Thursday, April 03, 2008

Wharton on the Future of Securitization

by Tanta on 4/03/2008 04:04:00 PM

Sigh.

Some days all I can do is sigh.

The Wharton faculty interviewed for this piece, "Coming Soon ... Securitization with a New, Improved (and Perhaps Safer) Face," seem mostly to believe that securitization of mortgage loans isn't going to to away after the recent debacle, nor is it going to go unchanged. I certainly agree with both claims. But I hate stuff like this. Those of you who think my posts on the subject are much too long and tediously detailed might like it a lot. Certainly you can read it and make up your own mind.

This is where I started heaving the great sighs:

Allen believes financial markets will get back into the business of securitizing mortgage debt, but only after making some major changes. One new feature of future securitization deals, he says, could be a requirement that loan originators hold at least part of the loans they write on their books. Before the current crisis, loans were bundled into complex tranches that were passed through the financial system and onto buyers with little ability to assess the real value of the individual assets.

"The way the collateralized debt obligations (CDOs) and other vehicles are structured will change. They are too complicated," says Allen. "I'm sure the industry will figure out how to do it. There will be a lot of industry-generated reform and the industry will prosper. This is not, in my view, something that should be regulated."
This is the logic of the text (if not, perhaps, Professor Allan's logic):

1. Originators will be required to retain some credit exposure, [because?]

2. Investors have insufficient information to assess loan level risk, [because]

3. Securitization structures are too complex.

4. Therefore the industry will change these things and no regulation is needed.

Forget item four; that follows from any set of premises for some people. I just want to know how having originators (I do not think this means "security issuers" like investment banks) retain some exposure to the underlying credits will solve either the investor due-diligence problems or the complexity problems. I do not know why failing to distinguish between mortgage-backed securities and such vehicles as CDOs is being helpful in this context. I don't know why having simpler MBS structures will necessarily make investor due diligence any better: am I the only one who imagines that investor due diligence could be minimal on a plain old single-class pass-through if you waved a distracting enough coupon in front of them?

Well, the next expert brings that up:
According to Wharton finance professor Richard J. Herring, for decades, mortgage securitization was backed by government guarantees through Fannie Mae and Freddie Mac, and it worked well. Of course, these agencies were regulated and bound by less-risky underwriting standards than those that ultimately prevailed in the subprime market which was also, potentially, more profitable. Indeed, default rates were so low in the mortgage-based securities market that banks and other private financial institutions were eager to take a piece of the residential business.
Those GSE securities that worked so well for so long were quite simple in structure, and I'm still willing to bet that a lot of people invested in them without knowing much at all about the credit risk of the underlying loans. They relied on the guarantee, on the liquidity of the GSE MBS market, and on the "homogeneity" of the mortgage world back then.

And, you know, they didn't earn much. Conservative credit risk is not the sort of thing that generates princely yields. It was not, in fact, just that margins on originating subprime loans were fatter. (Even at the height of the subprime boom, subprime was never more than 20% of originations, as far as I know. Low volume and high margin.) It was that prime quality mortgage-backed securities don't have high yields, and they can't. In a "normal" market, they aren't risky enough to have juicy yield. And furthermore, everyone (more or less) has one. Including investors in MBS. There's a limit to how willing we all are to put high-yielding MBS in our retirement accounts if it means we can't afford our mortgage payments. We have met the source of the underlying cash-flow, and it is us.

The lurking concept here is "leverage." You want to make the big bucks investing in MBS? You leverage them. That's where those CDOs came from. A whole lot of this complexity is driven by the "need" to goose the yield, not by some essential opacity of the underlying credits or the failure of originators to retain residuals--which, in fact, they actually did quite a bit of in there. The complexity came in because you can't get a tranche paying 12% out of a bunch of loans that pay 8% unless you create complex cash-flow structures hedged by complex rate swaps leading to re-securitization of tranches in new vehicles (parts of the MBS become CDOs, for instance).

So are all the rest of you convinced that market participants are going to give up on the chase for mo' better yield without regulation?

Leverage does get a mention later on:
Wharton real estate professor Joseph Gyourko notes that significant differences exist in the performance of commercial and residential real estate securities. "Securitized commercial property debt will come back once the market calms down," he says, adding that there has been very little default in commercial real estate finance. "You'll be able to pool mortgages and securitize them, but almost certainly won't be able to leverage them as much as you did in the past."

The residential side, where there is significant default, is more problematic. Gyourko believes the residential market will go back to what it was in the mid-1990s and most borrowers will have to put
down at least 10% of the sales price. "We will get rid of the exotic, highly leveraged loans," he says. "That will lead to lower homeownership, but it should. We put a lot of people into homeownership that we shouldn't have."
Worry about commercial RE is just a side-effect of market tizzies? It's only individual homeowners who need to just get used to being left out of the party if they don't have the down payment? I begin to stop sighing and start to mutter . . .

Ah, but one distinguished professor has his eye on another party who could share some of the pain with the homeowner:
Wharton real estate professor Peter Linneman offers an intriguing prescription to bring prices down to the point where the industry can start to rebuild. He suggests that the government tell banks that if they want to maintain their federal insurance, they should fire their CEO by the end of the day, and the government will pay the CEO $10 million in severance. Ousting the former CEOs gives the new bank CEOs an incentive to write down all the bad assets immediately, so that any improvement will make them look good going forward. That would speed the painful process of gradual price declines.

"There's plenty of money out there waiting for these assets to be written down to bargain prices," says Linneman. In another quarter or two, the lenders would have new cash and be ready to lend again. Meanwhile, he says, the government should tell bankers it will keep interest rates down but raise them after the end of the year. "That says, 'Get your house in order in the next nine months because the subsidy ends at the end of the year.'" Linneman figures that 1,000 CEOs are accountable for about 80% of the current lending mess. If the government were to spend $10 billion to restore liquidity to the market in nine months with only 1,000 people losing their jobs, it would be the best investment it could make to restore the economy. "I'm only half-kidding," he quips.
Yeah, I'm only half-throwing up. If these CEOs are indeed "responsible" for 80% of the worst credit crisis we've seen in most people's lifetimes, and they've been doing quite well out of it, why do we need to pay them another $10 million to go away? Because it's "innovative"? Because the First Deputy Assistant CEO waiting in the wings has a whole nuther plan we haven't heard yet for some reason, but they will pipe up with it as soon as the Guilty 1000 are gone? And it's going to be about how to wean themselves off the leveraged carry-trade in nine months? Maybe that plan would be worth $10 million, but do we get the money back if we aren't satisfied?

I am trying to avoid suggesting that maybe we'd get better CEOs out of better business schools. I'm not trying that hard, but I'm trying. As is quite often the case, I never know if some of these things sound so half-baked because of the writing--the urge to simplify complex ideas into palatable chunks--or because of the paucity of the underlying material. But I'm happy to blame a lot of the bad writing I see on what they teach in business schools.

There is some sense in here--but you'll have to go dig it out yourself. I've sighed so much I need a good drink. I certainly agree that the CDO is dead. So, probably, is the SIV. I'm still wondering about the multi-class structured MBS, but that doesn't seem to be the article's particular interest.

(Hat tip Bill & Bill)