October 30, 2011
Home Affordable Refinance Program
Last week the Federal Housing Finance Agency, Fannie Mae and Freddie Mac jointly announced changes to the Home Affordable Refinance Program (HARP) with the goal of making it easier for some households to refinance their mortgages at lower interest rates. Here I offer some thoughts on this proposal.
The graph below plots what some of us see as a key factor in the economic problems that the U.S. has been experiencing over the last four years. Household mortgage debt grew 60% faster than income between 2000 and 2007, and we may not see a return to a healthy economy until that ratio returns to more typical historical values. That could come from a combination of (1) foreclosures-- a painful process with significant deadweight loss for all parties--, (2) higher saving rates-- which make it harder to realize short-term growth in total spending-- or (3) strong GDP growth-- which is hard to achieve given (1) and (2).
Another way that the debt burden could be reduced is through refinancing. Often U.S. fixed-rate mortgages allow the borrower the option to pay off the loan if interest rates decline, replacing it with a new loan based on a lower interest rate. For the borrower, that means either lower monthly interest payments or a loan that gets paid off more quickly. Although under the FHFA proposal the household would owe the same notional amount as before, the real burden of the debt on the household is lower, and the household is clearly better off, if it can refinance the loan at a lower rate. The household's gain is naturally the loss of the party who was set to receive the original mortgage payments. If the lender receives a lump-sum payment of the full principal today, they could not use it to purchase an asset with as high a monthly income stream as the payments promised under the original mortgage.
One obstacle to refinancing has been mortgages that are underwater, which means that, as a result of declines in house prices since the time of the purchase, the principal owed on the mortgage exceeds the current resale value of the home. Previous rules would not allow Fannie or Freddie to guarantee a loan whose value exceeds that of the home, which refinancing an underwater loan would require. The new FHFA proposal relaxes that requirement so as to allow refinancing for underwater loans that were originated more than 2-1/2 years ago and on which the borrower is current on the payments with no late payments over the last 6 months.
One question of interest is, who will ultimately end up losing the income that corresponds to the household's gain from refinancing? Since the original loans are currently guaranteed by Fannie or Freddie, and since Fannie and Freddie's liabilities in turn are now de facto guaranteed by the U.S. Treasury, one's first thought might be that the household's gain is ultimately the loss of the U.S. Treasury. However, Fannie and Freddie guarantee against default but not against the loss that comes from pre-payment, so it's the holder of the loan, not the U.S. Treasury, that loses. On the other hand, Fannie and Freddie own over a trillion dollars of the loans themselves, and the Federal Reserve owns another trillion. The Federal Reserve contributed $82 billion to the U.S. Treasury this year from its earnings on the mortgage-backed securities and other assets that it holds. A lower income flow from these would reduce the size of the payments to the Treasury that the Fed is able to make and increase the net contribution the Treasury needs to make to keep Fannie and Freddie solvent.
Other holders of agency-guaranteed loans include U.S. commercial banks, institutions outside the U.S., mutual funds, retirement funds, and state and local governments.
In terms of the agencies' default guarantees, they currently have partial recourse in terms of representations and warranties issued by the originators of the loans. The new proposal is to waive these for the refinanced loans, which in principle gives the government exposure to a little more of the default risk than it had before. Here is the FHFA's analysis of this concern:
Nearly all HARP-eligible borrowers have been paying their mortgages for more than three years, and most of those for four or more years. These are seasoned loans made to borrowers who have demonstrated a capacity and commitment to make good on their mortgage obligation through a period of severe economic stress and house price declines.
The program would only be available to households that have been making their loan payments, despite being underwater. These are people who want to pay what they owe and are trying their best to do so. Insofar as refinancing makes those payments easier, one would expect the refinanced loans to exhibit lower default rates than would be observed if refinancing is not allowed.
My conclusion is that the direct consequences of the proposal potentially may entail a modest adverse budget impact, but that indirect benefits to the overall economy and perhaps even to the Treasury as well outweigh these. The proposed modification of HARP looks to me like a reasonable plan.
Posted by James Hamilton at October 30, 2011 06:47 AMdigg this | reddit
CBO has already done the analysis.
