A Discretionary Wealth Approach for Investment Policy

Abstract

In this article, the authors suggest a paradigm for broad investment policy. After briefly reviewing a point-estimate model based on growing discretionary wealth, they improve that approach using Bayesian logic. Risk in the outcomes of saving and spending plans is treated in parallel to risk in investment returns and can be fully comprehended by a Bayesian approach. This approach is particularly useful, for example, in retirement planning where longevity risk is an issue. For many other cases in which investment funding and spending outcomes are nearly certain, the investment implications of a fully Bayesian model are more subtle and its application may not produce big effects. Using illustrative asset allocation problems, tests of the model appear to show material potential improvement when long–short positions, high leverage, long time periods between rebalancing, or option positions are present.