The Atlanta Fed's macroblog provides commentary and analysis on economic topics including monetary policy, macroeconomic developments, inflation, labor economics, and financial issues.

Authors for macroblog are Dave Altig, John Robertson, and other Atlanta Fed economists and researchers.

February 24, 2009

Foreclosure mitigation: What we think we know

One of the most important challenges facing policymakers today is reducing the rate of mortgage foreclosures. It is a goal many think is at the heart of a sustained recovery in the U.S. economy. But as past attempts to reduce financial stress on homeowners have shown, the task is not easy. One of the complicating factors in formulating successful foreclosure mitigation policy is getting at the heart of the relationship between negative equity (the situation where the remaining mortgage balance is greater than the value of the house) and actual foreclosure.

Economic theory poses one categorical prediction about this relationship, which is that negative equity is a necessary condition for default. In other words, if a borrower is not in a position of negative equity, then he or she should never default. This conclusion follows simply from the fact that positive equity implies a borrower can sell the house, pay off the mortgage, and keep the difference—a better outcome under any circumstance compared with stopping payment on the mortgage and leaving the home.

What economic theory does not say is that if a borrower has negative equity, he or she should always default. The reason for this is that the owner could always default in the future, and thus there is value in waiting to see if house prices recover. Now, this value to waiting differs across borrowers and is sensitive to both the depth of negative equity and a borrower's financial situation. Why does a borrower's financial health matter? Well, the cost of waiting includes the monthly mortgage payment the borrower must continue to make. Borrowers who have plenty of wealth and a steady stream of income will be more willing to continue making payments than borrowers who are in financial distress, perhaps related to an unemployment spell or some other adverse financial shock.

So why does all of this matter in terms of thinking about a successful foreclosure mitigation program? Well, the appropriate policy prescription depends on the particular reason a borrower is currently considering default. I think it is useful to break things down in terms of three (not necessarily mutually exclusive) groups of mortgage borrowers:

  • those in unaffordable mortgages from the very beginning, who were implicitly relying on increasing house prices to refinance or sell for a profit;
  • those who have been hit by an adverse, but temporary, income/financial shock; and
  • those who purchased the house for strictly investment purposes and now see little or no hope of making a profit.

Borrowers may find themselves with unaffordable mortgages for many reasons. One might be an unscrupulous mortgage broker, who steered the borrower into an unaffordable subprime loan in order to generate high origination fees. Another, related situation would be an unaffordable interest rate reset on a subprime adjustable-rate mortgage. Finally, some mortgages may be permanently unaffordable because a buyer misrepresented income or assets during the origination process, a situation made easier by the growth of low documentation mortgages.

A large part of the administration's new housing plan—summarized succinctly by the New York Times, with lots of commentary (negative and positive) rounded up at Economist's View—is reasonably interpreted as being directed squarely at borrowers in the unaffordable-mortgage group. If policy is to be aimed at helping this group, the prescription is to offer the borrower a permanent reduction in monthly payments, whether it comes from lowering the interest rate, lengthening the maturity, and/or reducing the outstanding principal balance on the loan. The measuring stick often used in such plans is the debt-to-income ratio (DTI), which is the borrower's monthly mortgage and/or total required debt payments relative to his or her gross monthly income. While the administration's plan would succeed in lowering DTIs, the policy is temporary in nature (five years), and it is unclear what would happen to these borrowers after the plan runs its course—especially if negative equity is still an issue.

Many borrowers might have been able to afford their mortgages while employed but can no longer do so after they have lost their jobs. When housing prices are rising and homeowners enjoy positive equity, then distressed borrowers are able to sell their homes to pay off their mortgages. Alternatively, such borrowers can undertake cash-out refinances to gain some much-needed liquidity. Note that problems can occur for people in this situation even when positive future equity is a realistic hope. If the borrower is unemployed and liquidity constrained, the cost of waiting to default is very high and potential future price gains are of little value. Default in this case is much more likely, even though future prospects might be reasonably good. In this case, foreclosure-prevention policy could simply be used to eliminate the financial friction. In this case a lender would offer "forbearance," in which the borrower pays significantly lower payments for some period, with the arrears made up (with interest) later on. In this light, it is notable that the administration's key payment reduction plan has a five-year window.

