There are several posts between amv's last contribution to our discussion concerning the Greek debt crisis and the points made by John Cochrane. Amv is right: our opinions differ only marginally. But let me elaborate again on these small deviations:
- In my opinion, it doesn't make a difference for policymakers whether the recent developments were triggered by fundamentals or by irrational and self-fulifilling market behaviour or - most likely - a mix of both. We are in a very serious situation now - regardless of the cause. And in a first step, all should be done to avoid panics and contagion. But of course making the second step afterwards is crucial. Adressing moral hazard issues on the level of market participiants and aiming for sound fiscal policy in the member states is absolutely necessary.
- Amv's arbitrage argument is not convincing to me. In fact, there were institutional incentives which avoided arbitrage. Just take the ECB's collateral framework treating Greek bonds equal to German ones. I suppose that the "missing markets argument" might be applicable, too. This argument was brought up in the light of the subprime crisis. It basically states that there was no market to bet on the bust of the housing market. The major share of the innovations in mortgage finance focused on participating from the boom. This led Bob Shiller to the conclusion that we need even more deriviatives to cover every possible future state, which is a quite remarkable statement given the current political pressure on the financial sector. And CDS markets for EMU government debt don't seem to have been very liquid in the years on from 2004. This is at least what a quick look on the pricing data suggests. So buying protection and betting on a Euro debt crisis might have been very difficult.
- Nevertheless, I am fully in line with amv's call for sound policy and a reform of the institutional framework. Economic theory is well equipped for making proposals. OCA theory tells us what to do, and political economy tells us why these measures are difficult to implement.
