Uncategorized

The slow-going refinancing of HECM loans

admin | September 6, 2019

This document is intended for institutional investors and is not subject to all of the independence and disclosure standards applicable to debt research reports prepared for retail investors. This material does not constitute research.

With the MBA Refinancing Index at its highest levels since mid-2016, most investors continue to comb through instruments with some prepayment protection. HECMs may be an underappreciated niche in this market, arguably because many investors may not know the nuances of the program. But the pace of HECM refinancing historically has been slow due in large part to high costs. That hasn’t changed, and, in fact, refinancing could get even harder.

The latest HUD proposal

On Thursday HUD proposed that the HECM-to-HECM refinance program should be completely eliminated. This was included as part of the Trump Administration’s comprehensive housing reform proposal. This would slow prepayment speeds and further improve convexity in the product. This change only requires administrative, not legislative, action so has a reasonable likelihood of being enacted. Of course there will be strong pushback from the reverse mortgage industry so it is not a given that HUD will make this change.

Even if HECM-to-HECM refinancing is allowed to continue, the pace of refinancing historically has been quite slow for a variety of reasons. This should continue to be the case. Investors in HECMs get a product with substantial call protection and the option on a policy change to improve convexity even more.

HECM refinancing is driven by cash-out refinances

The majority of HECM loans are floating rate loans, resetting either monthly or annually and do not have an initial fixed rate period. Therefore borrowers have very little incentive to refinance when rates fall in order to save on financing costs, since the loan’s note rate will naturally reset lower.

Instead borrowers refinance if they can capture more equity from their home. Historically the primary driver is home price appreciation, and the only way to increase the size of the HECM is to refinance it.

Changes made to the HECM program in 2017 open the door to rate-driven cash-out refinances, since lower interest rates lead to larger initial loan sizes even if the home retains the same value. The reason is the way HUD sets the maximum loan size. The program first requires a property appraisal, and that is set as the maximum claim amount, or MCA. The HECM program then determines the amount a homeowner can borrow by applying a factor—the principal limit factor, or PLF—to the MCA. The factor itself is a function of the 10-year Treasury or swap rate and borrower age. Borrowers do not pay interest on HECM loans, so unpaid interest gets continually added to the loan principal balance. At lower interest rates, payments accrue more slowly and therefore the borrower is permitted to borrow a larger portion of the loan. When interest rates drop, slower accrual allows a larger loan balance. And borrowers may refinance to access more of their home equity and take out cash. The result: prepayment speeds could increase.

But cash-out refinancing is expensive

Historical prepayment speeds for HECMs show that these loans exhibit a very flat S-curve. Many borrowers choose not to refinance even when presented the opportunity to refinance due to home price appreciation. The expected S-curve due to changes in interest rates appears flatter than the S-curve for a typical 30-year MLB pool. A 100 bp drop in interest rates might increase speeds by only 3 to 4 CPR, which is an extremely small prepayment response to that rate move.

A major reason for slower prepayment speeds is that the cost of the additional cash provided by the refinance is significantly more expensive than the cash provided by the original loan. The floating rate loan implies the borrower will not save any money from lower financing costs after refinancing. The only benefit is the increase in the current loan amount, but closing costs are paid on the entire amount of the second loan.

Consider a borrower that took out a loan with a $200,000 MCA. The upfront mortgage insurance premium is 2.0% of the MCA and, based on discussions with originators, origination fees might be an additional 1.0%. A 70-year-old borrower with a 4.0% expected rate has a principal limit factor of 0.522 and by the end of 12 months will be able to get $104,400—the program limits the initial draw to 60% in the first 12 months to protect the program. This borrower paid $6,000 to borrow $104,400.

If rates were to fall to 3.5% the borrower’s PLF increases to 0.554. Assuming the MCA is unchanged the borrower could access $110,800 today, a 6.1% larger loan. The borrower shouldn’t have to re-pay the upfront MIP on the refinance but will have to pay $2,000 for origination fees again. But the new loan is only $6,400 larger and the closing costs consume almost a third of the total increase.