However, one important concern regarding the plan is that servicers/investors don't have enough incentives to substantially decrease current DTI ratios. For example, if a household has a DTI of 60 or 70 because of a job loss, the servicer is responsible for modifying the loan to get DTI down to 38 and then still has to kick in a 50 percent match to further reduce it to 31. The costs borne by the servicer/investor are much larger than those borne by the government, which may not be such a bad thing in principal but in practice may result in low participation rates.

Note also that while permanent relief is the prescribed course for borrowers in the unaffordable-mortgage group, temporary relief is indicated for those in the temporary economic distress group. This highlights the difficulties in constructing policies when the underlying sources of stress differ by individual. The existence of a class of borrowers that purchased and financed residential real estate primarily for investment purposes further complicates matters. People in this group are in much different circumstances than those in the other groups and will default much more ruthlessly. A so-called "ruthless defaulter" has given up hope of positive future equity and hence there are no potential price gains to value. Under the theory of ruthless default, one effective policy intervention is to lower the outstanding balance of the mortgage so that positive equity—or even the hope of positive equity in the near future—is restored. Alternatively, the lender could forestall default at least temporarily by cutting the monthly payment below the cost of renting an otherwise observable house.

Aimed as it is at owner-occupied housing, the administration's plan does not offer direct assistance to those in the investment class. That may not be too surprising, as it is hard to generate much political sympathy for a group carrying a label like "ruthless defaulter." In addition, the perverse incentives of government assistance that usually go by the name of moral hazard are arguably more severe for individuals who purchase properties for investment purposes. However, abandoned properties do add to the stock of unsold homes, independent of who owned them or why they owned them. This does not necessarily argue for policy relief for investment buyers, but it is potential issue that bears watching.

Finally, there may be commentators with the view that loan modifications are a failing proposition as a few studies have shown extremely high default rates on modifications performed in early 2008 (for example, see OCC and OTS Mortgage Metrics Report, Third Quarter 2008). But, according to the table below (based on my calculations), the problem seems to be that the wrong type of modification was being performed. Approximately two-thirds of the modifications performed by servicers in the first two quarters of 2008 had the effect of increasing the principal balance of the mortgage and, as a result, also increased the borrower's monthly mortgage payment. In light of the above discussion, we should not be surprised by high re-default rates on these loans. On the other hand, there is reason to believe that successful implementation of payment reduction programs may indeed help to stem the pace of foreclosures.

  # Loans
    # %  
# %  
# %  
# %  
Q107 13,900 200 1.25 600 3.75 13,100 81.88 2,100 13.13
Q207 21,600 700 3.00 200 0.86 20,700 88.84 1,700 7.30
Q307 24,600 700 2.55 300 1.09 23,600 86.13 2,800 10.22
Q407 32,300 3,600 9.65 1,000 2.68 28,000 75.07 4,700 12.60
Q108 33,000 7,100 18.11 400 1.02 25,500 65.05 6,200 15.82
Q208 41,200 10,600 22.36 900 1.90 30,100 63.50 5,800 12.24
Q308 52,600 17,300 28.22 200 0.33 36,000 58.73 7,800 12.72
Source: Lender Processing Services

By Kristopher Gerardi, research economist and assistant policy adviser at the Atlanta Fed

February 24, 2009 in Fiscal Policy, Housing | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference Foreclosure mitigation: What we think we know:


"those in unaffordable mortgages from the very beginning, who were implicitly relying on increasing house prices to refinance or sell for a profit;

those who purchased the house for strictly investment purposes and now see little or no hope of making a profit."

I am one who would say that there should be no help for the people in these groups. They were gambling, purely, and should take their own losses. F- them, and investor-horses they rode in on.

And a question: you say, "The costs borne by the servicer/investor are much larger than those borne by the government, which may not be such a bad thing in principal but in practice may result in low participation rates." But isn't the relevant question, at least for servicers/investors, whether those costs are more than they would face if there were more foreclosures? Isn't that always the question for them in doing loan mods? If the servicers don't participate on those terms, aren't they then assuming that their foreclosure losses won't be that bad? I mean, I have no sympathy for the investors; they were careless, and should lose.

