Monday, February 15, 2010

Sovereign CDS: Cat or Canary?

Right, time for a little bit of perspective on the sovereign CDS are good/evil (delete as appropriate) debate, courtesy of BarCap.

The above charts show the relative size of the CDS market in the eurozone periphery as proportion of outstanding government debt.

As you can see they range from a few percentage points to 10 per cent in Portugal. For other developed countries (Germany, Frank, UK and the US) the proportion is even smaller, according to BarCap. A maximum of 1.5 per cent and even lower for the UK and US.

Now, here are some more charts, which show how spreads move relative to market activity.

What this shows, BarCap says, is two things:

Spread widening, triggered by real issues, was exacerbated by the sovereign CDS market, where the price discovery process is more skewed towards ‘shorts’ than in the cash markets.

The ensuing widening of spreads, and as importantly, volatility of these spreads, then caused cash market participants to adjust their positions (reducing longs or going underweight), with the move feeding on itself, and leading to a deterioration in the liquidity of the cash markets. Essentially, an initial CDS-driven move would thus have been followed by a generalised risk reduction and loss of liquidity. This is somewhat supported as well by the evolution of the cash-CDS basis.

And on that final point, BarCap makes the following observation:

These charts show clearly that the CDS market has been much more active when spreads are volatile and widening. On the one hand, this is not unexpected; market volatility tends to generate activity. The charts do not establish that activity in the CDS market leads the widening of spreads as such; the moves seem to be contemporaneous (similarly, an analysis of the evolution of cash and CDS sovereign spreads points to no particular lead or lags between them).

On the other hand, both charts show that CDS activity drops quite a bit when spreads tighten. This would suggest that the CDS market tends to be dominated by players who are looking to buy protection (ie, be short in cash terms). This may be particularly the case in markets where it is more difficult to be short.

For example, the Greek repo market is not centrally cleared, which limits the appetite of dealers and investors to be short in specific bonds (there are only around €5bn of shorts currently – much the same as in Portugal, which is a 50% smaller government bond market, but is LCH cleared). We suspect as well that the mark-to-market sensitivity of a number of these players might be higher than the one of those active in the cash markets (which would tend to be dominated by longer-term, more passive types of investor) – a factor that could generate more volatility.

Bringing it altogether and considering the relative sizes and liquidity of the markets, BarCap reckons investors should be looking to government bond markets (and in particular to measures of government swap spreads versus OIS curves) rather than the CDS market for an accurate gauge of sovereign default risk.

Sovereign CDS, it says, are neither the ‘canary in the coalmine’ or the ‘cat among the pigeons’.

No comments: