Finance Department


Date of this Version


Document Type



Journal of Actuarial Practice 6 (1998), pp. 113-148


Copyright 1998 Absalom Press


We develop a methodology to ensure that a Monte Carlo simulation of the distribution of the primary rates, used for determining an interest crediting rate, is stable regardless of the initial random number seed. We consider the implications of the use of antithetic random normal deviates upon the scenario process and modifications to the candidate list and the choice function within the representative process. It is shown that the use of antithetic random deviates alone does not have a statistically significant effect on our results. The other two modifications (candidate selection algorithm and choice function) are statistically significant. Furthermore, the synergistic effects of the antithetic random deviates, candidate selection algorithm, and choice function are significant.