On the feasibility of reverse mortgages in Colombia

Purpose The Colombian Government proposed a reverse mortgage mechanism to complement retirement income in Colombia. This paper aims to study its feasibility by valuing its premia. Design/methodology/approach Under a reverse mortgage scheme, banks issue put options on an owner's home. To value the option, the authors apply a risk-neutral canonical approach modeling its three sources of risk: home future value, interest rate levels and homeowner life expectancy. Findings In all, premia values do not seem too high. However, if future interest rates are above the simulations or home appreciation is below its historical behavior, the premia could sharply increase, jeopardizing the system viability. Limiting the loan-to-home-value ratio or fixed-term annuities are feasible alternatives to keep premium increases at bay. Complementary mechanisms may also help. Research limitations/implications The home price and interest rate path estimation do not include inflation dynamics; in recent years inflation level was very low. However, the future does not offer any warrants. Future research also should cap the maximum loss the bank can endure. The pandemic may cause demographic changes affecting the viability of the reverse mortgage (R.M.) program in Colombia. Practical implications Based on the analysis, this work suggests possible government policies to help creditors and to maintain bank risks at a reasonable level. Social implications An adequate reverse mortgage program may help the policymakers in Colombia to face the adverse environment for Colombia's housing market and the pressure of its pension system. A good R.M. program generates incentives to purchase homes, given the possibility of receiving an additional rent after retirement. Originality/value The paper develops an econometrical improvement over previous work. The authors present a time-series analysis that includes stationarity and co-integration information to model the data-generation process of house prices and interest rates in a multivariate fashion. The authors also improve the valuation formula. Moreover, the paper presents a novel application to Colombia. The authors obtain our demographic data from the United Nations Population Division applying the Lee-Carter method to model mortality rates, which provides ample possibilities to extend reverse mortgage assessment to additional. Finally, this is the first scholarly effort to evaluate the R.M. for the Colombian case.