Posted by: mort_fin at October 30, 2011 07:46 AM
Your First item “(1) foreclosures-- a painful process with significant deadweight loss for all parties” isn’t entirely true. Not everybody loses. In fact … it’s “only” those who have moral hazard exposure that lose.
The investor or to-be homeowner who picks up the property at its true economic value isn’t hurt.
The home repair guys who get called in to fix the property up aren’t hurt because its new owner isn’t insolvent.
This is not a liquidity problem. It is a solvency problem.
So also is this crazy idea that abrogating bankruptcy laws is somehow better for society. The only thing it does is keep the current 1%’ers’ in their seats of power. If you’d let capitalism do its work, a redistribution of the malinvestment to this class would stop.
Posted by: ArkansasAngie at October 30, 2011 08:01 AM
I think prepayment will actually bring up the value of MBS the fed holds. At the time these were created, there was lots of turnover in housing, so most of the bets were likely on prepayment and against interest payment more than several years out.
Posted by: aaron at October 30, 2011 08:07 AM
Dear Professor Hamilton,
I notice that the mortgage information from Federal Reserve for the Z.1 Release, Table B.100, line 33 shows quarterly data for 2010 and 2011. For 2010, did you use the fourth quarter GDP and Mortgage data for your chart and annual data prior to 2010 (humble learning point)? Also, I notice that if the “Y” axis uses 10 point increments rather than 20 point increments, the upward slope of the curve looks even more unsustainable from about 2000 to 2007 (also your comment relating household mortgage debt to income is staggering-where were the regulators regarding who is qualified to borrow?). A couple further comments, I notice that for 1975 the Mortgage/GDP ratio is about 28% and for the 1960 the ratio was about 27%. It seems like the ratio was fairly stable for many years. Both the borrowers and the investors had a party that seems to have started about 1985, got more raucous and each thought the other would pay for the party.
Posted by: AS at October 30, 2011 09:44 AM
AS: The graph uses quarterly data throughout (obtained from the historical data download option).
Posted by: JDH at October 30, 2011 10:41 AM
Dear Professor Hamilton,
Posted by: AS at October 30, 2011 11:23 AM
This should have been done years ago. HAMP was created in 2008. Obama has had the money to do it and the authority since he took office, but not the will or courage.
There is a political price to be paid. People seem not to mind too much if millionaire banksters get government handouts, but they scream bloody murder if their next door neighbor gets a dime. It's just the way people are. They are more concerned about their status relative to their near peers.
Posted by: Joseph at October 30, 2011 02:20 PM
Looking back to Q4 1951, the Mortgage/GDP ratio was 15% and as of Q2 2011 is about 66%. What should the ratio be to avoid a future Home Affordable Refinance Program?
Posted by: AS at October 30, 2011 04:38 PM
It is a perverse commentary on modern welfare capitalism that the Treasury and Fed have been busy spoon feeding hundreds of billions of zero interest money to financiers that have demonstrated a total lack of talent in allocation of capital and inability to manage risk.
Meanwhile, millions of ordinary citizens of the 99% who have faithfully paid their mortgages every month are cut off from refinancing at the lower interest rates the Fed broke their backs creating with quantitative easing because of a loss of equity for which the financiers are responsible. This is a perfect example of the corruption of capitalism.
Posted by: Joseph at October 30, 2011 05:38 PM
A straightforward case of the administated globalization in thoughts and economic planing "asinum,asinus fricat" (the donkey rubs the donkey)
Prices of real estates have increased by almost 300% in a 20 years life span.
Indebtness reads as follows:
Non financial debts.
The French household bears the less financial debt when compared to public and municipalities (as a percentage of gross income) 77.42% in 2010 against 52.04% in 2000.The predominant shares of the savings are in real estates.
It is worth remembering that Japan real estates prices inflation of the 90s, came after a virtuous cycle,increased industrial production,increased exports, increasing foreign currencies reserves, same for Taiwan,same for Korea,same for China.The other way around for Europe and the USA.
The public debt is and will crowd out the private sector debt for many years to come.
Posted by: ppcm at October 31, 2011 05:03 AM
professor, "who loses from a rise in prepayments" is not such a simple question to answer.
the prepayment option in mortgages is a call option, sold to households. households gain because the call they own rise in value, sure (but keeep in mind, mortgages are priced with the option included, so they are paying for the option they bought with a higher rate).
