A two-parameter family of pension contribution functions and stochastic optimization (Q1111307)

In ibid. 5, 141-146 (1986; Zbl 0587.62191), the author has suggested a linear function of A(t) (present value of future benefits) and F(t) (fund) as pension contribution function in place of the form given in \textit{C. L. Trowbridge} [Trans. Soc. Actuaries 15, 151-169 (1963)] which is a one-parameter family of funding methods. Here we provide some theoretical justification for such a method by showing that, in the simplified model of this paper, the optimal solution of a stochastic control problem yields, as contribution function, an affine function of A(t) and F(t).