Optimal Surrender of Guaranteed Minimum Maturity Benefits Under Stochastic Volatility and Interest Rates
25 Pages Posted: 5 Jun 2016
Date Written: June 2, 2016
In this paper we analyse how the policyholder surrender behaviour is influenced by changes in various sources of risk impacting a variable annuity (VA) contract embedded with a guaranteed minimum maturity benefit rider that can be surrendered anytime prior to maturity. We model the underlying mutual fund dynamics by combining a Heston (1993) stochastic volatility model together with a Hull and White (1990) stochastic interest rate process. The model is able to capture the smile/skew often observed on equity option markets (Grzelak and Oosterlee, 2011) as well the influence of the interest rates on the early surrender decisions as noted from our analysis. The annuity provider charges management fees which are proportional to the level of the mutual fund as a way of funding the VA contract. The fair management fees to be charged by the annuity provider are computed with aid of the Fourier Cosine expansion (COS) method which is a proven computationally efficient algorithm. To determine the optimal surrender decisions, we present the problem as a 4-dimensional free-boundary partial differential equation (PDE) which is then solved efficiently by the method of lines (MOL) approach. The MOL algorithm facilitates simultaneous computation of the prices, optimal surrender boundaries and hedge ratios of the variable annuity contract as part of the solution at no additional computational cost. A comprehensive analysis on the impact of various risk factors in influencing the policyholder’s surrender behaviour is carried out, highlighting the significance of both stochastic volatility and interest rate parameters in influencing the policyholder’s surrender behaviour.
Keywords: Variable annuities, optimal surrender, GMMB, stochastic volatility, stochastic interest rates, Fourier Cosine expansion, method of lines
JEL Classification: C63, G13, G22, G23
Suggested Citation: Suggested Citation