Holders of MBS do not necessarily own the option. They may hedge the option (many do) by buying protection through interest rate derivatives. In fact, the risk policy of most organization *requires* the bank to keep exposure within certain bounds, hence they must buy at least a portion of the option back. If they did not hedge the option, and they falsely bet that rates would never get so low and borrowed at a fixed rate against to lock in the net interest margin on the MBS, then right now they would be loosing a truckload of money (the MBS gets refinanced, the debt stays on the books, giving them negative interest margin).
so the "looser" of the refinance is whomever owns the interest rate derivatives hedging the call option, and whomever owns unhedged MBS. Not banks, probably hedge funds, or other investment funds (like PIMCO) who do not have strict risk guidelines to protect them from interest rate risk.
Posted by: dwb at October 31, 2011 05:33 AM
Help me out a bit. To get the Mortgage/GDP ratio down to what it was in 2000 at 47% of the GDP, it will take an increase in the GDP of about 47% if there is no increase in mortgage debt. At a GDP growth rate of 3.0% per year this will take over 15 years. How can small relief programs impact anything on the scale that is needed for real estate and general economic recovery? I am beginning to think that the Reinhart and Rogoff book, “This Time is Different” is an optimist’s hymnal.
Posted by: AS at October 31, 2011 07:54 AM
This is a middle (or uppper middle class) subsidy, no? To qualify, you had to buy at the height of the market but be curent on your payments. So we're talking about someone with a job and not under such duress as to walk away from a mortgage.
I have any number of friends--investment bankers, fund managers, PhD pharma researchers and the such--who could really benefit from such a program!
Posted by: Steven Kopits at October 31, 2011 08:37 AM
"GSEs guarantee against default but not against the loss that comes from pre-payment."
incidentally, you also left out an important point on the guarantee -there is a liability but also a revenue stream associated with it. In GSE lingo this is the "G-Fee" business. The impact insofar as the guaranteee business (to the GSEs) may be either positive or negative, there is not enough information, but very likely positive for the budget (my sense, based on long experience with the GSEs).
GSE's actually earn a margin for insuring default, which is priced based on their credit models. It is both revenue (the fee) minus the liability (the expected charge). Think of it this way: a portion of your mortgage interest goes to GSEs for the guarantee. Ideally, the G-fee is priced to earn a return on capital (which means a positive income stream as long as the model is right!). When there's a refinance, the revenue/liability/income stream are replaced at the new g-fee.
The impact from this refinance could be either positive or negative for the GSE's. But, if the old loan had negative margin (meaning the liability is larger than the revenue stream, because the credit model says they are way more likely to default than at loan inception), then it seems likely they are replacing the negative-margin G-fee stream with a less-negative or even positive margin G-fee stream. Better consumer cashflow due to lower rate ought to improve the credit & lower the g-fee liability, while GSEs may even be able to charge the *current* g-fee - which is now much higher than the past.
In other words, if the GSE's get to reset the G-fee and simultanouesly lower the liability, the impact on the g-fee business is very likely a solid positive (meaning a positive budgetary impact). Even if they do NOT increase the fee, the impact is positive.
on prepayments, GSE's have strict interest rate risk limits, plus the g-fee book of business is much larger, so the impact from prepayments is likely very small.
So who "loses" on the prepayments? mainly retirement funds, PIMCO, and other holders of the "naked" prepayment option. I want add these are all sophisticated investors and have sophisticated mathematical models to assess the call risk. but not banks or the GSE by and large.
Posted by: dwb at October 31, 2011 01:57 PM
Regardless of who wins or loses with this refinance arrangement, the good thing about it is that it is open, above-board, and legal, as opposed to the many fraudulent and illegal practices that have been condoned by regulators in the housing market, like no-doc mortgages, 'rocket-dockets', robot signers on legal foreclosure documents, and MERS not having legal standing for the foreclosures they pursued.
So here we have a legal program that will make some positive impact on macro-economic recovery. Its a breath of fresh air.
Posted by: Mike Laird at October 31, 2011 02:33 PM
"it will take an increase in the GDP of about 47% ... at a GDP growth rate of 3.0% per year this will take over 15 years."
Only about 12 years (compounding).
