Economic Stimulus Through Refinancing — Frequently Asked Questions
- What is our solution to the housing and economic slump?
- How would the program work?
- How much can an affected homeowner save?
- Why isn’t this program just another subsidy for the housing market? Shouldn’t we just let the free market work?
- Are there any direct costs of the program to taxpayers?
- How will this program affect the federal budget deficit?
- Isn’t this plan unfair for bondholders? Why should bondholders take a loss to subsidize homeowners?
- Isn’t this just a transfer from bondholders to households? How can such a transfer provide a net economic stimulus
- What would the impact be on the bond market and future mortgage rates? Would this program disrupt the operation of bond markets?
- How much will a refinancing program stimulate the economy?
- Isn’t the federal government already helping troubled borrowers?
- This program seems to be an expanded version of the federal program, HARP. How will this proposal succeed in expanding the reach of HARP?
- How would this program be implemented? Are there impediments to pursuing this program?
- What incentive do servicers have to participate in the program?
- Why not help borrowers who do not have government-backed mortgages, or renters?
- What about mortgages with second liens?
- Will refinancing under this program change my recourse status?
- What about loans with private mortgage insurance?
- Do “representations and warranties” or other liability claims provide an impediment for this plan?
1. In more typical periods, mortgage rates as low as the current prevailing rates would spark a refinancing wave, freeing up cash for consumers to spend, stimulating the economy. Because lenders are nervous about the state of the housing market and the overall economy, refinancings have not reached the levels that would provide such a jump-start to the economy. However, because the government guarantees about 37 million outstanding mortgagesi, an opportunity for action arises. We propose that the government direct federal agencies involved in the mortgage market (Fannie Mae, Freddie Mac, Ginnie Mae, the Federal Housing Administration, and the Department of Veterans’ Affairs) to assist recent borrowers in refinancing their loans. The agencies would then direct mortgage servicers to send brief refinance applications to borrowers, allowing them to refinance their mortgages to current market rates without much paperwork.
2. To fund the program, the agencies would issue new mortgage-backed securities (MBS) to fund the refinanced mortgages and use the proceeds to pay off the existing MBS, just like what currently happens when borrowers refinance. Agencies would receive the same cash flow to cover default risk that they do now, passing along reductions in financing costs to borrowers.
3. Suppose that you bought your house in 2006 for $225,000, taking out a $200,000 mortgage at a fixed rate of 6 percent. Your payments are about $1,200 per month. Thus you owe around $189,000, but home values have fallen to the point that your home is worth less than that amount, making conventional refinancing impossible. Adjusting the interest rate to the current market rate (about 4.3 percent) would reduce your mortgage payments by 15 percent, resulting in savings of more than $2,000 each year.
4. The private market is currently not functioning in funding mortgages, to a great extent due to the breakdown in securitization. The U.S. government is currently backing 19 out of every 20 new mortgages. In addition, the government already bears the risk of the 37 million mortgages it has backed. The program we propose would better manage the government’s portfolio of mortgages to reduce losses at the same time as helping homeowners lower their mortgage payments. This opportunity would also be seized by the private market would do if it were operating normally without credit-constrained and underwater borrowers and lenders who are reluctant to make new loans. The program can be implemented without costs to taxpayers.
5. The fee paid to servicers would be added to the mortgage; it would not be paid by the government. The modest refinancing costs under our plan would be unlikely to deter borrowers from refinancing. The only costs to taxpayers would be whatever publicity costs are required to announce the plan.
6. The refinancing program to help reduce the federal deficit over the long term. According to Laurie Goodman of Amherst Securities Group, the future cost to taxpayers of loan guarantees to the GSEs (Government-Sponsored Enterprises, such as Fannie Mae and Freddie Mac) will add up to $448 billion. The plan we propose will lower that amount, reducing foreclosures and working to stabilize home prices.
The program might also increase net tax revenue, as lower mortgage interest payments would result in fewer tax deductions. We think this might cause, at best, a small additional reduction in the deficit. In essence, homeowners pay Mortgage Interest x (1 - individual tax rate), and the government collects Tax Rate x Mortgage Interest. However, the government loses revenue from any taxable interest payments not made by investors whose bonds are prepaid, which could be a wash. There are two additional considerations. On the one hand, not all homeowners itemize (perhaps two-thirds of homeowners itemize). On the other hand, a majority of all bonds ($3.7 trillion out of $5.7 trillion) are held by non-taxable investors (GSEs, the Federal Reserve, and overseas investors). These effects might net out modestly in the government’s favor, but we do not think it is a major factor in assessing the idea.
