The enormous mortgage-bond scandal
You thought the foreclosure mess was bad? You’re right about that. But it gets so much worse once you start adding in a whole bunch of parallel messes in the world of mortgage bonds. For instance, as Tracy Alloway says, mortgage-bond documentation generally says that if more than a minuscule proportion of notes in a mortgage pool weren’t properly transferred, then the trustee for the bondholders can force the investment bank who put the deal together to repurchase the mortgages. And it’s looking very much as though none of the notes were properly transferred.
But that’s not even the biggest potential problem facing the investment banks who put these deals together. It also turns out that there’s a pretty strong case that they lied to the investors in many if not most of these deals.
I mentioned this back in September, and I’ve been doing a bit more digging since then. And I’m increasingly convinced that the risk to investment banks isn’t only one of dodgy paperwork; there’s also a serious risk of massive lawsuits from the SEC or other prosecutors, as well as suits from individual mortgage investors.
The key firm here is Clayton Holdings, a company which was hired by various investment banks — Goldman Sachs, Bear Stearns, Citigroup, Merrill Lynch, Lehman Brothers, Morgan Stanley, Deutsche Bank, everyone — to taste-test the mortgage pools they were buying from originators.
Here’s how it would work:
First, the bank would put in a winning bid for the pool of mortgages, with the intention of slicing it up into mortgage bonds and selling those bonds off to investors at a profit.
After submitting the winning bid, the bank would commission Clayton to take a closer look at a representative sample of loans in the pool. Clayton controlled as much as 70% of the market for this service, which is known as third-party due diligence. But Clayton’s not at fault here, and the problem is likely to apply no matter who performed this service.
The size of the representative sample would vary according to the size of the loan pool; it could be anywhere between 5% and 35% of the loans in the pool. Essentially, Clayton would go back to the loans, one by one, and re-underwrite them after the fact, checking that the originator’s underwriting standards were in fact being upheld.
Clayton would either accept or reject the loans it was looking at, according to whether or not they met underwriting standards. Here’s the results of what it found for one bank, Citigroup; the chart comes from this document filed with the Financial Crisis Inquiry Commission. I’m just using Citi as an example, here; all banks behaved in basically exactly the same way.
Look at the first line. Clayton reviewed 1,280 loans on behalf of Citigroup in the first quarter of 2006. Of those, it accepted 554 outright: they lived up to the originator’s underwriting standards. It also waived another 144, on the grounds that there were mitigating factors (a large downpayment, say). And it rejected 582 for a rejection rate of 45%.
This kind of information was valuable to Citigroup: it showed them that the quality of the loan pool was much lower than you’d think just by looking at the ostensible underwriting standards.
Armed with this information, Citigroup would do two things. First of all, it would take those 582 rejects and put most of them back to the underwriter. Essentially, they said, the loans weren’t as advertised, and they didn’t want them. But Citi would still keep some of them in the pool.
But remember that Clayton had tested only a small portion of the loans in the pool. So Citi knew that if there were a bunch of bad loans among the loans that Clayton tested, there were bound to be even more bad loans among the loans that Clayton had not tested. And those loans it couldn’t put back to the originator, because Citi didn’t know exactly which loans they were.
If there had been any common sense in the investment banks, that would have been the end of the deal. But there wasn’t. Rather than simply telling the originator that its loan pool wasn’t good enough, the investment banks would instead renegotiate the amount of money they were paying for the pool.
This is where things get positively evil. The investment banks didn’t mind buying up loans they knew were bad, because they considered themselves to be in the moving business rather than the storage business. They weren’t going to hold on to the loans: they were just going to package them up and sell them on to some buy-side sucker.
In fact, the banks had an incentive to buy loans they knew were bad. Because when the loans proved to be bad, the banks could go back to the originator and get a discount on the amount of money they were paying for the pool. And the less money they paid for the pool, the more profit they could make when they turned it into mortgage bonds and sold it off to investors.
Now here’s the scandal: the investors were never informed of the results of Clayton’s test. The investment banks were perfectly happy to ask for a discount on the loans when they found out how badly-underwritten the loan pool was. But they didn’t pass that discount on to investors, who were kept in the dark about that fact.
I talked to one underwriting bank — not Citi — which claimed that investors were told that the due diligence had been done: on page 48 of the prospectus, there’s language about how the underwriter had done an “underwriting guideline review”, although there’s nothing specifically about hiring a company to re-underwrite a large chunk of the loans in the pool, and report back on whether they met the originator’s standards.
In any case, it’s clear that the banks had price-sensitive information on the quality of the loan pool which they failed to pass on to investors in that pool. That’s a lie of omission, and if I was one of the investors in one of these pools, I’d be inclined to sue for my money back. Prosecutors, too, are reportedly looking at these deals, and I can’t imagine they’ll like what they find.
The bank I talked to didn’t even attempt to excuse its behavior. It just said that Clayton’s taste-testing was being done by the bank — the buyer of the loan portfolio — rather than being done on behalf of bond investors. Well, yes. That’s the whole problem. The bank was essentially trading on inside information about the loan pool: buying it low (negotiating for a discount from the originator) and then selling it high to people who didn’t have that crucial information.
This whole scandal has nothing to do with the foreclosure mess, but it certainly complicates matters. It’s going to be a very long time, I think, before the banking system is going to be free and clear of the nightmare it created during the boom.
Update: KidDynamite asks a good question in the comments: were the bond investors able to do their own due diligence on the loan pool? The answer is no, they weren’t — the prospectus did not include the kind of loan-level information which would enable them to do that.