Monday, November 22, 2010
Last week the US Bankruptcy Court for the District of New Jersey issued an opinion in a case captioned Kemp v. Countrywide Home Loans, Inc. Case 08-02448-JHW. This case looks like the first piece of evidence in what might turn out to be the Securitization Fail or, in homage to Michael Lewis, The Big Fail.
Briefly, Countrywide as servicer filed a proof of claim for a mortgage in a bankruptcy case on behalf of Bank of New York as trustee for a securitization trust. The bankruptcy court denied the claim because there was no evidence that Bank of New York ever owned the mortgage. The mortgage note had never been negotiated or delivered to Bank of New York, despite the requirement to do so in the Pooling and Servicing Agreement (PSA) that governed the securitization of the loan. That meant that Bank of New York as trustee had no interest in the loan, so the proof of claim filed on its behalf was disallowed.
This opinion could turn out to be incredibly important. It provides a critical evidence for the argument that many securitization transactions simply failed to be effective because non-compliance with the terms of the transaction: failure to properly transfer the mortgage meant that the mortgages were never actually securitized. The rest of this post explains the chain of title issue in mortgage securitizations and how Kemp fits into the issue.
Note and Mortgage Transfers in Securitizations
A residential mortgage securitization is a transaction that involves a series of transfers of two types of documents: mortgage notes (the IOUs made by mortgage borrowers) and mortgages (the security instrument that says the lender may foreclose on the house if the borrower defaults on the note). Ultimately, both the notes and mortgages need to be properly transferred to a trust that will pay for them by issuing securities (backed by the mortgages and notes, hence residential mortgage-backed securities or RMBS). If the notes and mortgages aren't properly transferred to the trust, then the securities that the trust issues aren't mortgage-backed and are worthless.
So the critical issue here is whether the notes and mortgages were properly transferred to the securitization trusts. To determine this, we need to figure out two things. First, what is the proper method for transferring the notes and mortgages, and second, whether that method was followed. For this post, I'm going to focus solely on the notes. There are issues with the mortgages too, but that gets much, more complicated and doesn't directly connect with Kemp.
1. How Do You Transfer a Note?
A. The American Securitization Forum's Argument (ASF white paper)
The American Securitization Forum (ASF) has a recent white paper that purports to explain how notes and mortgages are transferred in a securitization transaction. The white paper explains that there are two methods for transfer and that either can suffice, although typically both are used. Those methods are a negotiation of the notes per Article 3 of the Uniform Commercial Code (UCC) and a sale of the notes per Article 9 of the UCC (take a look at the definitions of security interest, debtor, and secured party to understand how UCC 9-203 functions to effect a sale). (The ASF argues that the mortgage follows the note, meaning that a transfer of the mortgage effects a transfer of the note. I've got my doubts on this too, but that's for another time.)
B. Trust Law and the UCC Permit Parties to Contract for a More Rigorous Method
The ASF white paper is correct to the extent that is explaining how notes could be transferred from, say, me to you or from Citi to Chase. But that's not what happens with a securitization. A securitization involves a transfer to a trust, and that complicates things.
It's axiomatic that a trust's powers are limited to those set forth in the documents that create the trust. In the case of RMBS, that document is the Pooling and Servicing Agreement (PSA). Most PSAs are governed by NY law, which provides that a transaction beyond the authority of the trust documents is void, meaning it is ineffective.
PSAs typically set forth a very specific method of transferring the notes (and mortgages) that goes beyond what is required by Articles 3 or 9. This is perfectly fine under the UCC, which permits parties to deviate from its default rules by agreement (UCC 1-203), which can be inferred from the parties' conduct, including the PSA itself. So what this means is that if a securitization transaction did not meet the requirements of the PSA, it is void, regardless of whether it complied with the transfer requirements of Article 3 or Article 9. The private law of the PSA, not Article 3 or Article 9, is the relevant law governing the final transfer in a securitization transaction.
There is some variation among PSAs, but typically a PSA will have two relevant transfer provisions. First, it will have a recital stating that the notes (and mortgages) are "hereby" transferred to the trust. This language basically tracks the requirements of an Article 9 sale. Second, it will have a provision stating that in connection with that transfer, there will be delivered to the trust the original notes, each containing a complete chain of endorsements that show the ownership history of the loan and a final endorsement in blank. The endorsement requirement invokes an Article 3 transfer, but it imposes requirements (the complete chain of endorsements and the form of the final endorsement) that are not contained in Article 3.
There is a very good business reason for having the full chain of title in the endorsements: it is evidence of the transfers needed to ensure the bankruptcy remoteness of the trusts' assets. Bankruptcy remoteness means that the RMBS investors are assuming only the credit risk on the mortgages, not the credit risk of the originators and/or securitizers of the mortgages, and RMBS are priced based on this expectation.
It is also clear that historically the method of transfer for RMBS securitizations was endorsement, not recital of sale. The promissory note sales provisions of Article 9 only went into effect in 2001 (in 49 states). Pre-2001 PSAs contain the sale language, however, as post-2001 PSAs. This indicates that the "hereby" language is really carryover boilerplate; in 2001, it was ineffective to transfer a note under Article 9 of the UCC. While that language might have been sufficient for a common law sale, it wouldn't work for a transfer to a trust under NY law. When assets are transferred to a NY trust, there has to be actual delivery in as perfect a manner as possible; a "mere recital" doesn't cut it (and frankly, endorsements in blank might not suffice either because there is nothing that indicates that something endorsed in blank is trust property, rather than the trustee's or someone else's).
