Calculating expected income replacement rates can be an imprecise and misleading way for defined contribution plan executives to assess participants' retirement readiness, according to new research from the Ibbotson Associates unit of Morningstar Inc.
That's because there are so many variables and economic assumptions used in calculating how much of a participant's final-year compensation will be replaced per year during retirement.
“Plan sponsors are increasingly requesting — and making decisions based on — replacement rate studies where each provider conducts an analysis of its own product” to forecast the best replacement rate, said David Blanchett, a research consultant at Ibbotson in Chicago and the author of a report on the research. Because different providers use so many different assumptions, plan officials can be presented with wildly varying recommendations, he added.
“It is virtually impossible to compare the results of various studies and form any type of reasonable conclusion,” Mr. Blanchett said in an interview. “Each provider uses a different methodology.”
Instead, Ibbotson recommends hiring a single consultant to create a “controlled experiment to meaningfully compare replacement ratios,” said Mr. Blanchett. But he thinks an even better choice would be for sponsors to hire a consultant to analyze the different methodologies of the providers to see which approach best matches the needs and demographics of the sponsor's workforce.
“Replacement rate estimates can vary materially based on different assumptions and provide varying levels of insight,” according to his report, “Inaccurate Precision: The Danger of Replacement Rate Calculations.” “Plan fiduciaries should never compare replacement ratios calculated by different vendors.”
To illustrate how minor changes in assumptions can have a big impact on replacement rate calculations, Mr. Blanchett created a replacement rate calculator using 10 factors ranging from specific assumptions — such as a participant's retirement age, outside assets, change in savings rate, and asset allocation — to broader ones, such as market forecasts and inflation.
Then, he varied these assumptions for 20 sample plans to determine their impact on replacement rates. The range of median replacement rates among the 20 plans was between 48% to 107%, based on the different assumptions, the report said. The replacement rates ranged from 67% to 261% for individuals based on the different assumptions, the report said.
Mr. Blanchett's report said several factors typically will have a greater impact on results than others. They are a higher percentage of younger workers, participants with higher levels of pay, workers with higher savings rates and workers with lower plan balances.
Although sponsors approach replacement-rate strategies with good intentions for retirement readiness, “these studies are fundamentally flawed,” Mr. Blanchett said in the interview. “Plan sponsors need to better understand the methodology.”
Jack VanDerhei, research director for the Employee Benefit Research Institute, Washington, said in an interview that the Ibbotson research is “something different than anyone has done.”
Mr. VanDerhei, who reviewed the Ibbotson Associates' research at the request of Pensions & Investments, said the Ibbotson report achieved its goal of illustrating the “dramatic” variability of factors in trying to figure a replacement rate. However, Mr. he also noted the analysis didn't include factors such as health care expenses or long-term care costs.
Like Ibbotson Associates' questioning the adequacy of sponsors calculating replacement rates, EBRI has questioned the adequacy of participants relying on replacement rates, Mr. VanDerhei said.
Replacement rate calculations are “overly simplistic and potentially inaccurate,” according to a 2006 EBRI report. These calculations often rely on replacing pre-retirement cash flow after adjusting for taxes, savings, age and/or work-related expenses, but they ignore investment risk, longevity risk and/or the risk of “potentially catastrophic health care costs,” the report said.
“Replacement rates can give you an answer,” Mr. VanDerhei said in the interview. “Whether it's the right answer — or not — is debatable.”
