Wednesday, May 26, 2010

Reply to amv's quotation of John Cochrane (ls)

I doubt if Cochrane is right with his assessment of systemic risk emerging from a Greek default. In my impression it is not about "rationally" quantifying individual exposures of banks any more. Concerning this point, I am fully in line with him. But it's rather about trust in these days. Therefore it's not relevant whether a Greek default is a systemic event in a quantitative dimension, it's relevant whether the market considers is as one. ECB President Trichet emphasized several times that market conditions got worse in the days before the bailout consensus and the ECB decision to purchase gonvernment bonds (see here).

On 6 and 7 May [...] the markets were massively disrupted, and market participants all over the euro area and the rest of the world observed this situation, which was deteriorating extremely rapidly. This was a very serious situation, in which we had to intervene in order to preserve [...] the integrity of the monetary policy transmission mechanism.

We all don't know what would have happened to the interbank and bond markets without this consensus. But it's plausible to assume that the market sentiment would have been severely affected which might have led to a evaporation of both market and funding liquidity.

And it's also wrong to blame banks for betting on the Greek default. The underlying problem is several years old. The ECB collateral framework treats government bonds of Eurozone members equally. They are subject to same-sized haircuts despite of their different ratings. Hence, Greek and e.g. German bonds were perfect substitutes for ECB refinancing operations. Therefore, the incentive to distinguish in terms of credit quality was weakened. I guess, some banks were even happy to pick a few basis points extra yield on holding Greek bonds instead of German ones. Willem Buiter and Anne Sibert already pointed to this problem in 2005. When spreads began to widen after the Lehman default, the secondary market for Greek bonds didn't work well any more. So banks probably weren't able to reduce their physcial exposures. This could partly explain the dramatic increase of the CDS prices, as banks might have started to demand protection. But that's speculation, which is - as we are extensively told in these days - evil.