More on Draghi's almighty OMT bluff

A while ago I posted this, explaining how I thought that OMTs were an almighty - if so far successful - bluff.  The reasoning being that the promise is to buy whatever quantity it takes to eliminate a gap between actual and 'fundamental' prices (whatever that means) of troubled sovereign bonds.   That promise exposes the ECB to unlimited capital losses for which there was and is no political support, and possibly no legal support either.

At  a recent lunch this subject came up, and the following line of thought was pursued.  Since at any point in time there is a known and finite quantity of eligible short-maturity bonds, does that cap the fiscal liability of the ECB, and therefore render OMTs not a bluff but a credible promise?

I don't think so, for two reasons.  First, there is probably not even the political support to sustain losses equivalent to the outstanding stock of short-maturity troubled sovereign bonds.  Second, the thought experiment we need to carry it is to ask how the stock of eligible bonds would evolve in a hypothetical 'speculative' attack on a troubled sovereign.  (Speculative in quotes, as such a run would involve calculations no less rational or probabalistic than in any other outcome).  That attack would involve rising yields, and an increasing requirement to issue more bonds, probably tilted ever more towards short maturities, in theory without limit.  So although the current stock of eligible bonds is limited, in the status quo equilibrium where the ECB bluff is believed, the hypothetical stock in the event that there were a run would not be.  And hence the promise is still a bluff.  A remarkably and fortuitously successful bluff.