The reverse mortgage industry prides itself as the lenders of the Good Mortgage: the mortgage that not only provides essential retirement funds but also the mortgage that comes to the rescue, preventing foreclosures by paying off old “forward mortgages,” delinquent property taxes and unpaid insurance bills. A growth sector in an otherwise ailing mortgage industry, reverse mortgages reward investors with lower prepayment risk and superior performance in securitized form.
Until recently, the reverse mortgage held another distinction: the mortgage that actually makes money for the federal government, a “negative subsidy” of positive cash flow that enabled the expansion of the Federal Housing Administration’s Home Equity Conversion Mortgage (“HECM”) reverse mortgage program during a time of otherwise considerable financial difficulty. However, FHA’s recent announcement that the HECM program is expected to lose over $800 million in FY 2010 has exposed a weakness in this relatively small but rapidly growing sector. (With the recent demise of virtually all proprietary or “jumbo” reverse mortgage lending programs, HECMs comprise the vast majority of reverse mortgage loans originated in the U.S.) The Trouble with HECM is that the program encourages senior homeowners to borrow a very high percentage of their home value, which not only increases FHA’s losses, but also unnecessarily drives up costs to borrowers, lenders and investors. The lack of a more efficient, low-cost alternative has also discouraged many senior borrowers who might otherwise benefit from a HECM loan, but are discouraged by the size of the upfront fees. This three-part discussion seeks to outline the dimensions of the Trouble with HECM and concludes that the problem is fixable, if the FHA acts quickly enough.
FHA insures the investor against “Crossover Loss,” that is, the risk that the negatively amortizing HECM increases to the point where it “crosses over” the value of the home securing the reverse mortgage. FHA insures the HECM investor against crossover loss by purchasing the loan (at par plus accrued interest) from the investor when the HECM loan balance causes the HECM to reach a 98% loan-to-value ratio (“LTV”). The value used as the denominator in this formulation is called the Maximum Claim Amount (“MCA”), equal to the lesser of the property value or the FHA lending limit at the time the loan is originated. When the investor exercises this put, FHA effectively steps into the shoes of the investor and bears the crossover risk. When the loan finally does pay off, FHA’s losses are equal to the excess of the loan balance over the net property value.
In return, the HECM loan investor pays FHA a Mortgage Insurance Premium (“MIP”) equal to a whopping 2% of the Maximum Claim Amount at the time of loan origination plus 0.50% per annum, based on the unpaid principal balance of the HECM. In other words, a borrower whose home value is equal to or exceeds the FHA maximum of $625,500 is charged a $12,510 upfront MIP at closing, in addition to any other fees. These amounts are added to the loan balances, so the borrower ultimately bears these costs.
From the HECM program’s inception in 1990 until this year, these substantial premiums, combined with rising home prices, consistently resulted in ever-increasing positive cash flow to FHA. In its first decade, the program grew slowly from 157 loans in its first year to a mere 6,600 loans in 2000, but then grew rapidly, with a record of over 112,000 loans originated in FY 2008. Year after year, FHA reaped the additional benefit of collecting an ever-growing amount of upfront MIP from the new loans while the older vintages, with the help of ever-increasing home prices, largely avoided crossover loss. But beginning in 2008, that trend reversed abruptly: The HECM program’s profit (or negative subsidy) declined to $462 million in FY 2008, down from $697 million in FY 2007.
FHA projects that HECM originations will exceed $30 billion in FY 2010. They further project, using the Office of Management and Budget’s base case assumptions on home price appreciation, a positive subsidy (similar to a loss reserve) of nearly $800 million — in other words, a negative swing of $1.5 billion in a program with approximately $50 billion in outstanding loans. $800 million is nearly equal to the entire amount of all upfront premiums collected in the history of the HECM program from inception to 2006, and nearly one fourth of all of the premiums, monthly and upfront, ever collected. This $800 million does not include losses on HECM loans outstanding from previous years of HECM origination. This reversal of fortune follows a year, FY 2009, where FHA will collect nearly $1 billion of upfront MIP, due to another year of record new lending volume (over 150,000 loans projected) and much higher balance loans. (The latter is the result of the very large increase in FHA’s HECM lending limit mandated by the Housing and Economic Recovery Act of 2008 and the Economic Stimulus Package of 2009). The total dollar volume of new HECMs could exceed $20 billion in FY 2009. But despite these record premiums, FHA’s results are far different from the banner years of the past.
Declining home prices have played a part in these losses, just as they have wreaked havoc on every mortgage sector. But there is another culprit: the astonishingly high LTVs of HECM loans. Each HECM’s LTV is determined by borrower age and interest rate. The older the borrower, or the lower the interest rate, the higher the LTV. With interest rates hovering at or near FHA’s minimum value for interest rates (5.56%), a 72 year-old HECM borrower can receive as much as 71% of the value of their home, in an upfront lump sum if they wish. An 80 year-old qualifies for a 78% LTV loan and a 90 year-old 86%. This is for a loan with no monthly payments, with negative amortization from day one! Not every HECM borrower elects to borrow the maximum amount in one initial lump sum, but many do so, and are doing so in increasing numbers.
The HECM program began at the dawn of a great bull market in residential real estate. From the year of the program’s inception to 2006, home prices rose at an average of about 6% per year, outpacing the historical average of 4%. With such rapid home appreciation, it’s easy to see why the HECM program was so profitable. Even beginning at an 80% LTV, a loan accreting at 6% per year will take about 12 years to reach the crossover point if the underlying property is appreciating at 4% per year. The same loan will take 19 years to reach crossover starting at 70% LTV, 27 years starting at 60% LTV, and so on.
