Incremental housing help

The details of the foreclosure settlement this morning mostly reflected what had been reported late last night, and here was the final take from the FT’s Shahien Nasiripour and Kara Scannell:

Under the agreement, Bank of America, Wells Fargo, JPMorgan Chase, Citigroup and Ally Financial will be forced to improve their mortgage procedures, reduce borrowers’ loan balances and monthly payments, and make about $4.2bn in cash payments to an estimated 750,000 aggrieved homeowners and state governments. …

The mortgage component of the proposed deal involves about $3bn in benefits for borrowers who would refinance out of high-cost loans into cheaper mortgages, and a system of roughly $17bn in so-called “credits” that would be earned in relation to the loan principal reduced by the banks as well as other types of payment relief. …

The US Department of Housing and Urban Development forecasts the credit system could yield more than $32.3bn in lower payments and balances for borrowers, bringing the estimated total benefit to $39.5bn. About 1m borrowers are forecast to benefit.

This will surely be weak beer to the 750,000 people foreclosed on because of improper servicing procedures and who will get a $2,000 check, but we suppose it’s better than nothing and about what had been expected for a while.

But it’s especially worth noting (as most commentary has) that the banks aren’t in the clear on all foreclosure-related issues. The Huffington Post reported last week that the scope of the settlement would be mostly limited to robo-signing abuses, and so it is. This was a necessary part of convincing the California and New York attorneys-general, each of which is pursuing other legal options against the banks related to securitisation practices, to sign on to the agreement.

As to the impact on the housing market, Goldman Sachs economists yesterday published one of the more helpful overviews we’ve come across about the many housing-related initiatives either now being implemented or recently proposed by the Obama administration.

The economists estimate that taken together, they will reduce the flow of distressed properties to the market by up to 800,000 homes in the next two years, boosting house prices by 0.8 per cent:

This could occur through a combination of (1) a bulk REO sale program, (2) additional federal financial incentives for mortgage modification, and (3) a potential legal settlement by large mortgage servicers involving mortgage modification targets. A reduction of this magnitude in vacant units for sale could boost house prices by 0.8% over the next two years, all other things equal.

Refinancing activity should also increase as a result of the policy moves, though most of the incremental activity we expect is due to policies announced last year, rather than new initiatives. While helpful to the borrowers who are able to refinance as a result, this seems likely to have modest macroeconomic effects.

Today’s settlement takes care of the third item, though Goldman economists calculated that HUD’s estimate of one million homeowners getting a principal writedown is too high:

The Dept. ofHousing and Urban Development (HUD) has estimated that around 1 million borrowers could be offered a write-down on their mortgage principal. This seems high, as it implies an average write-down of only $35,000–assuming a 2:1 ratio of total write-down to penalty payment–compared to an average write-down in the HAMP program–which until now has had less generous incentive payments–of $67,000. If we assume the latter per-loan writedown amount, this would imply 520,000 borrowers would receive writedowns.

On the bulk REO (real-estate owned) sale program, the FHFA and Treasury started thinking about the best way to do this last summer. The idea is to dispose of the REO portfolios of Fannie, Freddie and the FHA — including, especially, the idea of selling the properties to investors looking to rent the units.

We were sceptical at the time that much would come out of this, but Goldman has an update on recent progress:

Other programs, such as the recently announced bulk sales of REO properties to investors for rental, are still in the startup phase. The effects of these programs are likely to become evident in Q2 and Q3…

In prior work on a potential REO-to-Rental program, we noted that the GSEs and the FHA have been selling nearly 500,000 distressed properties off their balance sheets at an annualised rate over the last few quarters. However, we estimate the location or condition of the properties likely reduces the number that could feasibly be sold in bulk to investors for rental to less than half the total.

Also, it seems likely that in some areas investors would have already converted a significant number of properties to rentals. We therefore estimate that the REO-to-rental program is likely to increase rental supply by between 20% and 40% of the distressed inventory, or 200,000 to 400,000 over 2012 and 2013.

As for item number 2, referring to the added incentives for mortgage modifications, the economists are referring to recent changes to the disastrously inadequate Hamp. The excerpt from the Goldman note is long but worth reproducing in full, as it includes a useful breakdown of the numbers involved:

The administration has announced two sets of changes to the HAMP program: (1) a tripling of the financial incentives (to 18 to 63 cents for every dollar of principal reduction) for servicers to provide principal reductions rather than other modifications that rely solely on reduced monthly payments, and (2) the eligibility of investor-owned properties and loosening of minimum debt-to-income ratios.