One of the things I object to in the mortgage bailout plan is the notion that the government can or should prevent house prices from falling further. The problem with this is that prices in many markets are still fairly inflated relative to incomes; that is, they're still basically unaffordable. As long as prices remain unaffordable, there are going to be a lot of foreclosures--it's just prolonging the pain. The only good long-term solution is to allow prices to reach a level that's actually affordable to buyers under normal (pre-bubble) credit standards. The housing market shouldn't get more stimulus--it should get less!

My wife and I make a decent, middle-class income, and yet we can't find reasonably affordable houses in our area. And now the government wants to use our tax money to make sure it stays that way! And to pay the mortgages of fools who got in over their heads. Do they understand why we might resent that a wee little bit?

Posted by: Moopheus | February 24, 2009 at 12:14 PM

Could you please clarify why some loan modifications have resulted in increases in total principal balances and in monthly principal repayments? I suppose that an increase in the former could be OK if it were accompanied by a decrease in the latter (i.e., if cash flow was the dominant issue) and that an increase in the latter could be OK if it were accompanied by an decrease in the former (i.e., if negative equity was the dominant issue), but why on earth would a loan modification be expected to work if both the total principal balance and the monthly repayments increased? That doesn't make any sense to me.

Posted by: Rich F | February 24, 2009 at 01:20 PM

No, I answer my own question: the reason servicers and investors will resist participation in a deal in which they have to take a loss on the loan mod is they want to pressure the government to give them a better deal, and protect them against any loss. It would be a bad and stupid move for the government to give in to them, but the Fed and the Treasury seem to have a hard time saying no to Wall Street.

Posted by: Moopheus | February 24, 2009 at 01:46 PM

Your entire premise is based on treating a symptom (foreclosures), rather than the cause (house prices). By any historical metric, house prices are too high (rent-price ratios, income-house price ratios, Case-Shiller, OFHEO, etc.). When house prices drop to prices that are affordable and cost competitive with other forms of shelter, a bottom will naturally form. Government intervention is not the right solution for this problem.

Posted by: uber_snotling | February 24, 2009 at 04:35 PM

I have been waiting for the housing to becom affordable to me in my area, and have been renting since 2002. I have an above average income. Why should I pay for those who live in a big house that they cannot afford for while I'm still renting?

The key problem is the housing is still too expensive. The natural market force is driving down the price. Why does the government wants to keep it expensive? Why does the idiotic government want to waste tax money paid by those who are renting, in order to keep the housing expensive to these renters?

Posted by: alex | February 24, 2009 at 05:20 PM

Hi Rich,

The reason why some payments go up on a loan mod is because the homeowner may have had a Pay Option ARM and was accruing negative equity.

When the payment is modified to a fixed rate loan, even if the rate is lower, the loan is now fully amortized.

I'm assuming that these homeowners actually READ their loan mod documents this time around but perhaps that's a false assumption.

If they couldn't afford the modified payment but signed anyways, this was just a step to buy the homeowner more time to possibly sell or to save up money before ruthlessly defaulting later.

Posted by: Jillayne Schlicke | February 24, 2009 at 05:34 PM

"Could you please clarify why some loan modifications have resulted in increases in total principal balances and in monthly principal repayments?"

What happens in a lot of cases is not a real loan mod, but a repayment plan, where past due amounts are added to the principal, and the payments are readjusted (upward) to reflect the new balance. And you're right--it's not a great deal for the borrower, which is why these "mods" have a high rate of failure. The borrower stands a better chance if the amount of actual debt is reduced, but then the lender has to be willing to write off the difference.

Posted by: Moopheus | February 25, 2009 at 12:01 PM

"One of the most important challenges facing policymakers today is reducing the rate of mortgage foreclosures."

Dave, please can you explain WHY? If home ownership is at 'unsustainable levels', if debt/income ratios are 'too high', then why should policymakers prevent an adjustment? This is not intended to be partisan/political, I'm genuinely interested in the rationale behind your opening sentence. Thanks, MW.

Posted by: MW | February 26, 2009 at 08:51 AM

Is there a way out of this mess. Let the finger pointing begin. All the Rep. are say :look at the Dems. they are screwing up!" But 8 years of asleep at the wheel can not be fixed overnight.

Posted by: Orlando | February 28, 2009 at 04:44 PM

I second Moopheus...