However, reverse that trend and the numbers change dramatically: without home price appreciation, the 80%, 70%, and 60% LTV loans reach the crossover point in about 4, 7, and 8 years, respectively, well under the expected average life of HECM loans. Bear in mind that we are not even considering property disposition costs, which both hastens the arrival and magnifies the severity of the crossover loss.
Needless to say, if home prices decline at the double-digit rates of the past two years, the homeowner’s equity is wiped out very quickly, and FHA is deep in crossover loss territory. If history were a smooth curve, the HECM program would have continued its smooth sailing. But history is a bumpy road.
A more conventional view of risk, and the approach employed in the design of private-label reverse mortgage, takes into account not only the worst-case scenario for a particular variable (like home prices), but also the correlation between risk factors. For example, the current housing crisis involves not only dramatic declines in home prices, but also lower prepayment rates, resulting from lower mobility and reduced ability to refinance. In a downturn, these correlations magnify losses: the crossover point is closer and fewer loans pay off before that loss threshold is reached. Standard & Poor’s requires, among other criteria, that a AAA-rated reverse mortgage bond be able to withstand a decline in home prices of approximately 30% over 3 years, prepayment rates of about half of the usual rate and sharply higher interest rates. This necessitated the much lower LTVs for the jumbo products. As a consequence, the five outstanding private label reverse mortgage securitizations have thus far weathered the housing storm, and continue to pay back bondholders.
With the rapid decline in home prices, FHA’s reversal of fortune is not surprising, but it also has a lot to do with where the HECM starts out. The majority of the fully-drawn HECMs originated in the peak valuation years of 2005 to 2007 is probably underwater, with 20% – 35% home price declines combined with two to three years of negative amortization removing any excess home equity. For those vintages, the only HECMs loans with significant home equity remaining are those whose home value was substantially higher than the Maximum Claim Amount (recall that FHA limits were much lower then), or where the borrower elected to borrow much less than the maximum LTV. Even then, with the combination of borrower draws and negative amortization increasing the loan balance, and home price declines decreasing the value, this equity cushion is disappearing in a hurry.
Of course, the loan balances for these loans already past the crossover point will continue to increase through negative amortization, draws, and servicing fees; all this subsequent increase piles on more losses. With so many loans from past vintages already underwater, FHA is effectively losing much if not most of the accrual from past vintages, the vast bulk of which were originated during the peak years of home prices. Estimating these losses is tricky, because it is not a simple cash flow equation. The HECM program probably still has positive cash flow due to the collection of the upfront MIP in a rapidly growing market, as we noted above. However, the 2% upfront MIP is meant to cover losses over the life of the HECM loan. Consequently, it should not be fully credited upon collection, but instead amortized over the life of each loan.
FHA’s loss estimate for 2010 seems reasonable. Assuming that $30 billion of HECMs are originated, the FHA estimate implies a 2.66% net loss, in present value terms ($798 million = $30 billion times 2.66%) . Using New View Advisors’ HECM cash flow model, and assuming a typical distribution of borrower ages, prepayment speeds, and loan-to-value ratios for HECM loans, and a home appreciation rate of about 2.9% per annum, the FY 2010 vintage would suffer about $19 billion in losses over its entire life. However, these losses would not be significant for about 6 years. Using a 6% discount rate that approximates an average HECM interest rate, the present value of these losses is about $4 billion. Offsetting those losses are over $1 billion in upfront MIP and the ongoing MIP, totaling about $5.5 billion over time and about $2 billion in present value terms.
Using these numbers, one arrives at FHA’s estimated $800 million loss:
$4,000 million (Crossover losses)
– 2,000 million (Monthly MIP)
– 1,200 million (Initial MIP)
____________
$ 800 million
This approach is very “back of the envelope,” but is an altogether plausible picture of where FHA stands today. In a subsequent blog, we will dive into these numbers further.
As for the future, HECMs originated after 2008 are based on much lower appraisals, but the effective initial LTV is much higher than previous years, because of the higher loan limit. A return to the home price appreciation of five years ago seems unlikely; home prices could fall even further.
What does not seem possible is a return to surplus in FY 2011. Every factor determining crossover loss is going in the wrong direction: home prices, prepayment rates, interest rates, monthly draws, LTVs, initial draws, and cost of disposition, which is driven significantly higher by foreclosures, and tax and insurance (T&I) defaults. Unless something changes in the housing market or the HECM program, the problem will get worse before it gets better. More loans will pass the crossover point and underwater loans will sink deeper underwater. Also, what if interest rates go up? What if home prices decline further? These could result in even higher losses.
Recall that the root of this problem is the amount lent to the borrower in the first place. This “Principal Limit,” or Loan-to-Value ratio means that even with flat home prices, the crossover point can be reached in short order. Therein lies the problem and solution to this dilemma. In Part II, we will attempt to quantify the problem more specifically for this year and subsequent years, and project the magnitude of the problem under various assumptions and scenarios. In Part III, our final installment, we will propose a new HECM choice for senior homeowners, a “HECM II” that substantially reduces FHA’s risk, and removes the high cost stigma unfairly attached to reverse mortgages. With proper product innovation, the HECM program can achieve long-term viability and serve the interests of senior homeowners in a more efficient and sustainable way.