A new study finds that dynamic strategies increase the utility adjusted wealth during retirement.
The use of “static” models is common in retirement-income research. Static models imply that a retiree makes decisions only at retirement and that the original strategy is followed the entire retirement period regardless of its ongoing efficiency. For the purposes of this article, we use “dynamic” models, which allow for intelligent change during retirement. Dynamic models are a more accurate perspective of retirement choices and, therefore, represent a better approach when determining optimal retirement-income strategies.
Through our analysis, we find that dynamic strategies can significantly increase the utility-adjusted wealth for retirees. We note an average overall increase in utility-adjusted wealth of 7.28% for our optimal, most-dynamic model compared with our simplest model, although the average improvement varies materially by scenario from less than 1% to more than 20%. We estimate that it would take an annual fee of 1.26% (i.e., negative alpha) to eliminate the 7.28% increase in utility-adjusted wealth that can be obtained from the most efficient dynamic strategy.
A Dynamic Perspective
Seeking to achieve a constant level of real income may be reasonable from the perspective of a goal realization. It is, however, unlikely that when faced with realized asset returns and changing retirement circumstances, a person will find the static path optimal. In reality, it is likely a retiree would change his or her consumption based on the likelihood of achieving a goal. Portfolio performance above the historical average may result in a greater desired consumption path because a retiree would have greater wealth than expected. If the portfolio performs poorly, a retiree would likely reduce spending to minimize the likelihood of financial ruin.
A withdrawal strategy where the portfolio amount is revisited at some frequency throughout retirement to take into account experiences and expectations is known as a dynamic approach.
Similar to the idea that dynamically updating the withdrawal rate can lead to improved outcomes for retirees, there are also potential benefits to purchasing annuities throughout retirement. We define dynamic annuitization as the process where the amount and the timing of annuity-purchase decision can vary as a retiree moves through retirement.
To test the value of dynamic approaches to retirement income, we used a variety of models. We made the following key assumptions of each model (the primary working paper should be consulted for additional information for each model):
The income-preference model considered two preferences. The first had to do with the risk aversion associated with having an income shortfall during retirement (relative to the income goal). The second was the relative importance of a potential bequest. The model seeked to determine optimal tradeoff between certainty of income (which would favor annuities) versus maximizing wealth at death (which would favor nonannuity strategies).
The optimal strategy was the one with the highest “utility-adjusted funded ratio.” This was a metric that was effectively the utility-adjusted wealth created by a given strategy within the context of the cost of the desired retirement goal.