Journal for Financial Analysts
Retirement, like life, is fundamentally uncertain. That’s why we want to offer our clients more context about what it’d appear like in the event that they don’t meet their retirement income goals, and accomplish that in a thoughtful way.
In my previous two articles, I examined how retirees are likely to have more flexibility of their retirement spending than traditional models imply and discussed a basic framework for dynamically adjusting their spending. Here, I examine how commonly used financial planning metrics are flawed—particularly probability of success—and why we must always consider other consequence measures that may provide additional and higher insight into clients’ retirement income situations.
The rise of Monte Carlo
Financial advisors often use Monte Carlo forecasts for instance the uncertainty related to funding retirement income and other retirement goals. The element of likelihood or randomness is the important thing differentiator of Monte Carlo projections in comparison with time value calculations and other methods.
While it will be important to reveal the likelihood that a goal is probably not achieved, it’s also necessary to stipulate the range of possible scenarios. Probability of Success is probably the most common consequence metric in Monte Carlo tools and refers back to the variety of runs or attempts through which the goal is fully achieved in a given simulation. For example, if a retiree goals for an annual income of $50,000 for 30 years and achieves this goal 487 times in 1,000 runs, the prospect of success is estimated to be 48.7%.
However, success-related metrics treat the consequence as binary and don’t describe the extent of the failure or how far the person was from achieving the goal. By such standards, it doesn’t matter whether the retiree fails within the tenth or thirtieth 12 months, or by $1 or $1 million. All errors are treated equally. So a retiree could have a comparatively small deficit but additionally a low probability of success, especially if their retirement income goal is funded primarily by guaranteed income and over a comparatively long assumed time period, reminiscent of 30 years.
Goal achievement
However, a financial goal isn’t a discrete series of “pass” or “fail” outcomes. It’s a spectrum of possibilities. This is why it’s so necessary so as to add context to the extent of possible failure. The percentage of goal achieved is a critical metric. The graphic below illustrates this effect with an assumed goal of $100 per 12 months for 10 years.
Percent likelihood of reaching the goal of $100 per 12 months for 10 years
For example, in runs 1 to five the goal is barely partially achieved. The percentage varies across the five simulations, but each run represents a “failure” based on success-related metrics. Other metrics tell a distinct story. Average goal achievement involves achieving a median of 90% of the goal, while success rate indicates a 50% likelihood of success. Although these two metrics are based on an identical data, they supply very different perspectives on the safety of goal spending.
The relatively low success rate suggests that achieving the goal is much from certain. However, the goal completion value offers a far more positive picture. This is especially necessary for longer-term goals reminiscent of retirement, where “failure” is most definitely in the ultimate years of the simulation.
Diminishing marginal utility
While goal achievement percentages provide a more nuanced perspective on the outcomes of Monte Carlo simulations, in addition they don’t bear in mind how variable the profit or pain that comes with missing a goal may be. For example, not funding essential expenses like housing or healthcare is prone to cause greater dissatisfaction than cutting travel or other flexible spending.
The concept of diminishing marginal utility describes this relationship: the pleasure of consuming or financing something typically increases, but at a decreasing rate. This could explain why people buy insurance despite the fact that it reduces wealth on average. They guarantee that they’ll finance a certain minimum consumption.
Goal attainment percentages may be further modified to account for diminishing marginal utility, whereby the implicit satisfaction related to achieving a given level of consumption changes, particularly depending on whether consumption is discretionary or non-discretionary. I developed a framework to make these adjustments based on prospect theory. These values may be aggregated across years inside a given run and across all runs. This leads to a goal achievement rating which will require very different advice and guidance than modeling based on probability of success rates.
Working with what we now have
Our industry needs to include higher earnings metrics into financial plans. Such metrics must take goal achievement under consideration and incorporate utility theory more directly. Certainly relatively few tools can do that today, so financial advisors might have to offer higher advice using the present toolset.
If you proceed to depend on success rates, it’s best to cut back your goals a bit. Based on my research, 80% might be the suitable goal. That could appear low: Who desires to have a 20 percent likelihood of failure? However, the lower value reflects the proven fact that “failure” in these situations isn’t as catastrophic because the metric suggests.
Customers also need more context about what exactly a foul consequence entails. As financial advisors, we are able to explain what the income from unsuccessful legal proceedings is. How bad are the worst case scenarios? Does the client have to earn $90,000 by age 95? This is way more powerful than a hit rate and shows how bad things could go in the event that they don’t go well.
Conclusions
Probability of success often is the primary consequence metric for advisors using Monte Carlo projections, however it completely ignores the magnitude of failure. Success rates may be particularly problematic for retirees who’ve higher life expectancy-backed or guaranteed income and more spending flexibility. Alternative outcomes metrics may also help us bridge this gap and ensure we offer our customers with appropriate and accurate information to assist them make one of the best possible financial decisions.
If you enjoyed this post, do not forget to subscribe.
Photo credit: ©Getty Images / Gilaxia