7. Bondholders purchased securities earning 1.5 percent to 3 percent above comparable U.S. Treasury yields while bearing no default risk. The risk they faced was that when mortgage rates fell, their bonds would be paid off early, which is what we propose should happen today.
Bondholders have been and continue to be large beneficiaries of government programs. The Federal Reserve’s purchase of $1.25 trillion of mortgage-backed securities (MBS) has increased the price of MBS, a large ongoing subsidy for holders of these long-term bonds. As well, bondholders received an additional windfall when the federal government turned its implicit guarantee of GSE bonds into an explicit credit guarantee in 2008. We believe that the net impact of government interventions in credit markets has been enormously positive for bondholders. For those who advocate a “free market” solution, were the federal government and the Fed to suddenly withdraw its support for bonds, bondholders would suffer large losses much bigger than they would due to early prepayment of bonds.
Bondholders will be paid off early at par, not at a “loss.” Regulated institutions such as banks and pension funds that carry the bonds on their balance sheet in a “hold to maturity” account will not suffer any loss of capital, as such bonds are being carried at par. The improvement in the overall economy would benefit all investors, including existing bondholders.
8. The impact of refinancing the bonds would provide a substantial net stimulus to the U.S. economy. Lower mortgage payments would result fewer foreclosures, saving money for lenders, homeowners facing possible foreclosure, their neighbors and other community residents. Those now trying to sell their home benefit through fewer fire sales of previously foreclosed houses. Society benefits through fewer homes damaged or destroyed during the foreclosure process itself.
In addition to net benefits from fewer foreclosures, any net transfer from bondholders to homeowners does not have a zero-sum impact on the economy. A refinancing wave would bring an appreciable net stimulus.
All savings from refinancings would go to homeowners. Many homeowners are liquidity constrained and suffer from a lack of available credit, so reducing these borrowing constraints will provide a substantial lift to spending. Importantly, the impact of lower mortgage payments would be both large and long-lived. Thus far, lower mortgage rates have translated to a very small increase in economic activity because credit markets are not enabling new loans to households.
9. Because refinancings replace existing bonds with new bonds, this program would not impact the net supply of outstanding bonds, thus limiting the impact on overall mortgage rates and on long-term interest rates. However, the logistical process of refinancing more than $5 trillion of bonds has the potential to lead to greater market volatility. However, the Federal Reserve, which already owns about $1.25 trillion of the bonds, could ensure an orderly market by buying or selling bonds as needed for stability.
Newly issued bonds are extremely unlikely to bear a large mortgage spread as some commentators have suggested. These newly issued bonds would be backed by mortgages with the lowest rates in 70 years, so the likelihood of a widespread refinancing wave at even lower rates is small. Nonetheless, the government could place wording in the prospectuses of new bonds committing not to pursue an expedited refinancing program for those bonds in the future without compensating bondholders if this appeared to be a major impediment to executing the program.
10. According to estimates from analysts at Morgan Stanley and JPMorgan, a large refinancing program such as this one can reduce mortgage payments by $50 billion annually; if done well, we believe it can do even more. Because the reduction is permanent, the program will have a long-term impact on the economy, allowing consumers to spend while still increasing savings.
11. The current federal programs are reaching a limited number of homeowners. Some borrowers at high risk of default have been helped by existing programs such as HAMP (Home Affordable Modification Program) and HARP (Home Affordable Refinance Program), but only $1 billion of mortgages has been refinanced with loan-to-value ratios above 105 percent under HARP, according to JPMorgan’s Matthew Jozoffii; the number of Americans who can benefit from refinancing is far greater than the number who have received it. An extensive refinancing program like our proposal would put money in the pockets of a wide variety of homeowners, a large enough number to stimulate the economy effectively and help borrowers struggling to keep up with their payments who are not in immediate danger of serious default or foreclosure.
12. The HARP program was very well intentioned, but has not performed up to its potential. The program we outline could be considered as HARP version 2.0. HARP was not widely publicized, especially compared to HAMP. Refinancing through HARP requires substantial upfront costs for borrowers, and has important limitations on who can qualify. Borrowers with second liens, impaired incomes, and high loan-to-value ratios are often locked out. Additionally, many borrowers do not know their home value and are reluctant to pay an upfront cost to get an appraisal, particularly if the result may prevent them from being able to refinance. Servicers also failed to push HARP for homeowners, possibly due to extensive paperwork barriers and other implementation costs.
