Mortgage Refinancing Proposals: Defining the Objective

A recent New York Times article suggested that the President may launch a new mortgage refinancing initiative as part of next week’s speech to announce a “pivot to jobs.” Current mortgage markets are so dysfunctional that a reaction to the trail balloon seems to depend, in part, on what specific aspect of the dysfunction the proposal is attempting to address. The current situation also provides a reminder of the problems inherent with the 30-year fixed rate mortgage and why it would be no great loss if eventual privatization of the mortgage markets made this instrument less ubiquitous.

There are two basic problems a mortgage refinancing program can seek to address:

1. About one-quarter of all mortgage borrowers are underwater, which means the outstanding principal balance on their home mortgage(s) exceeds the market value of their home. Households in this situation have an incentive to walk away from their property because doing so would increase their wealth (they shed liabilities worth more than the corresponding asset). Households that would prefer to remain in their house, even if they’re underwater, may not be able to do so if the primary income earner loses a job.
2. The economy is struggling. A huge portion of household cash flow is used to meet the monthly mortgage payment. Reducing that mortgage payment through a downward adjustment in the effective rate paid on the mortgage would allow households to increase consumption of other goods and services.

The Obama Administration has tried to address (1) through a number of refinance programs under the “making home affordable” banner. In total, more than 1.6 million mortgages have been modified. These mortgages are generally underwater, delinquent, and modified through a reduction in the interest rate or extension of amortization schedule rather than principal forgiveness. Fannie Mae reports that the average modification saves the borrower about $487 per month. That’s equal to about 15% of average monthly earnings. While one might anticipate that would be enough to make the program a success, improvements in cash flow generally don’t solve a balance sheet problem. Only about half of trial modifications have made it to the permanent modification stage. And only about half of the permanent modifications are still performing after 15 months (page 15).

In the absence of principal forgiveness, it does not seem reasonable to anticipate a new refinance program to stop additional foreclosures. If a household owes $300,000 on a house worth $200,000, the problem is not the interest rate. It’s the loan balance. The interest rate could be modified to zero and the effective interest rate would still be 50% (the monthly payment on a 0% interest rate 30-year fixed rate mortgage with a $300,000 loan balance would be $833.33, which is 50% more than the $555.56 monthly payment on a 0% interest rate 30-year fixed rate mortgage with a $200,000 loan balance). In short, if a new refinance proposal on this front is to be judged on the merits of “keeping people in their homes,” it is almost certain to fail.

If the proposal is based instead on improving household cash flow more generally, it makes a great deal more sense. Before the mortgage crisis, refinancing was generally understood to be something people did in response to lower interest rates (as opposed to something to extract equity or to get into “the right” mortgage). The option to prepay allows homeowners with a 6% mortgage, for example, to pay off their existing loan balance by getting a new mortgage when rates fall below 6%. For a $300,000 loan, this would save nearly $200 per month if the rate fell to 5%.

However, the freedom to prepay the current mortgage is not the same thing as having a lender that is willing to make a new loan to allow you to prepay. Since the mortgage loan is secured by the house as collateral, an underwater borrower generally cannot refinance unless he or she is willing to pay down the underwater portion of the mortgage. In the case above, this would mean a homeowner would have to write a check for $100,000 to pay the balance down to $200,000 so he could save a few hundred dollars per month in lower interest costs. For this reason, the initial refinance program had to be extended to current mortgages with loan-to-value (LTV) ratios of 125% because it was impossible to issue new mortgages to refinance old ones if these new mortgages were limited to the value of the house. Again, making a mortgage eligible to be refinanced is not the same as having a lender eager to make a new loan to replace the existing one in these situations.

In addition to underwater borrowers, there are many households who simply cannot refinance because of tougher credit conditions. Lenders – including Fannie Mae and Freddie Mac – have toughened underwriting conditions substantially. New loans to purchase homes or refinance existing mortgages have lower average LTVs and higher FICO scores today. A borrower able to secure a loan in 2007 might not qualify for the same loan now. As a consequence, the borrower may be locked into the existing mortgage despite lower rates.

What is the impact of these frictions and how are they inhibiting more refinancings? The table below compares the effective mortgage rate – the average rate paid on existing mortgages – to the “new” mortgage rate, which is the average rate charged on a 30-year fixed rate mortgage a well-qualified borrower could refinance into, and the Fed policy rate, which is the fed funds overnight lending rate controlled by open market operations. The table below that measures the spread between the effective mortgage interest rate and these two other rates.

