Why using eminent domain for liens is a bad idea

A couple of weeks ago, Matt Goldstein and Jenn Ablan had an intriguing story: Mortgage Resolution Partners, a politically well-connected firm in San Francisco, was shopping to municipalities the idea of using eminent domain to restructure mortgages. Then, on Tuesday, Cornell University’s Robert Hockett weighed in, saying that the idea was a compelling one. “To solve a collective action problem, we need a collective agent,” he wrote. “That’s what governments are.”

According to Imran Ghori of the Press-Enterprise in San Bernadino, where the idea seems to be furthest along, Hockett “has been working with Mortgage Resolution Partners” but “said he has no financial interest in the proposal”. I don’t really know what that means, but I think it’s fair to assume that if this happens, Hockett is very well placed to make a lot of money from it. So it’s worth approaching the idea with a skeptical eye.

In principle, I think I can like this idea. On Monday I met with Jorge Newbery of American Homeowner Preservation, whom I’ve written about a few times in the past; his company buys pools of defaulted underwater mortgages from banks, often for just $1 each, and then, having bought the mortgages at massive discounts to par value, can come up with any number of ways to successfully modify the mortgage, nearly all of which involve principal reduction. This is a very successful outcome for nearly everybody involved, but there’s a problem: while Newbery can buy pools of bank-owned mortgages, he can’t buy mortgages which have been securitized. And those mortgages represent the vast majority of defaulted subprime debt.

Newbery started buying pools of mortgages when his original idea didn’t work. That idea was elegant: investors would buy a house in a short sale at the market price, and then lease the home back to the homeowner until the homeowner had the ability to get a mortgage and buy it back at a pre-set price. The idea might have been elegant, but it didn’t work in practice, because the banks wouldn’t play ball: they (and Freddie Mac) simply hated the idea of a homeowner being able to stay in their house after a short sale, and often asked for an affidavit from the buyer saying that the former owner would certainly be kicked out.

The idea from Mortgage Resolution Partners and Robert Hockett basically does an end-run around the banks’ objections: they can’t object to the short sale, because they’re being forced to do the short sale. Clever.

But then things become extremely murky. Here’s Hockett:

Using their traditional eminent domain authority, municipalities can “take” – it’s a constitutional term of art – underwater mortgages from holders for fair market value. They can then write down the loans to just under the values of underlying homes, bringing these back above water. They can finance these takings with moneys supplied by investors, who then are repaid on the refinanced mortgages.

Got that? Me neither. Here’s Goldstein and Ablan, trying to explain the same idea:

Mortgage Resolution Partners would work with local governments to find institutional investors willing to provide tens of billions of dollars to finance the condemnation process to avoid using taxpayer dollars to acquire millions of distressed mortgages.

A local government entity takes title to the loans and pays the original mortgage owner the fair value with the money provided by institutional investors.

Mortgage Resolution Partners works to restructure the loans, enabling stressed homeowners to reduce their monthly mortgage payments. The restructured loans could then be sold to hedge funds, pension funds and other institutional investors with the proceeds paying back the outside financiers.

The key here — which is spelled out in much more detail in Hockett’s 56-page paper on the idea — is that the eminent domain powers are not being used to buy the actual houses in a short sale, as would have been the case under the original AHP scheme. Instead, they’re being used to buy the mortgage. Hockett doesn’t spend any time in his paper or his op-ed explaining why eminent domain should be used to buy mortgages rather than houses, and it’s here, I think, that his plan moves from something which could be a very good idea, to being something which is actually a pretty bad idea.

Here’s how a scheme like this should work. MRP, or a company like it, borrows short-dated money for a term of say three months at very low interest rates. Meanwhile, underwater homeowners in San Bernadino are invited to volunteer for the scheme. Once the money has been raised, it is used to buy those homeowners’ houses at the market rate. The homeowners then buy their houses back from MRP at say a 2% premium to the price paid, using a mortgage given to them either by MRP itself or by some other lender. MRP then repays the short-term loan. MRP’s profits come from that 2% premium, and from its separate mortgage-lending arm; if it wants, it can restrict the houses it buys to only the ones owned by people it would be willing to lend money to.

Under this scheme, the banks or investors who hold the mortgage would receive in return the fair market value of the home in question, just as they would in a short sale. That’s a very reasonable amount to receive for an underwater mortgage, so the banks can’t really complain. MRP would make a modest amount as a middleman and facilitator, and the homeowner would end up with a house mortgaged at its fair market value, rather than at some inflated old purchase price.

But that’s not what Hockett is proposing. Instead, Hockett wants MRP to be able to buy the mortgage, rather than the house. That’s very weird: while it might be legal under eminent-domain law (I have no idea about that), the spirit of the law is very much that the government can buy property, rather than liens. But after talking to Newbery, I understand what MRP is thinking here. His company, AHP, is buying up underwater mortgages for much less than the value of the underlying property — sometimes only 10% or so, and never more than about 25%. Admittedly, all of those mortgages have been in default for some time. But MRP clearly wants to be able to buy mortgages at a deep discount to the value of the home, and then “restructure” the mortgage so that the principal amount is very close to the value of the home. The result could be massive profits for MRP.

In both cases, the homeowner essentially ends up refinancing into a new mortgage with a principal amount just below the value of the home. But in the first case that money is essentially used to pay off the old mortgage holder, while in the second case — the one MRP is proposing — the money goes largely to MRP, the middleman.

What’s more, the market for liens is much more opaque than the market for houses, and as such MRP could probably make a colorable case that fair value for the mortgages it wanted to buy was extremely low. Since MRP would have all the important political relationships, the owners of the mortgage — especially if they’re just distant bondholders somewhere — would have very little ability to contest the valuation, and might end up getting paid much less than a genuinely reasonable price for it.

It seems to me that MRP is not adding a huge amount of value here — certainly nothing commensurate with the amount of money it’s likely to make. The real value is added by the use of eminent domain to buy the liens, and it’s the municipal government, rather than MRP, which has that power. So if anybody makes money from using eminent domain, it should be taxpayers: not some private-sector middleman.

If I represented the municipality of San Bernadino, I would respond to MRP’s proposal by giving them two choices. Either cut the city in to a very large proportion of MRP’s profits on these deals, or else force MRP to buy houses rather than liens. Both of those options seem fair to me. Hockett’s scheme, as it stands, doesn’t.