On the negative basis trade

Repost: On negative basis trade

This is a repost of something i wrote in May 2010. The reason for the repost is that I want to use the power of Markdown to make it look “nice”.

On FT Alphaville there was a discussion on the persistence of a gap between the price of CDS on Greek Government Bonds (GGB) and the spotmarket price of GGB: As noted by Barclay Capital, the GGB spreads have widened relative to CDS premiums, allowing an arbitrage trade without default risk from the issuer of the bond.[1]
One possible explanation for the persistence of the gap between CDS premia and spotmarket may be the counterparty risk. First, approximate the probability of default of GGB by the spread:

\[\Pr(D_{GGB})=r_{GGB} – r_f\]

The value of the CDS from the perspective of a buyer is – for some value of recovery rate \(R\), assuming no counterparty risk:

\[CDS = (1-R)\cdot \Pr(D_{GGB})\]

If we allow for counterparty risk \[\Pr(D_{CP})\] this becomes:

\[CDS = (1-R)\cdot \Pr(D_{GGB}) \cdot \Pr(1-D_{CP})\]

Now, if we allow for the weak form of the efficient market hypothesis, a widening between GGB on the spotmarket and CDS-premia could have two sources: a rise in the expected recovery rate, or a rise in the perceived counterparty risk. Since S&P has set a low expectancy value on the recovery rate (between 30-50%, far below the average recovery rate for sovereign defaults), this persistent, and widening gap is only consistent with a widening in the counterparty risk.

[1]: One could buy the bond and buy default protection – a CDS – and cash in the difference between the bond interest rate and the insurence premium.