This refinancing proposal is designed to reduce the pitfalls in the existing well-intentioned efforts to help homeowners. The upfront costs of the refinancing program should be minimal, and they would be passed into the mortgage payments. There is no new underwriting required for homeowners or lenders. A fixed fee to compensate lenders and minimal paperwork will encourage servicers to provide refinancing to as many borrowers as possible.
13. From an implementation perspective, “simplicity” is an advantage. Avoiding appraisals and income checks will mitigate the biggest hurdles and costs for servicers and homeowners. The underwriting process itself is one of the most significant hurdles to the existing program. Limits such as LTV caps and income rules, as well as the cost of refinancing and anything more than a simple one-page application, are unnecessary. If limits are placed on the program, underwriting becomes necessary, which would create a substantial delay. As well, the high-LTV borrowers are the ones who likely need the most help and are at the greatest risk of walking away, so that is a critical group for this program.
What about the legal impediments of higher-LTV loans? REMIC (Real Estate Mortgage Investment Conduit) rules appear to limit the ability to re-securitize pools with mortgages with LTV above 125 percent. We see several ways to address this issue. We had a conference call with senior partners at SNR Denton to address this question, and we see a number of practical ways to address this. First, by modest wording changes in the laws regulating REMICs, it is possible to get around this problem; the rule changes would allow REMICs to take higher-LTV loans only if being refinanced from a previous REMIC. Second, the FDIC has already done an equivalent deal earlier this year to securitize high-LTV loans from a failed bank and did so by providing explicit government backing of the credit risk. It is possible to develop a similar solution for the GSEs. Third, because both the Fed and the GSEs themselves own about $2 trillion of the bonds and would also be paid back early, they could devote that money to buying the mortgages backed by high-LTV loans. Thus the REMIC rules would not be an impediment. If it were absolutely necessary to limit LTVs, the government should accept automated appraisals, rather than a physical appraisal. Several companies do these in bulk, so there will be no need to send an appraiser for most borrowers.
Other impediments? Title insurance is required and could be expensive. As it is critical that this program be low-cost, the easiest and best way to implement it would be for title insurers to agree to issue very cheap policies (costing around $250 each) to facilitate the process. The volume of refinancings would make this a feasible request; 37 million times $250 is a lot of money. The costs of reissuing policies by title insurers on mortgages they are already insuring is near zero. If worse comes to worse, legislation could address this issue if title insurance costs became a major barrier.
Some states may have a stamp tax to reissue a mortgage. The program could pay the stamp tax and roll it into the costs of refinancing. The stamp tax in most states, however (such as New York), is only for purchase money mortgages. Servicers would need to redo escrow accounts, but that is feasible and already standard for new refinancings.
14. Servicers would receive a fixed fee for each mortgage they refinance to cover the costs of closing the mortgage and to encourage them to cooperate. Currently, servicers are often discouraged or prohibited from contacting their own borrowers to encourage refinancing. In this program, the government-backed lenders would require them to do so. That change alone could have a drastic impact on take-up of the program. Also, an appropriately set servicing fee would provide economic incentives for servicers to cooperate with the government. Because the government-backed lenders are critical for new business, these lenders could let the servicers know that future business will depend on their success in implementing this program. Finally, servicers would also benefit with lower mortgage payments that lead to fewer defaults because defaults are costly for servicers to manage. So costs for servicers would go down under this program.
15. This program is targeted at the GSE mortgages, especially those homeowners who are current on their mortgages, because they are a group that the government has not addressed so far in the crisis, but that are suffering from poorly performing credit markets. The non-GSE mortgages are very hard to reach. Much of the problem with these 21 million loans revolves around second liens and getting banks to mark to market the true value of these loans (principal writedowns), things that have been impossible to do without strong pressure from banking regulators. Nonetheless, there is no reason to hold up progress on the 37 million GSE mortgages because it is harder to find a solution for the others. If anything, such a widespread refinancing solution for these loans will pressure the banks and servicers to do more for their portfolio and securitized loans, in part by providing a road map for how to accomplish this step.
The program we propose is not a subsidy or a bailout, so renters are not disadvantaged in any way by the program. If anything, they benefit from future deficit reductions from losses when owners walk away from their mortgages.
16. According to an estimate from JPMorgan-Chase, about 20 percent of all GSE loans appear to have second liens. A refinancing program such as this one requires the approval of second-lien servicers. For portfolio second liens, this is easy; the lender can just provide a blanket endorsement. For securitized second liens, there could be impediments to those mortgages refinancing that might require legislation to allow them to endorse the program. However, less than 1 percent of all outstanding mortgages have privately securitized second liens.
