The fatal conceit in Paulson's plan4462
Assume the plan works as expected. This means that $700 billion worth of MBS will be purchased from the banks holding them and stashed somewhere at the Treasury. In exchange for such securities, the lucky banks will receive cash or T-bills or something equivalent. So far so good.
Now, my recollection is that the outstanding amount of MBSs and similar securities is pretty much of the same order of magnitude as the outstanding amount of residential mortgages, i.e. $11 trillion. I have got to double check the numbers: the last one I recall dates to 2006
and it was $6 trillion, but I cannot recall if that was the total or
just single family, as the commercial/multifamily may be computed
separately and that is about another three and something trillions. Not all of those are sub-prime loans, but I have found no evidence that subprime loans are restricted into only a clear and easily identifiable subset of such securities. In other words, as far as I can tell there are about $11 trillion worth (face value) of securitized mortgages floating around, some of which contain subprime and some of which do not, some of which contain defaulting loans and some of which do not, some which (finally and more importantly) contain mortgages that will default in the next year or two, and some of which do not. No one knows exactly what is what, where it is and, in particular, no one can possibly figure out and reliably predict which residential loans will default in the next 24 months.
This explains the repeated drying up of the liquidity market: if the securities you are being offered as collateral for the liquid assets you are lending are MBSs, and you cannot figure out what those MBSs contain and what is the expected flow of money they will bring about, you refuse to lend unless they offer you an insane interest rate. Anyone could be trying to set you up with one kind of lemon or another. This makes sense.
Now, how much of this dangerous material can we possibly "mop up" with the $700 billion of the Paulson's funds? We read (NY Times, I believe):
According to Federal Reserve Chairman Ben S. Bernanke, who worked with
Mr. Paulson to develop the plan, the government would pay "hold to
maturity'' prices -- meaning a price based on some estimate of what the
asset would be worth once the crisis of confidence had passed, not on
what the asset holder could get by selling it today.
By doing so, they said, the government would provide troubled firms
with an infusion of capital, reducing doubts about their viability and
thereby restoring investor confidence.
Notice two things: contrary to my friend David K. Levine's good economic sense (and, indeed, the good economic sense of any decent economist I know) the goal is NOT to purchase the good securities from the banks in order to allow them to obtain good assets that the market recognize and accept as such, in exchange for assets that are almost equally good but the market does not recognize as such because of the lemons problem, letting the dumb guys with the bad assets fail. No, the goal is actually to mop up the bad stuff. Forget this, as this is one of the many bad aspects of the Paulson plan that other colleagues have already outlined. The auction theorists that favor the Paulson plan have claimed they can design auctions that will prevent this from happening, for a hefty fee (and for an even heftier one, devise techniques the banks can use to undo the revelation mechanism their colleagues set up for the Treasury). Let's wait and see ...
The second thing to notice is that the goal is not to buy outstanding MBSs at deep discount prices but at good prices based on holding to maturity, i.e. under the assumption that not a large chunk of the underlying mortgages will default. Good. No, bad indeed, but let's take this assumption for good and proceed from there.
What this means is that, with $700 billion you (well, ok, Paulson) can buy, maybe, $1 trillion worth (face value) of the ouststanding MBSs, give and take $100 billion. What this also means is that, EITHER all the bad mortgages that have defaulted and (MUCH more importantly) are expected to default in the next two years, are going to be in the $1 trillion worth of MBSs that the Treasury will purchase, OR lots of the toxic stuff will still be around AFTER the magic auctions have taken place and Mr. Paulson has spent the $700 billion of our money he just received from the Senate. Notice, also, that this will not take a week, or a month but a considerably longer amount of time ...
Let me ask: can we conceive of any possible mechanism that will guarantee the first outcome, i.e. that all and only the risky things will be purchased at the auctions and credibly taken away from the market? My answer is NO: we have no idea of how to do that because we have no idea of where the loans that WILL default 6, 12, 18, 24 months from now are sitting in the enormously vast MBSs market! No one can possibly argue the Treasury can do that, even if Alberto's suggestion (which I fully share and have also been preaching) will be implemented. It obviously will not because, if the Fed wanted to implement it, it could have started two years ago, and it did not nor it asked Congress for the powers to do so. Notice also that, to do this and to mop up the trillion dollar worth of MBSs, will take MANY MONTHS. And in the meanwhile? In the meanwhile the markets will dance the samba ...
If the previous analysis is correct, it follows that the Paulson plan will do zip to alleviate the liquidity crisis. In the best scenario it will take away 10% of the outstanding stock of potentially toxic securities, leaving 90% of them floating around the market. The lucky banks that managed to unload their MBSs to the Treasury will be better off, but (for the reasons everyone know and I mentioned above) those need not be the "good" banks. In any case, about 90% of the financial system (in the USA and in Europe) will still be saddled with securities their counterparts would not be willing to accept as collaterals. Hence, we will be back to square one and the lemons problem which is paralizing the world financial system will still be there, almost intact. No, ok, it will be only .9(intact).
Which is why I keep believing that, given that something MUST be done, we need to attack the virus where it is: at the mortgage level. I have been suggesting for a while (not to the world, just to people I talk to) that we (well, the Treasury of the USA) need to take over the defaulting mortgages at a discount rate of, roughly, .8 to the face value. All we need to do is to keep servicing those mortgages as if nothing happened, while the banks take care of foreclosing on the houses whose mortgages were defaulted and put them on the rental or resale market. The details of how this can be done and how the ownership of the houses can be handled in an (almost) incentive compatible way are boring, and I am still pondering over some sharp criticisms that my friend Sandro Brusco has raised. There are issues, in particular, about how to create incentives for the banks to maximize the value of the houses they take over, and how the gains from such effort should be shared between the banks and the Treasury that will be servicing the loans.
But the thing seems doable, it seems to cost a lot less than Paulson's plan (because we are not purchasing all the loans, we are just servicing those that default, as they come in) and, more importantly, it kills the virus at its source. No matter which MBS you own and what is behind it, you now have the Federal Government warranty that the underlying loan will keep performing at, at least, 80% of its face value. Hence those MBSs become usable as collateral, they regain liquidity. They do not become risk-free - as there is still an interval of about 20% of their underlying value within which they can vary depending on default rates (and then there are still all the old risks of prepayments and what not) - but they are a lot more liquid than they are now. Crucially, and contrary to the Paulsons's plan all $11 trillion of outstanding mortgages become liquid in the sense I just specified!
I have been told that what I propose is a, technically a bit more complicated, variations of the "insure the mortgages" request that the House Republicans had been pushing last week into the bill. Good, for once I stand with the House Republicans - it may have happened already in the past, I cannot recall and it does not matter. So, that's the Republican-Boldrin plan and I will write on its details tomorrow, after the House will have approved the fatally flawed Paulson's plan and after my colleagues have corrected mine, to the extent it can be corrected.
Nobody will pay attention, but it does not matter: this is just a blog of economic theorists after all.