A Primer on HECM Loans In September, RiskSpan announced the addition of Ginnie Mae’s loan-level Home Equity Conversion Mortgage (“HECM”) dataset to the Edge platform. The dataset contains over 330,000 HECM loans with origination dates from 2000 to 2018 and reporting periods from August 2013 to October 2018. This post is a primer on HECM loans, the HMBS securities they collateralize, and the structure of the new dataset. What is a HECM? HECMs are FHA-insured reverse mortgages that provide people 62 and older with cash payments Read More.. https://riskspan.com/a-primer-on-hecm-loans/ In September, RiskSpan announced the addition of Ginnie Mae’s loan-level Home Equity Conversion Mortgage (“HECM”) dataset to the Edge platform. The dataset contains over 330,000 HECM loans with origination dates from 2000 to 2018 and reporting periods from August 2013 to October 2018. This post is a primer on HECM loans, the HMBS securities they collateralize, and the structure of the new dataset. What is a HECM? HECMs are FHA-insured reverse mortgages that provide people 62 and older with cash payments or a line of credit in exchange for equity in their homes. Borrowers are not liable to make any payments on HECM balances until the house ceases to be their primary residence. In contrast to traditional mortgages that amortize down over time, reverse mortgage balances usually grow over time as accrued interest is added to the loan. The Federal Housing Administration (FHA) insures HECM lenders against default and loss and is paid a mortgage insurance premium in return. Because borrowers do not make principal and interest payments, the concept of HECM default differs from that of traditional forward mortgages. HECM default most commonly occurs when borrowers fail to keep current on property tax payments and insurance premiums or otherwise jeopardize the lender’s lien position on the property. Initial loan-to-value (LTV) ratios for HECMs average between 60% and 70% to allow for the balance to grow over time (taking into account borrower age and interest rate). The number of borrowers is arguably a more important factor when predicting HECM performance than when predicting traditional mortgage performance. Because reverse mortgages do not become due until all borrowers have left the property, reverse mortgages with multiple borrowers tend to have longer tenures—and consequently run a higher risk of growing beyond the point where the balance and accrued interest are supported by the underlying property’s value. Like traditional mortgages, HECM interest rates may be fixed or adjustable. Fixed-rate HECMs disburse a single, initial advance, while adjustable-rate HECMs combine a line of credit or monthly advance with an initial advance. Figure 1 (below), which was constructed using data from the newly available dataset, illustrates a steady increase in the share of ARM loans since 2013. Figure 1 One net result of this trend is fewer one-time lump-sum distributions and more line-of-credit (LOC) distributions over time. LOCs give borrowers access to a source of funds that they can draw upon as needed. While LOCs constitute (by far) the most common type of HECM, two other loan types—“term” and “tenure”—also occupy the HECM landscape. “Term” loans provide monthly payments for a set period of time. “Tenure” loans provide monthly payments for as long as the borrower lives in the home as a primary residence. The lender receives principal, interest and possibly a share of the home appreciation upon expiry of the fixed term (in the case of term loans) or upon borrower’s death or move-out (in the case of either loan type). The dominance of the LOC loan type relative to term and tenure HECMs is depicted in Figure 2, below. Figure 2 Fannie Mae had traditionally functioned as the primary investor in reverse mortgages for most of these loans’ 25-year existence. Since 2009, however, Fannie Mae has significantly scaled back its reverse mortgage portfolio, leaving the majority of the reverse mortgages to be picked up by the Ginnie Mae HMBS market. What is a HMBS? HECM loans are pooled into HECM mortgage-backed securities (HMBS) within the Ginnie Mae II MBS program. HMBS are made up of a pool of participations in the HECM loans. A participation in a HECM loan is a pro-rata share of the loan that is securitized in a HMBS. As explained above, many HECM loans are structured as a line of credit, which allows borrowers to draw on their lines as needed. When these draws occur, the drawn-down loans become a smaller pro-rata share of the loan and the participation balance doesn’t change. HMBS participations have a mandatory repurchase clause requiring a lender to buy back all the participations of a HECM loan when its LTV reaches 98%. For HECM loans, LTV is calculated as a proportion of the current HECM balance against the maximum claim amount. As of June 2018, participation unpaid balance stood at approximately $56.18 billion with 11,380,452 active participations. Figures 3 and 4, below, show the trend of participation composition (by number of participations and UPB) over time. These reflect the shift toward ARM lines of credit (and away from fixed-rate lump sum disbursements) illustrated in Figures 1 and 2. Figure 3 Figure 4 HMBS Dataset Ginnie Mae provides two monthly loan-level files related to the HECMs that collateralize its HMBS offering. One of these files contains fixed-rate and annually adjusting rate loans, and the other contains monthly adjusting rate loans. Because individual security participations are spread across several different pools (often with several column values repeating for a single loan) working with this dataset can be challenging. An example of a single loan spread across multiple security participations is illustrated in the table below. Note that for a single loan ID, the current UPB and Max Claim Amount columns are repeated for each participation. Loan ID Current HECM UPB Max Claim Amount Participation UPB 1000033608 260,784.73 365,000.00 860.70 1000033608 260,784.73 365,000.00 321.87 1000033608 260,784.73 365,000.00 12,079.98 1000033608 260,784.73 365,000.00 483.81 Table 1 The most important risk factors associated with HECMs relate to borrower mortality and mobility (i.e., borrowers’ remaining in their homes until the increasing mortgage balance exceeds the value of the property). Borrowers are more likely to move out of their homes for health reasons as they age, but they become less likely to move out for other reasons. Having more than one borrower tends to extend the life of a HECM because the loan does not become due until the last surviving borrower leaves the property. As of the most recent reporting period, about 43% of the aggregate HMBS balance was associated with HECMs with more than one borrower. In order to calculate HECM prepayment speeds, we look at the zero balance codes provided in the dataset to exclude loans which have reached a 98% LTV from the opening balance. (As noted earlier, loans must be purchased out of the HMBS once they reach this threshold.) Because interest is deferred in HECM loans, it is added to the opening balance. We calculate the total prepayments and obtain the single monthly mortality to calculate the CPR. Figure 5, below, shows the one-month CPR by vintage over the past five years. Figure 5 Because borrower mortality and mobility tend to remain stable over time, HECM prepayment speeds exhibit less variability than traditional mortgages do. An important aspect of evaluating CPR includes looking at the outstanding participation balance relative to borrower age. Figure 6 contains a heatmap plotting borrower age against HECM purpose for the most recent reporting period (July 2018). Figure 6 Because most HECM borrowers are younger than age 80, prepayments are likely to increase as this cohort ages and becomes more likely to move out or pass away. Figure 7 below shows the five largest HMBS originators by participation as of July 2018. As discussed above, lines of credit (LOCs) are the most popular HECM type with Single Disbursement Lump Sum the next most frequent. Stay tuned for future blog posts in which we will use the Edge platform to glean additional insights from this newly available and very interesting dataset. For information on how to use the Edge platform to conduct your own analyses of this or any other dataset, please contact us.