I also want to know why is it good to keep house prices artificially high?

Also, if we were to help underwater homeowners (for the sake of saving the economy) this thing has to be done such that irresponsible homeowners profit at the expense of taxpayers. They need to give-up something in return. For instance, some "option value", such that if their house appreciate, they have to repay the government.

Posted by: FC | March 02, 2009 at 08:16 PM

I think that the idea is that foreclosures represent a sort of collective action problem: Individual banks would like to foreclose and resell, but if everyone's doing it, prices are driven lower, more mortgages go underwater, and the cycle repeats. In today's market foreclosed homes often sit unoccupied, steadily losing value.

So by keeping people in their homes, even with modified loans, the banking system overall is better off. The goal is to stem the panic and reduce write-downs on toxic mortgage assets.

I've been wondering what would happen if, rather than setting up all of these hoops for homeowners and banks, the government simply imposed a temporary, $10,000 per foreclosure on banks per foreclosure? This would cost taxpayers nothing and would provide an incentive for banks to modify their own loans.

Posted by: Tom | March 02, 2009 at 08:36 PM

I understand the logic of the $10,000 (or whatever amount) per foreclosure. This is a good idea. The only problem I see is that banks would factor in this expense in future morgages.

Anyway, and I know I am being repetitive, I do believe that current homeowners need to pay back any help received. Besides an "option" triggered by appreciation, another idea is to void the tax exemption on capital gains when selling the house for all homeowners who get relief from the government.

Posted by: FC | March 03, 2009 at 01:42 PM

We can't artificially manipulate prices. We need the market to correct naturally.

Posted by: Brian Dickerson | May 20, 2009 at 11:39 AM

I like your break out of the 3 different homeowner categories. We tend too often to lump all homeowners in need of loan modification into the same category. But sadly, I believe from all the horror stories I constantly hear form distressed home owners or former owners through my mortgage business that many, many lenders are much more proned to go the foreclosure route than to do what I believe is the right thing for both the home owner and the bank's bottom line.

Posted by: Ron Stone | August 07, 2009 at 12:20 PM

I don't understand why banks don't take more measures to try to prevent foreclosure when it costs them a ton of money. It would be beneficial for the homeowners as well as the lenders.

Posted by: Gilbert | November 26, 2009 at 12:12 PM

Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in

February 18, 2009

The American Recovery and Reinvestment Act in pictures

A year-by-year look at where the money goes:

Note: The dollar amounts listed below are denominated in millions of dollars.








Which adds up to:



Finally, the relative size of each year's spending:


Some other breakdowns of how the money will be allocated can be found here and here.

There you have it.

By David Altig, senior vice president and research director, and Courtney Nosal, economic research analyst, at the Atlanta Fed

February 18, 2009 in Fiscal Policy | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference The American Recovery and Reinvestment Act in pictures:


David - THANK YOU. I've been trying to get various media outlets to do something similar for weeks now. Despite their increased predilections for graphics nobody really bit.

On that note perhaps you'd care to take the next step and map the various components to impact timeframes, perhaps with judgments about effects. Contemplate an array with "columns" defined by s.t., intermediate and strategic impacts. And then project the speed vs multiplier impacts.
Possible, interesting, good tracking and on-going evaluation tool ?

Posted by: dblwyo | February 18, 2009 at 06:27 PM

Very useful. Do the figures above refer to fiscal or calendar year? Thanks in advance.

Posted by: Andrea De Michelis | February 19, 2009 at 05:44 AM

WP graphics just before the House confirmation. Complements your graphics.


Posted by: Sanjay Bapna | February 19, 2009 at 09:46 PM

Thanks, useful stuff. As I was scrolling, I was thinking, "yeah, but, what is the total breakdown between years and groups" and the next charts appeared!

Posted by: Paul | February 20, 2009 at 12:16 AM

Thanks for this. It helps visualize rather than just hear numbers.

Posted by: jeff | February 21, 2009 at 05:57 PM

Hi Dave and Courtney,

Thanks. Don't know if you guys are still seeing the commments to this post, but I was wondering about the year-by-year totals. I get something closer to 20% (spending alone or with revenue reductions) in 2009. Am I interpretting this incorrectly?

