Cheiron has an industry-wide reputation for creating sophisticated interactive modeling tools such as P-Scan and H-Scan. We have recently created a new interactive tool called R-Scan that enables exploration of the risks embedded in pension funds during live meetings and thus facilitates better understanding by Plan Sponsors and Trustees.
Cheiron’s R-Scan platform uses our base P-Scan model to demonstrate a likely range of potential future outcomes and how that picture changes if the funding, investment, or benefit policies are modified. The graph shown below depicts a basic output screen from R-Scan. The colored bars represent the percentiles of possible results of this plan’s future contribution rates for the next fifteen years. This tool can be programmed to show results for longer periods as well as any number of statistics (funded percentages, amortization periods, contributions rates per hour, compounded investment returns, etc.) that are of interest as well as quickly switch between them.
The black line represents the expected, or “middle of the pack” in potential outcomes. However, future investment returns are uncertain and the expected outcome cannot be relied upon when considering how much cost uncertainty there is embedded in a plan. A plan sponsor might want to see how large contribution can get by 2021. From the graph below, one can see the 95th percentile for 2021 is approximately 110% of payroll. This means that given all the possible future outcomes, one can expect 95% of the time for the contribution rate in 2021 to be less than or equal to 110% of payroll and only a 5% chance of being greater than 110% of payroll. Similarly, the 80th percentile is approximately 90% of payroll. This means that given all the possible future outcomes, one can expect 80% of the time for the contribution rate in 2021 to be less than or equal to 90% of payroll and only a 20% chance of being greater than 90% of payroll.
The next graph illustrates how R-Scan can be used to compare potential outcomes under two different sets of policies. In this case, the two policies represent two different investment portfolios with correspondingly different discount rate assumptions. The more aggressive 7.5% assumption of the baseline model, with its accompanying higher standard deviation, displays a wider range of potential outcomes.
Finally, R-Scan can show the probability distribution for a single year’s results. This information can be isolated for any given test statistic and any year in the projection.
The costs of a defined benefit plan are fundamentally uncertain until investment returns, longevity, retirement behavior and other factors are known. Understanding the possible ranges of those costs is the first step in developing the appropriate policies to manage cost uncertainty and ensure the sustainability of the pension plan.