The number of loans likely to be modified under the latest version of the program will be determined by one of two constraints: the funding available for the program or the number of eligible loans. We expect the available funding to be the binding constraint, and thus base our estimate on the funds available. As of January 27, there was $29.9 billion in funding from the TARP program allocated to the HAMP program (since legal authority to purchase assets under the TARP program has expired, the amount cannot be increased). Of that total, it appears $17 billion remains available for modifications. At a current per-modification cost of almost 15,000, this implies an additional 1.3 million permanent modifications.

However, at the current annual rate of modifications, it would take three to four years to exhaust the available funds. Instead, it seems likely that in light of new principal write down incentives, HAMP funds will be exhausted before all eligible loans have been modified. Current writedowns average $67,000 in loan principal, at a cost to the Treasury of $25,000.

We estimate the new incentives would bring the cost per principal reduction to around $45,000. If modification activity shifted entirely to principal writedowns–an unlikely outcome–this would imply around 400,000 loans could be modified with the remaining funds available. We think a reasonable assumption is that somewhere between 20% and 50% of modifications will occur through principal reductions, implying permanent modifications of 600,000 to 800,000 loans under the program, most of which are likely to occur over the next two years.

Of course, such estimates of the earlier version proved wildly optimistic, so we’ll see.

On refinancing measures, we weren’t impressed by Harp 2.0 — not because it was a bad idea in itself, but because it didn’t seem to go very far. The FHFA’s own estimate was that it would merely continue the pace of refinancings we already had under the first edition. That said, we could well be wrong and a lot’s happened since then — mortgage rates have remained low, and it appears that refinancing activity has lately ticked up. Calculated Risk expects refinancings to further increase in the coming months, and he’s forgotten more about the housing market than we’ve ever known.

Here’s a matrix of where today’s settlement fits in with everything else going on in housing policy (click to enlarge):

You’ll notice that we’ve ignored those items with an X under “requires legislation”. This includes the Administration’s proposal to expand the range of people with GSE-backed loans eligibile for refinancing. We’re still unclear, to be honest, as to why this requires an act of Congress. (From everything we’ve read, the FHFA could do this on its own.) Regardless these proposals are unlikely to pass.

So how might this all of this lead to a mild boost to house prices this year and next? Here’s how Goldman closes the note:

Indirect benefits of reduced vacant units. Taking the mid-points of our estimates above implies a total reduction in distressed supply on the for-sale market of as much as much as 800,000 units over the next two years if we assume (1) HAMP loan modifications have a 30% redefault rate, (2) servicing settlement loan modifications have a 50% redefault rate, and (3) half of these redefaulted loans result in foreclosures and REO sales. Last year we introduced a house price model that, among other inputs, relies on vacant units to determine price appreciation. Based on the relationships estimated in the paper, a reduction of around 800,000 units over the next two years would imply that house prices should be roughly 0.8% higher two years from now than they would be if those distressed properties were listed for sale.

Direct benefit of removing distressed sales from house price data. There may also be a direct effect on house prices resulting from the removal of distressed sales, which typically occur at a large discount, from reported prices. Foreclosure reduction should clearly help in this way. The effect of an REO-to-Rental program would depend on how bulk sales of properties are recorded; in a worst-case scenario, discounted sales could all be recorded at the transaction price, appearing in price information used for appraisals and home price indexes and thus potentially putting additional pressure on prices. To put this in perspective, consider CoreLogic’s home price indexes. The index including distressed sales declined by roughly 4.6% from December 2010 to December 2011. With distressed sales of about 1 million per year, this implies that 40% of distressed sales could be removed from the transactions underlying the index, and would have resulted in an overall decline of roughly 3% instead of a 1.6 percentage point increase.

Related links:
Harp II’s lack of ambition – FT Alphaville
Harping on about housing – FT Alphaville
The foreclosure settlement – FT Alphaville

This entry was posted by Cardiff Garcia on Thursday, February 9th, 2012 at 21:03 and is filed under Capital markets. Tagged with The foreclosure settlement, US housing market.
Edit this entry.