Original posted on FT Alphaville by Tracy Alloway:
In addition to the difficulty of converting temporary mortgage modifications into permanent ones, one of the big question marks hanging over the US Treasury’s Home Affordable Modification Plan is the redefault rate. That is, the percentage of homeowners who redefault on their modified mortgage.
FT Alphaville has mentioned before that in cases of severe negative equity, it might make more sense for a homeowner to make a couple of Hamp-reduced interest payments on his or her mortgage and then walk away. The US Treasury hasn’t given an official default rate for the programme yet, but figures like 25 per cent and 50 per cent have been bandied about.
In their US Securitised Product Outlook for 2010, however, Barclays Capital take a slightly more negative outlook on the redefault rate. It might be better than previous mortgage modification plans — but it’s still likely to be pretty dismal, they say:
Re-default performance for loans modified in Q3 08 has been dismal, with more than 60% relapsing into deep delinquency already. However, HAMP has been more aggressive than earlier mods – reducing borrower payments by 30-40%, compared with earlier modification efforts that typically reduced monthly payments by 15-20%. As Figure 6 shows, higher payment reductions reduce re-default rates, but only by 5-10% for that magnitude of payment change . . .
On the flip side, HAMP does not address the issue of negative equity, which is one of the primary drivers behind default . . . Taking these factors into account, we expect overall HAMP re-default rates to show not more than a 10-20% improvement over the default rates seen in past mods.
With the redefault issue then, plus the conversion rate for permanent modifications and various servicer problems, BarCap thinks we’ll see some sort of revision or significant tweaking of the programme.
Possible changes could include further streamlining the documentation requirement for Hamp applications, creating a lower debt-to-income target or second-lien programme, or, perhaps most significantly, starting principal forgiveness instead of just forbearance.
Here’s a summary:
Finally, watch out for new policy changes from Washington on the mortgage front. If HAMP does not work well (as we expect), and foreclosures keep rising, Congress might revisit some of the more radical suggestions from earlier this year, such as cram-downs, forced debt forgiveness, etc. On the agency MBS side, one tail risk is the prospect of an off-market, low mortgage rate provided by the government. MBS investors fearful of this shift compressed the coupon stack sharply in Q1 09 – if such an off-market rate is actually offered by the government, it could greatly hurt premiums and, thus, all agency MBS valuations
. . .
A greater share of debt forbearance mods would lead to upfront losses on the pool, in turn leading to higher initial [constand default rates]. However, since debt forgiveness mods typically perform better than comparable rate reduction mods, re-default rates would be lower (Figure 27). Higher losses upfront on the forgiven amount would imply that subordinates would be written down faster on subprime deals, causing crossover to occur sooner. This would benefit the second and third cash flows at the expense of the first cash flow bond as the principal waterfall switches from sequential to pro rata.
Given all of the above, readers might well be scratching their heads as to what the Hamp is actually good for. And on that point Barclays is very clear — shadows and cans:
To be clear, the modification program (HAMP) is not a silver bullet. As Figure 6 shows, historical re-default rates for all types of modifications are high – HAMP should be better, but not hugely so. But the process of modification buys time. It increases the number of months between the borrower turning delinquent and the home hitting the market. This is shown in the REO (real estate owned) line in Figure 5; even as foreclosures keep rising, the REO bucket has gone down. So kicking the foreclosure ‘can’ down the road has helped prices stabilize.
Intuitively, if there are millions of foreclosures to still work through the system, it is better to spread them over a few years than have them hit the market in six months – this prevents prices from over-correcting to the downside. And with the Administration focused on modifications, we expect long delinquency-to-liquidation timelines to help home prices.
As a result, our forecast calls for prices to drop 8% from current levels, before stabilizing in Q2 2010. The macro impact of this decline should be muted. After all, a house worth $100 is now worth $67 (prices have fallen around 33% from peak in Case Schiller). A further 8% decline from current levels is simply another $5.3. As every month passes without a sharp increase in the REO bucket or a sharp drop in home prices, the tail risk posed by housing declines ever so further.
`Real estate owned’ means the properties owned by banks and mortgage companies — the stuff we call `shadow housing inventory‘ since it’s not included in official measures of unsold housing inventory.