Posted by: KevinM at October 31, 2011 05:12 PM
Thanks for the correction. I feel much better now, knowing it will only take 12 years.
Posted by: AS at October 31, 2011 05:49 PM
A not contradictory but alternative view looks at the fact that housing starts follow new household formation -- and up to 2006 there was a boom (bubble) in new household formation that was even larger than that in housing starts. Unprecedented.
Since then there's been an even bigger plunge in new household formation that housing starts have followed down, and lagged. By this measure, there are very arguably still *too many* homes being built (!).
Data here. To the extent that demand for housing is fundamentally driven by household formation, it is hard to see how the housing market is going to recover without an upturn in the creation of households. Moving interest rates on loans around a bit doesn't seem to have much to do with it.
The clear bubble in household formation that preceded the "housing bubble" is near never mentioned -- strange, considering its plain implications as to the cause of the latter -- and its own cause is a bit of an unexamined mystery.
But it hardly seems likely that there will be much of a recovery in household formation with unemployment >9% and the economy remaining so weak. Census in fact reports that household consolidation ("destruction") continues.
And with no demand for housing ... there doesn't seem there will be much recovery in the business of supplying it. Period.
OTOH, an increase in new househol formation, getting back up to the old long-term trendline, could float all boats. But what is that going to take?
Posted by: Jim Glass at November 1, 2011 12:19 AM
I helped one person qualify for the HARP program and I am currently working with another.
The program is essentially what in the past was done where the borrower and lender sat down together directly. Today the mortgages and the institutions are so convoluted that it takes a private detective to sort out who actually owns the loan so that you can negotiate with them.
The person I helped qualify had income slightly above the poverty level. She had bought a town house with her 2 daughters and one walked out stucking her with the whole mortgage payment. It took three years of negotiation, with three mortgage companies and two loan administrators because the loan was sold a number of times. It was not until we began to negotiate with the last loan administrator - actually when foreclosure was filed - that we found out that the loan was totally owned by FNMA. Our HARP negotiations became complete one week before the foreclosure mediation, needless to say at the mortgage companies and FNMA the right hand did not know what the left hand was doing. The good news is she got to keep her home.
I recommend the program for all levels of income, if someone is trying to save a home from foreclosure.
You might be interested in the way FNMA structured the final loan.
-30% of the principle was made a balloon payment due at the end of the loan or when the home was sold, whichever came first.
-The loan was changed from a 30 year to a 44 year loan with the remaining 70% of the loan financed at 5%.
The result was payments at about the level of a 2% loan on the outstanding principle. What would bank auditors think of this kind of deal at a commercial bank?
Posted by: Ricardo at November 1, 2011 07:55 AM
The corollary to household formation is the number of empty bedrooms, not the number of empty houses. When this ratio of examined, the inescapable conclusion is that America is far more "over-housed" than anywhere else in the world.
There is also a demographic effect to consider, i.e., boomers needing to rent bedrooms for income (due to lower pension, 401k, IRA and SS payments) and their kids needing to lower their rent to afford mandatory repayment of crushing student debt loads.
I believe these two factors will keep home prices from "catching up" to their peak for many years, if not decades
Posted by: Jeff G at November 1, 2011 08:56 AM
It may not make a huge difference, but may be more accurate to graph household debt / GDP over time as opposed to mortgage debt / GDP over time. Home equity lines and second mortgages were heavily marketed by banks starting in the mid 1980s, early on they were marketed as substitutes for other types of consumer debt due to tax advantages that I believe were subsequently eliminated. I'm sure the trendline is still upward, hopefully not steeper. I would bet that the increase from 1985 to 2000 was mainly a function of home equity and second mortgage lending since there was really no real estate bubble during this period. Post-2000, probably all real estate bubble.
Posted by: Frank at November 4, 2011 04:35 PM
Regardless of the general low regard human beings exhibit towards each other , trappedin a system of usary and disregad for the family and the need for most, of a fair housing oppertunity, the above discussion is a moot point. this financial system can slice and dice and play with numbers all it wants, it is going down and needs deep reform if any generation is to 'buy' into it again. Wss honesty, fairness,compassion and our own freedom to create our own safety to live within our birthrights all compromised by a legion of money changers? Yes and this is not what I want for the children to grow up into.
Posted by: jane at November 5, 2011 05:04 PM