brings down CDS and tightens the basis." The SSR defines 'naked' as CDS that are 'highly correlated' with long bond positions, and bonds have only tended to be quite correlated to their own CDS at periods of low volatility, but this correlation breaks down over sell-offs, which is precisely when hedging is needed most. This will leave portfolio managers unlikely to want to rely on sovereign CDS hedges (which they may now be forced to unwind at any moment) and presumably means they will be reluctant to take out initial long positions in both peripheral sovereigns and corporates in the first place - reducing demand for cash bonds. Once again - regulators and politicians should be careful what they wish for.
Matt King, Citi: As the shorts come off, what will we find underneath?
The European Commission’s publication this week [8] of the technical standards underlying its short selling ban reinforce the view we held previously: the ban seems likely to add to selling pressure on cash bond spreads in peripherals, even if it brings down CDS and tightens the basis.
The latest publication accompanies the SSR (Short Selling Regulation), and defines the methods which will be used to assess net short positions once the regulation comes into force in November. Specifically, it sets out the rules which will be used to establish whether or not they are “naked”, or instead form a permissible hedge.
The SSR will require market participants to net their sovereign shorts with any long positions they hold, either in that same instrument or in other “highly correlated” financial instruments. “Highly correlated” is defined as an 80% correlation over the last 12 months, with an allowance for it to drop below this threshold for a stretch of up to 3 months but remain above 60% for the entire time.
We think these levels are so high that few market participants will be able rely on even genuine hedges being permissible. In peripheral sovereigns, for example, bonds have tended to be quite correlated to their own CDS at periods of low volatility, but this correlation breaks down over sell-offs, which is precisely when hedging is needed most.
During the periods of greatest stress, correlations have dropped below 60% for both Spain and Italy:
[9]
In core countries, hedging will be harder still – correlations are lower and less stable:
[10]
For those who hope to use sovereign CDS as a macro hedge for longs in corporates, the case is harder still. Here, the correlation must meet the same “meaningful and consistent” requirement over the past 12 months, and will again be monitored on an ongoing basis. Even where correlations have temporarily been above 70%, they have seldom remained there consistently:
[11]
In sum, we struggle to think that portfolio managers will fancy relying on sovereign CDS hedges which they could be forced to unwind at any moment should correlations drop below 60%. This presumably means they will be reluctant to take out initial long positions in both peripheral sovereigns and corporates in the first place, reducing the likely demand for cash bonds. So while it does look as though many existing shorts through CDS will have to come off, we are not sure policymakers will like what they find underneath.