Why the eminent-domain plan doesn’t hurt second liens
Brad Miller, arguably the most sophisticated and well-informed member of the House when it comes to housing finance issues, has an op-ed in American Banker today about the eminent-domain plan being mooted in San Bernadino (which just voted to file for bankruptcy, by the way). Miller’s excited about the plan, because he thinks that it will force banks to take losses on all-but-worthless second liens. But, sadly, he’s wrong about that.
Miller actually makes two mistakes in his piece. The first comes when he explains how the price of the mortgages would be determined:
Deciding a fair price would not be hard. There are frequent auctions of mortgages with a sufficient number of informed, sophisticated buyers. The auctions are an almost perfect pricing mechanism. There would be comparable sales to determine almost any mortgage’s fair market value.
Miller’s right that mortgages do get auctioned relatively frequently, if not frequently enough that the market can even be considered highly liquid, let alone “an almost perfect pricing mechanism”. But here’s the thing: the private-label mortgages which tend to get sold off are precisely the mortgages that Mortgage Resolution Partners does not want to buy — the ones in default. Banks which own performing mortgages have almost no incentive to ever auction them off. And MRP has said that performing mortgages are the only mortgages it’s interested in.
What’s more, when performing underwater mortgages are traded, they’re often sold above par, since the homeowner is locked in to higher-than-prevailing mortgage rates. MRP, by contrast, is determined that it will only buy mortgages well below par: indeed, they’re saying that they’ll demand a discount not only to the face value of the mortgage, but even to the market value of the property. As a result, deciding a fair price might well be completely impossible: the owners of the mortgage would value it as a performing loan at a high rate of interest, while MRP would essentially ignore the fact that it’s performing, and value it on the basis that it cannot be worth more than the value of the collateral.
A free market copes quite easily with huge valuation discrepancies like that: there’s simply no trade, and the owner of the mortgage holds onto it, while companies like MRP find themselves unable to offer a price at which anybody is willing to sell. That’s why MRP’s whole idea is contingent on doing an end-run around the free market, and forcing the owners of the mortgage to sell. The point here is that if there really was a low market-determined fair price for the mortgages, then MRP wouldn’t need eminent domain at all: it could simply buy up those mortgages on the free market, directly from banks. Maybe, eventually, once it ran out of free-market mortgages to buy, MRP could try to use the eminent-domain method to buy mortgages from CDOs and MBSs. But at that point they’d have real-world market-based proof of how much such mortgages were worth.
MRP isn’t going down that road, however, because it knows that no one will voluntarily sell them mortgages at the kind of discounts it’s looking for. Which is prima facie evidence that the amount it’s willing to pay is not a fair price after all.
Miller then moves on to the thorny issue of second liens. While first liens are often owned by special-purpose investment vehicles, second liens are generally owned by banks. Miller writes:
So the real losers from the program would be the biggest banks, the holders of second liens, not investors in first mortgages. And even for the biggest banks, eminent domain would not cause losses but reveal losses.
But this isn’t true. If anything, the holders of the second liens would make money from this scheme, rather than losing money. Remember that MRP is not planning to buy houses using eminent domain, which would make much more sense. Instead, it’s only planning to buy mortgages — and it’s refusing to buy any mortgages held directly by banks. Instead, it’s only buying mortgages held by special-purpose investment vehicles, which tend not to have expensive lawyers willing and able to contest any and every valuation.
“Involuntary sales of seconds at fair market value would end fictitious valuations and require an immediate accounting loss,” writes Miller — and he’s right about that. Sadly, there’s nothing in the MRP plan which suggests that MRP has any interest at all in buying up second liens from banks. If MRP were buying houses rather than mortgages, then the banks holding the second liens would be forced to take an immediate loss. But it’s not. Instead, it’s just buying up performing first liens, and leaving everything else intact, including all second liens. At the margin, then, by reducing the amount of money that homeowners owe on their first liens, the MRP plan will increase, rather than decrease, the value of the second liens.
Why won’t MRP buy up second liens at what it considers to be a fair market value? For the same reason that it won’t buy up first liens directly from banks, either — the two sides will never be able to agree on a price. By MRP’s calculation, if the first lien is worth much less than par, then the second lien has to be worth something very close to zero. By the bank’s calculation, on the other hand, a performing second lien is a valuable revenue stream, worth a significant amount of money. And because MRP will be willing to pay very, very little for that second lien, it will not be willing to spend a substantial amount of money defending that near-zero valuation in court: its legal fees would almost certainly be greater than the amount it was willing to pay for the second lien in the first place.
If Brad Miller can point me to a plan where eminent domain is used to buy underwater houses rather than underwater mortgages, or if he can point me to a plan where eminent domain is used to by delinquent mortgages rather than performing ones, including seconds — then he’ll have me persuaded. But sadly the plan on the table is not the plan that Miller thinks it is. Which is why it’s a bad plan.