Toxic Assets: Promise, But Also Peril

It appears we are about to have served upon us (finally!) the plan from Treasury to "cleanse" the banking system of "toxic assets".

From the WSJ:

The administration plans to contribute between $75 billion and $100 billion in new capital to the effort, although that amount could expand down the road.

The plan, which has been eagerly awaited by jittery investors, includes creating an entity, backed by the Federal Deposit Insurance Corp., to purchase and hold loans. In addition, the Treasury Department intends to expand a Federal Reserve facility to include older, so-called "legacy" assets. Currently, the program, known as the Term Asset-Backed Securities Loan Facility, or TALF, was set up to buy newly issued securities backing all manner of consumer and small-business loans. But some of the most toxic assets are securities created in 2005 and 2006, which the TALF will now be able to absorb.

Finally, the government is moving ahead with plans, sketched out by Treasury Secretary Timothy Geithner last month, to establish public-private investment funds to purchase mortgage-backed and other securities. These funds would be run by private investment managers but be financed with a combination of private money and capital from the government, which would share in any profit or loss.

But then later on we get to the key provisions:

To target troubled securities, such as mortgage-backed securities, the government will create several investment funds. Treasury will act as a co-investor, in most cases contributing $1 for every $1 contributed by the private sector and sharing in the first-loss position.

To target troubled loans, the government will create a Disposition Finance Program with the FDIC. In that case, the government will be a co-investor, but could also agree in some cases to contribute 80% of the financing, with the government putting up $4 for every $1 in private financing. As part of that program, the FDIC would provide guarantees against losses on a pool of loans that a bank wants to sell. The program could guarantee as much as $500 billion in loan investments.

1:1 contributions into the program aren't all that troublesome, provided that "first loss" is taken by the investors.  After all, with 2:1 leverage you need to lose half (of an already distressed price) before the government loses money.

But 5:1 leverage is a bit more troublesome.  It amounts to 80% financing, which means that being off by more than 20% winds up back on the taxpayer's head.

The NY Times story implies even greater leverage - and thus taxpayer risk:

To entice private investors like hedge funds and private equity firms to take part, the F.D.I.C. will provide nonrecourse loans — that is, loans that are secured only by the value of the mortgage assets being bought — worth up to 85 percent of the value of a portfolio of troubled assets.

The remaining 15 percent will come from the government and the private investors. The Treasury would put up as much as 80 percent of that, while private investors would put up as little as 20 percent of the money, according to industry officials. Private investors, then, would be contributing as little as 3 percent of the equity, and the government as much as 97 percent.

97%?  Oi.

Now the plan itself is not necessarily bad.  In fact it would be an interesting gambling exercise were I so inclined (and had a bunch of people I knew with a lot of money) to pony up a few billion and take a gamble such as this.  If you're wrong and the assets never come back in value, continuing to decline (or go to zero) then you lose the entirety of what you put up.  But with anywhere from 20:1 to 50:1 leverage if you're right the gains will be enormous - even if you do have to share them with the Treasury.

Indeed, drawn properly this plan can actually work. 

Why?

Because there are a lot of people who will take this gamble if their losses are limited to capital put in but their upside is some multiple of that capital.  This activity can and will attract a lot of very bright analytical types who can comb through these assets and assign what they think is a probable value down the road if held to maturity, and then discount that back to current price.  If there is a delta, multiplied by the leverage advantage in that calculation this is a very nice risk:reward bet, and one that anyone with a brain would fall all over themselves to participate in with a piece of their risk-based capital.

The key though is the "drawn properly" part.

Two things need to happen for that to be the case:

• The reward must be outsized compared to the risk in order to provide the inducement to participate. That is, if I have 20:1 leverage, I need to get more than 5% of the appreciation, if any.  If you give me double my capital contribution percentage in appreciation I'm interested.  If I get 25% of the appreciation (five times my contribution on a ratable basis) I'm a lot more interested. If I get half of the gains I'd be insane not to take the gamble.
• You must prevent "gaming" of prices so that they are not bought too high, thereby guaranteeing losses taken by the taxpayer.

The second point is the problem although it doesn't seem so at first blush.  Why, after all, would a private party intentionally overpay .vs. what their analytics say is the true value over time of these securities?

The problem is that there is a perverse incentive for a bank to participate through some back channel if it can, given sufficient leverage.

Let's say I'm "Frobozz Bank" and have $100 billion of this trash (a lot!) on my balance sheet.  Its mostly performing (for now) on a cash-flow basis, but I know what the deterioration in on-time payment flow looks like, and as a consequence I know in advance that eventually this paper is going to be worth much less than my "internal" marks (that I'm reporting every quarter on Level 3.)

So here comes Treasury.  They offer 20:1 leverage (I put up 5%) and the "private parties" bid for the assets, with their maximum loss being capped at their contribution (that is, if there's more than a 5% loss the taxpayer eats it.)

Aha!  Now if I can be the "private party" I can overpay on purpose, capping my losses at 5% of whatever I "buy" from myself!  I am thus able to transfer the other 95% of the risk onto the taxpayer and I escape with a 5% penalty off the purchase price!

That, if it happens, is an enormous scam and Treasury and the FDIC must absolutely guarantee that it not occur.  If it does, we the taxpayers are going to be violated to an insane degree while the true "risk money" (and there's a lot of it out there) won't go anywhere near this program, because they, being unwilling to overpay, will simply lose the bidding contests.

So in order to prevent intentional overpayment you must as a matter of policy (and even law) enforce strict separation of the funds that are doing the buying from anyone that has an interest in the sellers, and make clear that if you catch anyone cheating extremely severe sanctions - like 100 years with Bubba - will be the consequence.

If you do not the process will get gamed and the taxpayers will lose.

IF, and I repeat IF, the proper protections are in place on this program it has promise.  It also has peril, but short of simply swooping in with the FDIC and taking these institutions over this may indeed prove to be the best alternative.

A year ago I said that the FDIC should simply come in and close all of these banks.  But you have now seen what indecision and inaction has done in the case of IndyMac - had the FDIC acted when it should, the losses would have been small or even zero.  Since they did not, its in the tens of billions.  As such while the "clean" solution remains having the FDIC step in this no longer the "reasonable or low" loss scenario - it is now a circumstance where a literal trillion dollars in immediate capital could be required by the FDIC (which it doesn't have and can't raise) and while some of that (perhaps even most) would be recovered over time, in the interim you have to finance it - and I doubt we can.

As such with proper protections I am willing to support this program.

Without proper protections Tim Geithner (who I do not trust as far as I can throw him), Ben Bernanke (ditto) and Sheila Bair (who did not step in when she should have in the case of IndyMac) should all face being grilled and eaten - literally, with BBQ Sauce - by the taxpayers if this blows up in our faces.

After all, these three clown-car riders have been the prime participants to date in what I can only characterize as a looting and willful-blindness operation of colossal size. 

If they intend to convince me otherwise in this case, the burden of proof is on them.

Let's see the rules for participation include sworn statements, under penalty of that 100 years with Bubba and full, public disclosure, that the funds used to participate have no tie, direct or indirect, back to the sellers of these "assets."