Posted by: Guhan | March 11, 2009 at 04:14 AM

Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in

February 05, 2009

There is no accounting for priorities

Just in case you are desperately seeking some refuge from the pervasive blogland commentary on the fiscal stimulus proposal winding through the Senate (which already made its way through the House of Representatives), be forewarned: You won't find it here, but you will find an update to the old adage, There's no accounting for taste (de gustibus non est disputandum), which is to say that when it comes to the fiscal stimulus package, there's no accounting for priorities.

The Senate bill is not yet a done deal, of course, but a couple of clear differences between it and the House bill have emerged. According to the Congressional Budget Office—or CBO, from whom all figures in this post spring—the Senate bill is slightly bigger ($884.5 billion versus $819.5 billion) and would implement the stimulus at a faster pace. The current Senate bill would introduce about 79 percent of the expenditures and tax cuts in 2009 and 2010. The corresponding figure in the plan that came out of the House is 64 percent.

Perhaps more interesting—and maybe more confusing—are the priorities reflected in the separate bills:



The share of the stimulus devoted to discretionary spending—the place where, for example, infrastructure and education spending reside—is pretty similar in both stimulus plans (about 28 percent in the House version, about 26 percent in the Senate version). What is clearly different is the much greater reliance on tax cuts in the Senate bill, compared with the House bill's emphasis on "direct spending."

In a sense, this distinction is as much an issue of labeling as anything. The majority of the items in this category of direct spending are "provisions that would increase direct spending for unemployment insurance, health care, fiscal relief for states through the Medicaid program, and other programs," according to the CBO. In the language of economists and national income accountants these are "transfer payments," or funds that are transferred to individuals. Formally, they are subsidies for certain types of economic behavior—job seeking and purchasing health care, for example—and hence are really just a negative tax.

There is a certain arbitrariness to the distinction between increases in transfer payments and reductions in tax payments. This arbitrariness is illustrated by a change the CBO made between its initial assessment of the draft House bill and its (largely unchanged) summary of the bill that passed:

"The Congressional Budget Office, in consultation with JCT [Joint Committee on Taxation], has concluded that the subsidy for health insurance assistance for the unemployed should be treated as an increase in outlays rather than a decrease in revenues. Although this treatment is different from that in the table provided in our estimate for H.R. 1 as introduced on January 26, the overall effect on the budget remains the same for each year. JCT has also adjusted its estimates of the mix of revenue losses and outlay increases associated with certain refundable tax credits; that change also has no effect on the budget totals for each year."

Still, if you are likely to be on the receiving end of one of these programs, the distinction is probably not so arbitrary. From this end-user perspective, there is an important economic distinction between approaches taken in the competing plans. So then, which approach to "tax cuts" is better? At this point, I will send you to the aforementioned pervasive blogland commentary. You will find no shortage of opinions.

By David Altig, senior vice president and research director at the Atlanta Fed

February 5, 2009 in Federal Debt and Deficits, Fiscal Policy, Taxes | Permalink


TrackBack URL for this entry:

Listed below are links to blogs that reference There is no accounting for priorities:


the stimulus is disturbing to me in where it is spending the money as well. it will go into non-productive things from a GDP perspective.

the other disturbing thing that I saw was an interview with Christine Roemer on CNBC. She was asked point blank if the Obama administration was confident in Bernancke, and she refused to give him that vote of confidence.

mankiw in his blog cites the multiplier effect on tax cuts versus spending. He calculates that tax cuts offer a 3:1 bang to the buck growth in GDP, where govt spending offers a 1:1 bang to the buck.

Economic results over the last 3 decades would tend to prove him right.

Posted by: jeff | February 07, 2009 at 02:03 PM

When are MBS yields affected by the Fed's purchases? At the time the purchase agreement is signed, or afterward, at settlement?

Posted by: Holden Lewis | February 09, 2009 at 02:44 PM

But as past attempts to reduce financial stress on homeowners have shown, the task is not easy.

I'm not sure I agree the purpose was ever to reduce financial stress on homeowners. That said, I agree the task is not easy.

Posted by: FutureRob | April 09, 2009 at 09:55 PM

Post a comment

Comments are moderated and will not appear until the moderator has approved them.

If you have a TypeKey or TypePad account, please Sign in

Google Search

Recent Posts



Powered by TypePad