17. Recourse laws vary from state to state. In states where first-lien mortgages already have personal recourse to the owner, refinancing the first lien will not change that recourse status. If a mortgage has recourse, the lender may seize some personal assets in addition to the house if the borrower defaults, until the entire value of the loan is collected. The impact of refinancing on recourse status varies by state. Basically, in some states, all mortgages have recourse; in others, recourse requires judicial approval; still others only classify non-purchase mortgages as recourse loans. Borrowers in this last set of states who refinance will see a change in their recourse status, as a refinanced mortgage is not a purchase loan, and borrowers should take this into account as they decide whether to refinance or not. However, the reduced likelihood of default coming from the lower payments should make the threat of banks exercising their recourse right more remote.
18. Private mortgage insurance provides insurance for the lender against losses from mortgages that have less than a 20 percent down payment at origination. Private mortgage insurance policies are typically extinguished when a mortgage is prepaid and a new policy is created. With expedited refinancing, private mortgage insurers would be asked to endorse the existing policy to the new mortgage with a lower rate. Because the lower rate mortgage is less likely to default than the previous loan, the private mortgage insurer would be better off (that is, the insurer is less likely to face a loss and/or foreclosure that would trigger a claim). Thus we expect that mortgage insurers would be supportive of such a plan.
19. Mortgages typically have a series of “representations and warranties” made by an originator who sells a mortgage into a securitization. Among other things, these “reps and warranties” typically give assurances that the lender has followed an appropriate process of verifying the ability of a borrower to pay and that the borrower and the collateral meet the underwriting requirements of the mortgage program. If the GSE finds sufficient evidence of violations of reps and warranties by the original lender, they can seek a remedy that includes asking the originator to buy back the delinquent loan at its par value. However, this right may be extinguished upon refinancing.
Losses from reps and warranties violations are likely to be a relatively small factor for loans that were originated at times when underwriting was poor; between 2005 and 2007. Most mortgage defaults due to violations of reps and warranties occur relatively early in the life of the mortgage, and thus they would have already occurred more than three years later. As well, a number of large originators are now in bankruptcy or liquidation, so claims against these originators will have little economic value.
Sources:
“prevailing rates of 4.3 percent” – St. Louis Fed - http://research.stlouisfed.org/fred2/series/MORTGAGE30US?cid=114
“37 million outstanding mortgages” – Morgan Stanley (Berner et. al.: “Slam Dunk Stimulus”)
“future cost to taxpayers of loan guarantees” – Amherst Securities Group – http://www.housingwire.com/2010/01/13/gses-could-lose-448bn-of-mbs-guarantee-business-says-amherst
foreign holders of agency securities – US Treasury International Capital System - http://www.ustreas.gov/tic/shl2009r.pdf
“the Fed and the GSEs themselves own” – Federal Reserve Flow of Funds - http://www.federalreserve.gov/releases/z1/current/accessible/l210.htm
“reduction in mortgage payments of about $50 billion” – Morgan Stanley (Berner et. al.: “Slam Dunk Stimulus”)
“30 million mortgage refinancings” – Federal Financial Institutions Examination Council / Home Mortgage Disclosure Act - http://www.ffiec.gov/hmdaadwebreport/NatAggWelcome.aspx
(the original numbers I got had 30m, an imputed number got 24m; if we get 30m independently let’s stick with that since we still have data that adds up to 30m. I’ve updated the graph accordingly.)
“expedited refinancing for borrowers with loan-to-value ratios up to 125 percent” – MakingHomeAffordable.gov - http://makinghomeaffordable.gov/borrower-faqs.html#10
“little more than $1 billion of mortgages” – JPMorgan (Jozoff et. al.: “Securitized Products Weekly, 7/16/2010)
“20 percent of all borrowers” – ibid.
“Private estimates of the future taxpayer cost” – Laurie Goodman, Amherst Securities Group, quoted in Diana Golobay, HousingWire, http://www.housingwire.com/2010/01/13/gses-could-lose-448bn-of-mbs-guarantee-business-says-amherst
“government is now financing nearly 19 out of every 20 new mortgages” – Joel Schechtman, Newsweek, “No Consensus on Future of Housing Finance” - http://www.newsweek.com/2010/08/18/no-consensus-on-future-of-housing-finance.html
