The last issue of Segal Consulting’s Ideas, “Identifying Retirement Plan Risk,” described how various risk factors could affect retirement plans. Understanding those risks is the first step to being able to measure them, toward the goal of effectively managing plan risk to achieve the retirement plans’ goals. This Ideas discusses how those risks can be analyzed and quantified.
When considering defined benefit (DB) retirement plan risks, it can be helpful to think about them as falling into three categories:
This Ideas, the second in a series, focuses on how to quantify measurable risks and outcomes. Subsequent pieces in the series will identify risk-management strategies (including plan design alternatives) and outline the major components of running a well-managed plan.
This section describes the two primary approaches to quantifying risk: deterministic projections and stochastic projections. There are other methods that can be used to quantify risks for pension plans; however, deterministic and stochastic projections provide the most effective and efficient method for displaying how the various risk factors affect the two major funding metrics, the credit balance in the Funding Standard Account and the Funded Percentage, for multiemployer pension plans.
Deterministic projections are based on the hypothesis that a particular event will occur — such as that plan assets will return 7.5 percent next year, and perhaps the same in every year thereafter. Or, the returns could be set assuming lower near-term returns, followed by higher returns based on expectations that the investment environment will change. Therefore, each variable in the projection, like expected investment return or the inflation rate on administrative expenses, is predetermined for each future year.
Individual variables within the deterministic projections can be changed to measure how the change could affect financial results like plan solvency, funded ratios, withdrawal liability, minimum funding requirements under the Employee Retirement Income Security Act (ERISA) and zone status under the Pension Protection Act of 2006 (PPA’06). Deterministic projections are invaluable for exploring multiple “what if” scenarios, varying such inputs as employment levels, investment returns, contribution rates and benefit levels.
A typical conclusion based on deterministic projections might be, assuming that the investments earn the projected rate of return each year in the future, that the plan will continue to be certified in “the green zone.” This type of projection is valuable for sensitivity testing of how varying inputs may affect where the plan is headed.
However, while this method accurately illustrates the impact of a particular outcome (i.e., the plan’s exposure to that event occurring), it gives very little indication of the likelihood of that event occurring. That is where stochastic projections can be helpful.
Segal Pulse®, an enhanced version of our Forecast tool, puts deterministic projections in trustees’ hands.
Featuring an easy-to-use interface, the tool allows trustees to make real-time projections — including a review of funding issues — to increase their understanding of the direction their plans are heading and to learn what factors could change that direction.
Stochastic projections aggregate thousands of deterministic projections to provide a range of results that can be used to determine likelihood or probability outcomes within a specified range. For example, using stochastic projections, one can determine, for a specific investment portfolio and set of capital market assumptions, what the probability of remaining in the green zone will be for each of the next 10 years.
Stochastic projections can also quantify the likely effects of plan design changes under consideration. By assessing the future impact of proposed benefit changes, stochastic projections can help identify which course of action is most consistent with the trustees’ resources, strategies, and objectives for managing the plan. In particular, stochastic projections can show how plan design changes affect the investment or contribution risk to the plan. For example, two designs could appear to provide the same results when using deterministic projections. However, a stochastic projection could show that one design over another has a better probability of improving the funding status.
The accompanying chart lists five concerns shared by many boards of trustees and notes related questions that deterministic and stochastic projections can answer.
Graph 1, which is based on an actual plan’s actuarial valuation assumptions, shows a deterministic projection of the plan’s credit balance in the Funding Standard Account. The credit balance is expected to erode very slowly, and range between $266 million and $374 million over the next 10 years and, as a consequence, remain in the green zone (hence the green color of the graph bars).
Graph 2 shows a stochastic projection of the same plan’s credit balance in the Funding Standard Account. The blue diamond represents the 95th percentile, which means only 5 percent of the time better results are expected to occur. The red diamond represents the 5th percentile, which means 5 percent of the time worse results are expected to occur. The yellow bar represents the 50th percentile where half the results are expected to be better and half are expected to be worse. The dark grey boxes represent the 25th to 75th percentile where where half the results will occur. Graph 2 shows the plan has a 25 percent probability of a funding deficiency occurring by the end of the 2027 plan year, which is an outcome not shown with the deterministic projection shown in Graph 1.
A deterministic projection of the plan’s Pension Protection Act of 2006 (PPA’06) funded percentage, which like Graph 1 is based on the plan’s actuarial valuation assumptions, is shown in Graph 3. The graph indicates the plan’s funded percentage is stable, decreasing in the short term, but increasing over the long term and it is expected to stay above 85 percent. Consequently, the plan is expected to be certified in the green zone (again, represented by green bars).
Graph 4 shows a stochastic projection of the sample plan’s PPA’06 funded percentage. The plan has a little more than a 50 percent probability of being more than 90 percent funded by 2027 and a 25 percent chance of having a funded percentage lower than 80 percent in 2021. Again, the stochastic projections paint a different picture of the same plan from that of the deterministic projections.
Graph 5 shows the probability of the same plan being in the green, yellow or red zones. By 2022, there is a 31 percent chance the plan will not be green. That chance increases to 38 percent by 2027. Compare this to the deterministic projections that show the plan always in the green zone.
All of the graphs above focus on the effect of market returns on various funding metrics of a pension plan. Stochastic projections are better suited to show how investment risk measures like the volatility (standard deviation of expected returns) of each asset class, as provided by the investment consultants, affect the various funding metrics of the plan. Deterministic projections are better suited to show how different investment return assumptions will affect the plan’s funding. Stochastic projections excel at displaying a range of results and measuring the probability of the plan’s various funding metrics changing over time. Both types of projections can be used to make decisions, and it depends on the question asked as to which type will provide the best answer.
It is important to keep in mind that projections, whether deterministic or stochastic, are only as valid as their underlying assumptions and the methods used to build them. Consequently, the assumptions for projections must be chosen carefully, with internal consistency and with input from professional advisors and the board of trustees.
All projections are based on assumptions about unknown future events and, as a result, what actually occurs will almost always differ from what was projected. However, in conjunction with other strategic planning information, such as projected industry employment patterns, quantifying risks through deterministic and stochastic projections can help retirement plan trustees answer questions based on a range of results and help them make more informed decisions.
The ability to quantify risk enables plan sponsors to plan for possible future events that may not be readily evident from a standard deterministic projection or an actuarial valuation. Once armed with the ability to quantify certain risks (i.e., investment risk, longevity risk, contribution risk), trustees can manage their plans to improve the plans’ long-term financial health and stability.
The next issue of Ideas will explore strategies for managing retirement plan risks. Refer to the Managing Pension Risk page on Segal’s website for links to past and future Ideas in this series as they are published, as well as links to videos on related topics.
To discuss how Segal Consulting can help you identify the risks associated with your fund’s defined benefit and defined contribution retirement plans, contact your Segal benefits consultant, or one of the experts listed below.
Consultants and actuaries from Segal Consulting, together with investment consultants from Segal Marco Advisors (the SEC-registered investment solutions member of The Segal Group), can be of assistance in developing the appropriate strategies for maintaining and enhancing benefit security. Our strategic consulting services for multiemployer pension plans include the following:
For information about these strategic consulting services, contact your Segal consultant, the nearest Segal office or one of our experts.
To receive Ideas and other Segal publications of interest to sponsors of multiemployer plans, join our email list.