Today, the effective interest rate on mortgage debt is 5.28%, which generates $548 billion per year in mortgage interest payments. The average new interest rate was only 4.66% in the second quarter and has since fallen by 0.3%. If the effective interest rate on outstanding mortgages were reduced to 4.66%, mortgage interest payments would fall by $67 billion per year. Taking the rate down to the average spread over the Fed policy rate a 4.14% effective mortgage rate) would reduce annual mortgage payments by $120 billion. This would be a huge stimulus for the economy, equal to between 0.5% and 1.1% of disposable personal income.

Closing the gap between effective interest rate and policy rates would also aid monetary policy. The Fed has judged the economy to be so weak as to require a 0% lending rate. But if frictions in mortgage finance markets prevent this rate (or something close to it) from being accessed by household borrowers, the policy rate becomes disconnected from any concrete economic reality. A refinance program that eliminates these frictions by automatically refinancing existing mortgages into a new, low rate would improve the monetary policy transmission mechanism and make current Fed policy far more stimulative. The challenges confronting underwater borrowers for whom the principal balance is the problem would remain, but the economy would receive a huge infusion of disposable income.

Although some commentators wrongly argue that there are no costs associated with this proposal, this is simply not true. A mass refinancing would impose large costs on the existing owners of mortgage assets, whose interest income would fall by the same amount as the mortgage borrowers’ income would rise. Although default would be averted in some situations, as a lower mortgage payment would make liquidity shortages less likely, in general it’s quite clear that servicers and investor much prefer to keep current mortgages active rather than aiding in a refinancing wave. When a mortgage pays a coupon that’s above the current interest rate, it sells at a premium to par value. Refinancing is effectively buying this mortgage back for par, which means that the investor loses the value of the premium. Since the largest holders of mortgages eligible for refinancing continue to be Fannie Mae and Freddie Mac, they’d also suffer the largest mark-to-market losses from a massive refinancing wave. Assuming these mortgages have a current market value of 5% above par value, a $2 trillion refinance wave – 10 million mortgages with average outstanding balances of $200,000 – could result in $100 billion of fair value losses. The GSEs would have the largest share of this total. Is this something the Obama Administration really wants to push onto these wards of the state?

It is interesting to consider how such a radical proposal would not be necessary if mortgage finance weren’t dominated by the GSE-backed 30-year fixed rate mortgage. Although the fixed rate mortgage is supposed to be a hedge against rising rates, the refinance rate has been lower than the effective mortgage rate in 37 of 42 quarters since 2001 (88% of the time). Rather than protecting borrowers from rising rates, the 30-year fixed rate mortgage has created transaction cost obstacles to gaining the benefits of more accommodative monetary policy that would flow naturally to households if more mortgages had adjustable rates. The average refinancing costs 1.5% of the mortgage balance – or $4,500 on a $300,000 mortgage – which must either be added to the mortgage balance or paid out of pocket. There are also significant non-pecuniary costs, such as identifying a lender and collecting required documentation. Moreover, underwriting is counter-cyclical, so credit terms are tougher precisely when households would be most likely to refinance (i.e. loan terms are often more strict when interest rates are low, as everyone is now realizing).

The GSEs argued that a key public benefit they provided was the 30-year fixed rate mortgage. But leaving aside their credit-related collapse, do these mortgages really deliver a public policy benefit? The interest rate risk doesn’t disappear, but is rather transferred to taxpayers due to their backing of the GSE balance sheets. Imagine instead if prime mortgages had an interest rate equal to the prime loan rate. Today, the average effective rate on these mortgages would be 3.25%; even if a margin of 0.5% were added, the boost to household income relative to today would run into the hundreds of billions of dollars. Does it really make sense, going forward, for taxpayers to assume huge interest rate risk exposure to ensure households have access to an instrument that inhibits monetary policy’s effectiveness by erecting barriers to lower rates?

The refinance proposal could be a modest success if the objective is improving household cash flow rather than addressing the problem of underwater borrowers. The costs would be real, however, and pretending they don’t exist is disingenuous. Yet, it’s interesting to consider how this stimulus measure wouldn’t even be necessary if not for the GSE dominance of mortgage lending. Just add the current conundrum to the cost of on-going federal “support” of mortgage finance.