In September 2013, the Federal Housing Administration introduced the Home Equity Conversion Mortgage (HECM—pronounced "HEK-um") that replaced the separate HECM Standard and HECM Saver reverse-mortgage programs. The new program provides some additional protections for the FHA in insuring these loans along with slight fee reductions and adjustments to allowable loan amounts. Funds can be accessed in a variety of ways, including as a lump sum, as monthly payments for as long as the owner continues to occupy the home or as a line of credit. The key advantage over other types of loans is that repayment is not required as long as the owner continues to live in the home – the loan is paid off from home equity when the owner sells the home or dies. Any remaining home value after paying off the loan goes to the homeowner or heirs, and any shortfall is absorbed by the FHA.

The Pfeiffer, Salter and Evensky research focused on the line-of-credit option. The homeowner can use a HECM to obtain a line of credit equal to a percentage of home value (up to a value of $625,500) called the principal limit factor (PLF), which varies with the age of the youngest owner and a term defined as the "expected interest rate." This rate is calculated as the 10-year Libor Swap Rate (2.75% as of early March 2014) plus a lender margin estimated at 2.5% in an illustration provided on the National Reverse Mortgage Association's website, giving a rate of 5.25%. This chart shows a sample of PLFs from an extensive set of tables provided by HUD.

What is immediately apparent from this table is that the amount of available funds decreases rapidly at higher interest rates, although it is level for rates below 5%. The size of the line of credit available will be significantly greater at current interest rates than if rates move up to more normal levels in the future.

The interest rate charged on a HECM line of credit is calculated differently from the Expected Interest Rate in the table above. It varies monthly based on the 1-month LIBOR rate, which is 0.15% as of early March 2013. A lender's margin, estimated at 2.5%, is added to this index, plus an FHA mortgage insurance premium of 1.25%, for a total of 3.90%. This is more expensive than the rates on conventional variable rate mortgages (currently averaging about 2.5%) or home equity lines of credit (with rates as low as 3.0%). However, the reverse mortgage is the only form of loan that allows balances to build without requiring repayment.

Up-front charges

Perhaps the biggest negative about reverse mortgages is the high up-front cost, including appraisal, origination, title search, insurance, inspections, recording, mortgage taxes, credit checks and an up-from mortgage insurance premium of 0.5%. Such fees are likely to total 3% to 4% of the value of homes in the $250,000 range and a bit less for higher home values. Most of these fees can be financed as part of the reverse mortgage rather than paid up front, but the financed fees are deducted from the amount that can be borrowed.

In the context of setting up a standby line of credit to support a retirement withdrawal strategy, the up-front fees can be thought of as purchasing insurance and spreading it over the life of the loan. If the fees are $8,000 and the loan lasts 10 years, that's $800 per year, or $400 annually if the loan last 20 years. It doesn't make sense to use a reverse mortgage if the owner expects to remain in the home for just a few years.

Impact on sustainable withdrawal rates

Let’s go back to the Pfeiffer, Salter and Evensky study. These authors ran 30-year Monte Carlo retirement projections without SRM support for a range of inflation-adjusted withdrawal rates, and they measured outcomes in terms of success rates (not running out of money before the end of the 30-year period) and median ending wealth. Then they ran the same tests with an SRM established at the beginning of the 30-year period and available to provide home-equity funds when needed.

They established rules for when funds would be drawn from the reverse mortgage to support the planned withdrawals. They set up a three-bucket strategy consisting of the investment portfolio (with an assumed 60/40 stock/bond mix), the HECM line of credit and cash. Withdrawals were assumed to be taken from the cash bucket, and the authors established rules for replenishing the cash bucket – whether the funds would come from the investment portfolio or the HECM line of credit.

The determination of where the funds would come from was based on a glide path consistent with the investment portfolio being drawn down to zero at the end of 30 years. The authors worked their way through the 30-year retirement in quarterly increments based on randomized investment projections from Monte Carlo simulations, and at the end of each quarter compared the size of the investment portfolio to the glide path. Running a variety of tests, they determined that they could make best use of the HECM line of credit by drawing on it to fill the cash bucket when the investment portfolio dropped below 80% of the glide path. They noted that using 100% or 90% resulted in overuse of the line of credit and poorer outcomes.

The research assumed an initial investment portfolio of $500,000. Tests were run for home values of $250,000 and $500,000. Tests were also run for low, moderate and high interest-rate environments, which resulted in different HECM lines of credit being made available, as the prior chart on Principal Limit Factors showed. The chart below shows a small sample of results from the study based on a moderate interest rate environment.

The survival rates for the cases in which the line of credit was not available (no SRM) are considerably lower than for well-known financial planning studies that show a 4% withdrawal rate to be very safe based on historical returns. These researchers used lower-than-historical returns, with the 60/40 stock/bond portfolio assumed to earn an arithmetic average of 7.15%, about 2.5% below what historical numbers would produce. They also assumed transaction costs, which diminished returns a bit further. Using the SRM improved the survival rates by a considerable margin, but with a reduction in median wealth as some home equity was used up.

The gaps in survival rates and median wealth for SRM versus no-SRM widen considerably when the withdrawal rate in increased to 5.75%. The no-SRM case preserves the home value ($608,000 the median wealth with no SRM and a 5.75% withdrawal rate, is the original $250,000 increased at an assumed 3% for 30 years), but the investment portfolio runs out of money in more than 70% of the simulations.

Overall, this study shows that setting up a SRM can have a big impact on portfolio survival rates. A more subtle, but important, conclusion is that the interest-rate environment, which determines the size of the line of credit available, also has a big impact on survival rates. There might be a temptation to wait until funds are needed before setting up an SRM, but if interest rates increase in the meantime, the amount of funds available and survival probabilities from using the strategy will be reduced.

The larger context

This study shows how home equity can be integrated into a total retirement-planning context. Clients who have investment portfolios but may need to also utilize home equity to support retirement face a complicated optimization problem. Studies that have looked at investment portfolios must be expanded to integrate home equity. This research by Pfeiffer, Salter and Evensky gets us off to a good start.

Joe Tomlinson, an actuary and financial planner, is managing director ofTomlinson Financial Planning, LLC in Greenville, Maine. His practice focuses on retirement planning. He also does research and writing on financial planning and investment topics.

Read more articles by Joe Tomlinson