So to start Tyler is saddened that no one took up his initial post on the issue where he says

Draghi says yes to loans to banks, for a three-year period and with weak collateral, and no to loans to sovereigns, in accord with current EU law.  (Admittedly his remarks requirefurther parsing and so this interpretation is subject to revision.)  What does the Modigliani-Miller theorem say?

A cash-strapped government will start a bank.  A cash-strapped government will induce a domestic bank to lend more to it.  A cash-strapped government will force a domestic bank to lend more to it.  Remember the era of “financial repression”?

To some extent governments will internalize the value of this guarantee to banks.  If you don’t think this guarantee to banks somehow transfers to governments, won’t ECB-guaranteed claims on the bank become the new safe domestic security, knocking out the market for the riskier sovereign stuff and thus mandating some kind of risk equalization to keep the whole show up and running?

If its any consolation I would have commented but was busy with other things.

I wouldn’t think of it the way his second paragraph reads, but this relates to what I’ll talk about below.

I am not sure I know what he is getting at in the third paragraph but right off I would say that there is no sensible “domestic safe security.” What matters are the securities available inside the Eurosystem. Now, it is true that there were problems in the money markets and for a while in equalizing effective money rates under the various National Central Banks, but solving that problem doesn’t alter the dynamics in a way that I see “knocking out the market for the sovereign stuff”

In any case what can we do with the 3yr loans to the banks.

Well, we can produce essentially unlimited levels of liquidity. Now, if you had one bank that was big enough the debt crisis for Italy, at least, would just go away in the near term. I’d have to look at the other countries to see what was possible there.

Here is how.

So, I am the big Italian Bank. First, I just pack up my entire stack of Residential and Business Loans and ship it to the ECB. Procedural details are important here, but if I took no haircut AND if I am paying the overnight rate AND if payment is not due until the end of the three years AND I keep the yield on my collateral then just ship it all. I am sure I don’t get all those things, so the deal is not as good but we can use that as an extreme point to anchor the thought experiment.

So, now I got a bunch of cash.

Then I buy a bunch of Italian debt. I face a tradeoff because I want to move the market as little as possible but I also want to buy as much debt as possible. There is some profit maximizing amount here that balances those two concerns.

Then I sign a deal extending the Italian government a line of credit for three years sufficient to handle all of its financing (this is why I have to be big enough.) at some clearly manageable rate say overnight plus 50 bps.

As soon as I do that, the yield on Italian debt will collapse. The three year and less yield should collapse all the way to the overnight rate. Probably the 10 year will collapse pretty far as well as this steers us towards the “Italy self-finances” equilibrium.

Then I sell all of my Italian debt and make a ton of money.

Now, my big risk is that I have this credit line outstanding with the Italian government. If we get down the road close to 3 years and things start to look like they are going badly then the yield on Italian debt is going to rise and the government will exercise the line. Then if things do go badly I go down completely.

The key question though is whether I am actually carrying any non-existential risk. This is a nice place to link up to the question of “Too Big to Fail”

Because, if things went badly what would probably happen is that the ECB would bail me out and then everyone would whine that the only reason I did this was because I knew the ECB would bail me out. And, we should end “Too Big to Fail” blah, blah, blah.

However, that is false.

The issue is more fundamentally this. If Italy can’t make it with the deal I am offering, then Italy certainly can’t make without the deal that I am offering. However, if Italy can’t make it then the fallout from an Italian default and likely Euro breakup would mean that I couldn’t make it without a bailout anyway.

So, in the world where I get bailed out this loaning money to Italy doesn’t adds limited risk because I will get bailed out. In the world where I don’t get bailed out loaning money to Italy doesn’t present much risk because if the fundamentals are bad enough to make this a bad loan then I don’t survive anyway.

Thus, there isn’t a state of nature in which loaning money to Italy alters my survival. Thus, if it may be profitable to loan money to Italy I should do it.

Now, that having been said the ECB could manufacture a state of nature in which loaning to Italy would alter my survival. They would do this by saying we will bail you out If and Only If you don’t loan to Italy.

This is like a conditional bailout which how bank regulating often works. But, absent that, the policy stance of the ECB doesn’t matter because it doesn’t change the consequences of my actions.

Okay, so all of that was with a big enough bank.

The reason it needs to be big enough is so that it can internalize the gains from arbing Italian debt and the ECB lending facility.

You could also have multiple banks that were big enough in co-ordination to do this but they would have to work closely because if you can’t guarantee the line of credit then you can’t buy up the secondary market bonds which means you can’t profit easily from guaranteeing the line of credit.

On the other hand if there were enough Italian banks already holding enough debt to make this scheme profitable then you could arrange it much more easily. You could possibly get some banks from other countries in on it too but it depends on what kind of existential exposure they face with respect to Italy.

However, generally speaking if the size of the banking sector is big enough you make this work.