In response, some people (including one of the country’s top bankruptcy lawyers) have told me they don’t buy it.
Specifically, they ask such questions as:
- With a mortgage sold to two different entities, wouldn’t the income from the mortgage be shown on the books of both entities?
- Was the interest/principal payments that were made by the homeowner before they stopped being divided between both entities? If so, wouldn’t this have rung alarm bells immediately?
- If only one was getting it, why didn’t the other entity immediately try to foreclose?
- If there was one servicer involved, was the servicer covering the difference between what was collected and the payments actually made? If so, how did the servicer do this and still remain in business?
- If two servicers were involved, why didn’t this come out sooner or were both servicers hiding this fraud?
So I wrote to some of the leading experts on mortgage fraud – L. Randall Wray (economics professor), Christopher Whalen (banking expert with Institutional Risk Analytics), and William K. Black (professor of economics and law, and the senior regulator during the S & L crisis) – to seek their insight.
Chris Whalen told me:
All good points, but the short answer is that nobody may have noticed until now. The issue of substitution and other games played by servicers makes exact tracking of loans problematic. It should show up in the servicers reports and should be caught, but there are a lot of things that go on in loan servicing that nobody talks about. Until about 2006, the GSEs and banks would advance cash and would substitute, but not now. The noble practitioners you heard from are all sincere and want to believe in intelligent design.
Prior to FAS [i.e. Financial Accounting Standards] 166/167, a defaulted loan might sit in a FNM/FRE pool for up to a year before the default was removed from the trust. The issuer would then place a new loan into the pool or “substitute” for the old loan. No purchase event was booked. The investor would never know. In fact, the issuer would keep paying interest on the original principal amount in those days. Now under FAS 166/167, the issuer must immediately repurchase the defaulted loan and take the loss less estimated recovery. That is why the pace picked up this year when it comes to repurchase demands.
You should refer your dubious and very naive friends to the case of National Bank of Keystone, WV. One of the worst failures per $ of assets in FDIC history. The management hid a Ponzi scheme in the loan servicing area for five years. Paid interest to investors with their own principal. Two auditors missed the fraud and later were sued by the FDIC acting as receiver for the dead bank. And this was a small operation. The big five are an even worse mess. Remember, when the seller of a loan and the servicer are the same, anything can happen. And it usually does.
Professor Black told me:
Double pledges (as they’re typically called, though one could pledge multiple times) are a well known fraud device. It is correct that one of the key purposes of adopting Article 9 of the Uniform Commercial Code (UCC) was to reduce the risk and frequency of this form of fraud. So, double pledges in the modern era require both (A) fraud (on the part of the borrower or purchaser) and incompetence, indifference, or corruption on the part of the original secured lender or their agents if the borrower is the fraudster or the purchasers if they are the fraudsters.
The two potential sources of fraud: A fraudulent borrower could pledge the same home as security for multiple mortgage loans. Title checks, by the lender/title insurer are so easy to conduct and so vital to protect the lender that this form of fraud is vanishingly rare. Alternatively, and far more likely, the lender could sell the mortgage to multiple buyers. Those buyers could have far lower incentives to check on prior pledges and less ability to check for prior pledges. The entity selling a loan to multiple parties (A) has a compelling incentive to hide the prior pledge(s), (B) is financially sophisticated, and
therefore more capable of deception than a homeowner, and (C) can pick who to make the multiple sales to — allowing them to select the most vulnerable targets for fraud.
Subpart (C) provides the logical transition to the second requisite for multiple pledge frauds — vulnerable victims. The characteristics they would exhibit include (A) growing massively, (B) purchasing nonprime loans without fully underwriting the quality of the loans (and quality in this context inherently requires superb “paperwork”), (C) poor internal and external controls, and (D) opaque systems that make it extremely difficult to determine the beneficial owner and locate key mortgage documents that would reveal multiple sales. Unfortunately, these four characteristics were characteristic of many purchasers of nonprime mortgages. That is why I have long stated that the process was dominated by the financial sector equivalent of “don’t ask; don’t tell.”
Bottom line: the elite bankers and the anti-regulators have been so unwilling to
find the truth that no one knows how bad these frauds became. Finding the facts
is essential and can and should be done by reviewing samples of the loans pledged or sold to Fannie and Freddie and the Fed.
And professor Wray told me that record-keeping by servicers was terrible, and pointed me to the following article from the Tampa Tribune:
Peter Bakowski, a 58-year-old former Tampa mortgage broker, has admitted orchestrating a Ponzi scheme that involved more than 30 investors and institutions and more than 150 deals, documents show.
Bakowski sold the mortgage assignments to multiple investors, promising high rates of return and using all the money he generated to “keep the scheme afloat,” according to his plea agreement.