2. Was There Compliance with the Trust Documents?
So to tie this all back to Kemp: the note in Kemp lacked the endorsements required in the PSA. That means, as the Bankruptcy Court concluded, that the note was never transferred to the trust at the time the bankruptcy claim was filed. The Bankruptcy Court did not need to opine beyond that point, but it is a small step to recognizing that if the loan wasn't transferred to the trust in the first place, it cannot be transferred now. PSAs contain numerous timeliness provisions about loan transfers, often related to ensuring favorable tax status for the trust. PSAs also require the transferred loans be performing (not in default). That means that for the securitization trust, the Kemp note is like caffeine in 7-up: never had it, never will. The securitization of the Kemp note failed.
Now here's the real kicker: there's no reason to think that the Kemp note was a unique, one-off problem. All evidence from actual foreclosure cases points to the lack of a chain of endorsements on the Kemp note being not the exception, but the rule, and not just for Countrywide, but industry-wide. Certainly on the non-delivery point (separate from the non-endorsement problem), Countrywide admitted that non-delivery was "customary." If either of these issues, non-delivery or non-endorsement is widespread, then I think we've got a massive problem in our financial system.
3. Implications for Various Parties
Below, a review of the implications for several types of parties.
Federal bank regulators should be all over this; there is monstrous systemic risk potential. The new Financial Stability Oversight Counsel, as well as the OCC and the Fed and FDIC should all be doing very targeted examinations of the large trustee banks' collateral files to grasp the scope of the problem. I don't know what they're actually doing, but I'm afraid that they aren't undertaking the proper investigation. Fortunately, this particular issue is easily within the expertise of bank regulators: just go to the collateral files and start looking at a large sample of notes. See how many are missing complete chains of endorsement or lack signatures altogether. That will be a very quick way to tell if there is a problem.
I'm also very concerned that some banks might decide to start filing in chains of endorsement and backdating. But that's fraudulent, you protest! Surely no bank would ever engage in fraud! Of course backdating signatures is fraudulent, but if the signatures aren't there, the banks are dead, so there's really no downside in having some underlings fill in their signatures. If caught there likelihood of jail time is low. Why not bet the farm? Bank regulators should be very sensitive to this potential problem. They should insist on being the ones who actually select the collateral files to be reviewed and that they are the ones who pull the actual note out of the file. The examiners should be making digital images of all notes that they review and keeping those for potential examination against the actual notes if those notes are produced in future foreclosure cases.
My concern here is that the bank regulators so badly don't want for there to be a problem that they won't look at the notes in the hopes that this issue goes away. I hope that they are sensible enough to know that if there is a problem, they cannot prevent it, and would do best by gathering up all the information they can.
SEC and Accountants
If the mortgages weren't properly transferred, there could be a variety of securities law violations, including servicers' regular Reg AB attestations. There could also be securities law violations on behalf of the banks--if the assets weren't properly transferred, they are still on the banks' balance sheets (as are the losses) and should be accounted for as such.
The ratings agencies should be all over this issue. It goes to the question of whether the collateral backing the MBS is there and whether the representations made to them about deals was in fact correct. I have heard, but cannot verify, that ratings agencies were themselves able to inspect the actual notes. If so, then there is a real conflict of interest on this point, as they should have caught this facially obvious problem. Unfortunately, the materials I've seen coming out of some of the ratings agencies make me concerned that they simply don't understand the legal issue involved and may not even understand the difference between the note and the mortgage.
Banks (Securitization Sponsors)
The banks are in serious trouble if there are widespread securitization fails. If the loans weren't transferred to the securitization trusts, then they are on bank balance sheets, which means that (1) the losses on the loans are the banks (to be sorted out with the investors), and (2) the banks need to be holding capital against the loans that haven't gone into foreclosure. Depending on the scale of the problem, the banks might not have enough capital to cover the securitization fails, which means we're in Dodd-Frank resolution territory.
If the notes weren't properly transferred to the trusts, then investors have the mother of all putback claims. Investors probably also have claims against securitization trustees and against the law firms that did diligence on the securitization deals. (Note that these same firms are the ones lining up to swear that there isn't a problem....). Of course, the danger for investors is that there is a huge problem, and the banks lack the money to fix it.
Let's be clear that investor interests here are split. AAA investors who are still well in the money would prefer to simply be paid out on their RMBS at 100 cents on the dollar than mess with putbacks. But mezzanine (like CDOs) and junior investors have a lot of potential upside here.
This could be very awkward for the monolines. Generally they promise timely payment of principal and interest to investors. If that coverage obligation continues while the monolines make rescission claims, they might have to pay out of pocket first and then look to the banks for recovery. If so, they would be in a heck of a liquidity pickle.
Chain of title doesn't affect whether homeowners are in default on their loans. The loans' validity is not in question because of chain of title. But chain of title does affect who has the right to foreclose. At the very least, if there is a chain of title problem, it means lots of foreclosures cannot properly proceed because of lack of standing. On the other hand, if the loans weren't actually securitized, they are on banks' books, which might, just might, facilitate workouts. More generally, if there is a widespread securitization fail, it means that there will have to be a legislative solution to the problem, which might facilitate real loan modifications.