Academy of Financial Services, Nashville
October 2014
Abstract
There is a strong desire in the financial services sector to be able to address the uncertainty of a projection or forecast for a client’s financial plan based on two primary factors – mortality risk and rate of return risk. The primary methodology for handling uncertainty has been and remains Monte Caro Simulations for investment returns and randomization of mortality assumptions based on mortality tables. This paper hypothesizes that it is possible to construct an algorithmic replacement to Monte Carlo Simulations. The advantage of MCS is that by generating a sufficient number of series of randomized results based on the expected arithmetic return and standard deviation we can then apply these to whatever forecast we desire and aggregate the results. The disadvantage remains the widespread use of limited numbers of iterations introducing estimation error and the lack of granularity in the aggregated results based on a “pass/fail” measure.