Put differently, the incremental borrowing provided by the cash-out refinance is almost seven times more expensive than the original loan and many borrowers won’t pay such an exorbitant fee.

Industry rules protect consumers from non-economic refinances

Since a HECM refinance is so expensive the National Reverse Mortgage Lenders Association (NRMLA) has established consumer protections to prevent abusive refinancing. These protections also provide additional prepayment protection to investors. But it is important to note that even without these protections prepayment speeds historically have been slow.

Seasoning requirement

NRMLA imposes an 18-month seasoning requirement before a loan can be refinanced. This is significantly longer than the roughly 6-month limitation imposed by the FHA, VA, and various GSE lenders on single family refinancing.

Closing cost test

NRMLA also requires that a refinance must pass a closing costs test. The test requires that the increase in the current principal limit of the new loan over the old loan must be at least five times larger than the total closing costs including all origination fees plus any upfront MIP, although the latter is usually zero. This test imposes a significant hurdle.

For example consider a 70-year-old borrower with a HECM that pays a total of 1% of the MCA to refinance. For simplicity assume the borrower is able to refinance immediately, ignoring the 18-month seasoning requirement, and that the MCA does not change. In this case the PLF must increase by 0.05, or five times the closing costs, which generally requires that rates must drop more than 75 bp. If the original loan used a 4.0% expected rate and received a 0.522 PLF then the new loan’s PLF would need to be at least 0.572. A new rate of 3.125% has a PLF of 0.576, but the PLF is too low for any rates higher than 3.125%.

As a borrower ages the PLF increases but the benefit from a lower rate decreases. This means the closing cost test requires larger rate declines for older borrowers, and that the prepay protection increases as a loan seasons.

Loan proceeds test

Finally, NRMLA requires that the increase in principal limit from the old to new loan, less closing costs, must be at least 5% of the new loan’s principal limit. In most cases the loan proceeds test won’t matter—a loan that fails this test will also fail the closing cost test.

However, when closing costs are low this test could become the binding test. Consider the limiting and unlikely example of a completely free refinance. The closing cost test can be passed trivially since any PLF increase exceeds zero closing costs. But the loan proceeds test still requires the new loan to be roughly 5% larger than the original loan.

Returning to the example of a 70-year-old borrower with a current loan carrying a 4.0% expected rate, a 1.0% closing cost implies the new PLF needs to be at least 0.560 [=(0.522+0.01)/0.95], which occurs with a 62.5 bp lower rate. This is less than the rate decline required by the closing cost test. But at zero closing costs the new PLF needs to be at least 0.550, which occurs at a 50 bp lower rate. Therefore this test ensures that a baseline rate improvement must always be achieved even if refinancing becomes cheaper.

Conclusion

The HUD has proposed to eliminate the HECM-to-HECM refinancing program, which would further improve convexity of these pools. This policy change would not require legislative action so has a reasonable chance of being enacted. Investors in these pools are buying an option on this policy change, with limited downside if it doesn’t come to pass.

Even if HECM refinancing is permitted to continue refinancing speeds have historically been very tame, since the primary reason to refinance is to do a cash-out refinance, and that is a very expensive way to access additional home equity. Even when presented with the opportunity to refinance most borrowers don’t choose to do so.

The changes made to the HECM program in 2017 raise the possibility that lower rates could entice borrowers to do cash-out refinances. However, the expected response should be quite low, on-par with some of the best call protection stories in standard MBS pools.

Furthermore, the reverse mortgage industry has enacted a number of consumer protections regarding when a loan can be refinanced. These enhance the significant prepayment protection already offered by these loans.

Finally, the nature of the HECM program is that older borrowers receive less benefit from lower rates—for the same drop in interest rates the PLF increase is smaller for an older borrower. Therefore the convexity of these loans improves as they season.