HECMs: Are We Still In Trouble?

Part I: The Ghost of Principal Limits Past

Three months have passed since we published the last installment of our three-part “Trouble With HECMs” blog, enough time for the dust to settle on a number of fronts. First, FHA made a major change to its Home Equity Conversion Mortgage (HECM) reverse mortgage program, lowering by 10% the Loan-to-Value (LTV) ratios (or Principal Limits, in HECM parlance) for all HECMs originated after September 30, 2009. Second, FHA released its Annual Management Report, in which it made the alarming disclosure that the capital ratio for its main fund, Mutual Mortgage Insurance (MMI) fell to a scant 0.53%, below its required 2% floor. FHA maintains that the recent changes they have implemented will improve its financial position, such as the raising of credit standards and the exclusion of non-compliant lenders, among others. Finally, Ginnie Mae’s HMBS securitization program experienced rapid growth, and is now the mainstay of the reverse mortgage secondary market. All of these developments have wide-ranging implications for the reverse mortgage industry, causing significant changes in the volume, product, and pricing of new reverse mortgages.

This entry is the first part of another three-part blog sorting through these events. Part I will review the events of the recent past and discuss their implications for the risk profile of the HECM program in the current economic climate. Part II will delve into the present financial situation of the HECM program and FHA in general, as reflected in their recent Annual Management Report. Part III will present our recommendations for the future.

In a nutshell, our main recommendations were first, keep the old HECM but fix it, and second, implement “HECM II,” a low cost, lower LTV HECM product. We also proposed replacing the Servicing Fee Set Aside (“SFSA”) with a Tax and Insurance (T&I) Set Aside. The SFSA has proved more trouble than it’s worth, whereas HECM investors are increasingly concerned about T&I defaults. With the LTV reduction, FHA has implemented the first recommendation; FHA has also announced that they are studying the HECM II concept, or “HECM Mini,” as they call it. A less fortunate development is that it is becoming more clear that FHA will suffer significant losses from the 320,000 or so high-LTV HECMs originated during the peak years of residential property values. The first order of business, therefore, was to reduce the extremely high LTVs in the HECM program.

As of October 1st, the beginning of FHA’s Fiscal Year (FY) 2010, FHA lowered HECM LTVs across the board by 10%. Therefore, a HECM borrower who could previously take out an 80% LTV HECM can now borrow no more than 72% LTV, borrowers who would have qualified for a 60% LTV will now get 54%, and so on. This necessary and appropriate action will go a long way to reducing FHA’s, and therefore the taxpayer’s risk, for reasons we discussed at length in our previous blogs. Suffice to say that it addresses our previous concerns and should allow the program to break even at lower home price appreciation rates in the 2% to 3% per annum range.

FHA’s action capped a series of significant changes that permanently altered the shape of the reverse mortgage industry. To summarize the sequence of events: the Housing and Economic Recovery Act of 2008 (HERA) raised the loan limits for HECM Maximum Claim Amounts (MCA) to $417,000 and moved all new endorsements for the HECM program from FHA’s General Insurance (GI) Fund to the MMI Fund starting in FY 2009. In February 2009, the American Recovery and Reinvestment Act of 2009 (ARRA) raised the MCA again to $625,500, which spurred a record fiscal year of HECM origination, just barely in number of loans, but by about 20% in loan balances. FY 2010 began with the 10% LTV reduction. Of course, this all took place with the backdrop of extremely difficult economic conditions, including a continuing fall in home prices.

All these changes have effectively created three groups of HECM loans: pre-FY 2009 HECMs, originated at the old (higher) LTVs, but lower loan limits, and classified in the GI fund (let’s call this “Group 1”), the FY 2009 HECMs with the higher LTVs, higher loan limit and placed in the MMI fund (“Group 2”), and HECMs endorsed on or after October 1, 2009, with lower LTVs, higher loan limit and placed in the MMI fund (“Group 3”). These three categories have very different risk profiles. Group 1 clearly poses the most risk to FHA, with loans originated at higher LTVs on properties appraised during the peak of the residential housing boom. Group 2 benefits from the more modest appraisals this year, but still has the high LTVs. Group 3 is the safest for FHA: lower LTVs on lower property values. The lower MCAs and lower initial utilization rates of Group 1 do mitigate its risk, but this is almost certainly overwhelmed by the other risk factors, as we discuss below.

Group 1 comprises about 320,000 of the approximately 450,000 HECM loans outstanding. Group 2, the FY 2009 production, has approximately 125,000 units, and Group 3 is the future. Pre-2002 production was tiny to begin with (only about 50,000 HECMs were originated in that period), and very few are still outstanding.

Home prices have generally fallen to 2002 levels, about 30% below their peak in 2007. That means essentially none of the loans in Group 1 have benefitted from price appreciation; their loan balances have grown as their property values have fallen. A 5.5% HECM loan originated in 2002 with a 60% LTV that has the same value 8 years later, is now a 92% LTV loan. A 2007 5.5% HECM loan that began at a 60% LTV on a property that has declined 30% in value is now underwater, especially taking the cost of property disposition into account (more on that in Part II).

Loss Mitigation vs. Lost Mitigation
Defenders of the old regime sometimes insist that there are overlooked mitigating factors to FHA’s HECM risk. Chief among these are utilization or draw rates, and loans with excess equity (that is, loans with property values exceeding their Maximum Claim Amounts). However, as we shall see, this loss mitigation looks more like lost mitigation when examined in light of the housing bust, changes in the HECM product, and borrower behavior.

Many adjustable rate HECM borrowers do not initially borrow the full amount available, instead opting to draw the remaining amount, or Net Principal Limit, over time. For example, a borrower with a $200,000 house may elect to borrow $100,000 initially and leave $40,000 on a line of credit. The line of credit can then be drawn down any time in the future. In the meantime, the unused portion grows each year at the loan interest rate.

This loan is initially less risky to FHA, but borrowers do make their draws. According to HUD data, line of credit draws average an additional 15% of the Net Principal Limit in the first year and nearly as high in the remaining years, so that the average Net Principal Limit rapidly approaches zero, even with the growth feature. It’s no surprise then that most loans put back to HUD are fully drawn, or nearly so, with utilization rates exceeding 90% for ARMs, according to one study. Of course, fixed rate HECMs are fully drawn from day one.

Bear in mind that the borrower who does not take down a high percentage of proceeds at closing, and also does not make a lot of subsequent draws, has a rather high cost loan. Using our example, the 2% Initial MIP alone represents a 4% fee on the loan balance. FHA may not lose money on this loan, but it is a pyrrhic victory, at a high cost to the senior homeowner. Even if there were enough of these loans, it would mean that one group of borrowers is heavily subsidizing another group. This is not the intention of the HECM program, which seeks to equalize risks and benefits. It also illustrates the unfairness of charging the same upfront insurance fee to each and every borrower, despite their widely variable degrees of risk.

Before the HECM loan limits were raised, FHA capped the Maximum Claim Amounts at much lower levels, from $200,160 to $362,790 depending on geographic location. But unlike now, borrowers then had the option of taking out jumbo mortgages, both forward and reverse. The forward jumbos offered generally higher LTVs and the reverse jumbos offered much lower upfront costs, albeit with generally higher interest rates. Due to its much higher LTV, the HECM was able to compete against the jumbos in the $300,000 to $600,000 property value range, when the borrower felt the HECM’s higher proceeds and lower interest rate outweighed the higher upfront fees. Consequently, a number of HECMs had some excess equity at origination. Since their underlying value exceeded the HECM loan limit, their real LTV was lower than that of other HECMs with the same borrower age and expected rate. For example, according to the General Accounting Office, some 42% of HECMs originated in 2006 were originated with some excess equity.

However, that was before the crash in residential home values. Most of these excess equity loans were originated in California and other regions that were hit hardest by the crash in home prices. So how much, if any, of that excess equity remains?

Recent borrower behavior provides a clue. When HECM limits were raised to nearly double their old amount, some predicted a wave of reverse mortgage refinancing. For the jumbo loans, this came true. Many jumbo borrowers took advantage of the lower rates and higher LTVs that were suddenly available for homes in their price range. Jumbo prepayment rates climbed from an average of 9% CPR (that is, the per annum rate) in October 2008 through March 2009 to 19% in April through September 2009, thereby recovering back to their approximate historical averages. But for HECMs, the refinancing boom was a refinancing bust; lenders reported significantly fewer HECM refinancings than expected. If there were so many borrowers with excess equity, the industry should have seen a much higher number of originations in FY 2009. Instead, unit volume was only slightly higher than 2008. The most plausible explanation: the excess equity was modest to begin with and has since vanished. That HECM borrower in California who borrowed $250,000 in 2007 on a home appraised at $475,000 might now owe $300,000 (with draws and interest roll-up), with little if any equity remaining.

With the HECM loan limit now at $625,500, the vast majority of new loans will have a Maximum Claim Amount equal to their appraised property value. The mean average Maximum Claim Amount jumped from under $220,000 in FY 2008 to over $260,000 for loans originated in FY 2009. The same GAO analysis that shows the percentage of 2006 HECMs with Maximum Claim Amounts capped by the loan limit at 42% also shows that percentage dropping to 25% in FY 2008, and to only 18% for the first four months of FY 2009. Therefore, it seems the percentage of Group 1’s excess equity has all but vanished (for the minority of loans which had any), and for Groups 2 and 3 it will scarcely exist.

With each passing year, and after massive home price declines, the slow prepayments, higher draw rates, and the negative amortization built into reverse mortgages, it seems less and less likely that by the time these HECM loans mature they will have excess equity or undrawn credit lines to cushion the blow to FHA.
But very few of these loans have paid off or even been put back to FHA yet … which brings us to Part II.