Experts: Mortality data, if faulty, may be the least of Detroit’s pension problems

Experts: Mortality data, if faulty, may be the least of Detroit’s pension problems

March 2, 2016

CRAIN’S DETROIT BUSINESS

Detroit and Southeast Michigan’s premier business news and information website

Originally Published: March 01, 2016 4:19 PM Modified: March 02, 2016 6:34 AM

Experts: Mortality data, if faulty, may be the least of Detroit’s pension problems

Detroit might have far bigger issues than some outmoded mortality tables used by actuaries,behind the projected $491 million shortfall in its two retiree pension funds starting in 2024,local and national pension and legal experts said.

Crain’s own Robert Snell reported here this week that an analysis by the Detroit Financial Review Commission found an $111 million balloon payment due in eight years to the pension funds under the city’s 2014 bankruptcy exit plan was calculated using mortality tables from the year 2000.

“They assumed people were not going to live as long as they would,” Detroit Mayor Mike Duggan said in his state of the city address last week.

Pre-bankruptcy valuations supplied to the General Retirement System pension board by Gabriel, Roeder, Smith & Co. in 2011 also relied on mortality tables known as the RP- 2000, published that year by the Society of Actuaries, with “a margin for mortality improvements of approximately 15 percent.” In early 2015, Gabriel Roeder completed an updated study based on “fully generational 2014 tables,” and projected that payment in 2024 at $194.4 million.

Making adjustments from old tables is not far out of step with industry practice, said Jeremy Gold, consulting actuary and owner of New York-based Jeremy Gold Pensions, so restructuring experts who relied on it were unlikely to face any sticker shock.

“Generally the RP-2000 is not particularly liberal or conservative. I don’t think these adjustments affected the generally unbiased nature of the tables..,” he said. “When (society of actuaries released a new table) two years ago they estimated this would increase pension liabilities by single-digit percentages relative to RP-2000. None of this is surprising, nor is it a major factor in Detroit underfunding.”

Other assumptions made at the time might be far more troubling than the mortality data, he said. For one thing, the Gabriel Roeder report noted that “total active payroll is assumed to increase 4 percent annually,” or that pension accounts would gradually receive the same slice of a larger pie over time, building fund assets based on a level percentage of payroll. But city payroll expenses had diminished from about $400 million in 2002 to $303 million shortly before bankruptcy, an average growth rate of -5 percent per year, so expectationsmight veer far away from reality.

 

Anthony Randazzo, director of economic research at the nonprofit Reason Foundation, also said the mortality table data could have been far better but it doesn’t seem as puzzling as other assumptions — and Duggan might need more to go on if he wants the city to pursue possible legal action against its bankruptcy advisers. If the city had not adjusted the 2000 data that might have been malpractice, he said, but as adjusted data it’s more like “an irresponsible practice, but not uncommon for the industry.”

“(U)sing old, static mortality tables (is) theoretically a justifiable practice when compared to the industry as a whole, but not a great practice considering the bankruptcy project as a whole was to put the city on more responsibly governed ground,” he said. “It is absolutely a good choice to update the mortality tables, and to continue to do so regularly in the future. I don’t understand why the emergency manager would not have recommended this in the first place, considering they were also pushing to use more responsible discount rates.” Some historical assumptions about the discount rate, or the annual rate at which projected

Some historical assumptions about the discount rate, or the annual rate at which projected future assets have to be reduced to a net present value in order to be fairly certain of covering future liabilities, were too generous, Randazzo said. Generally speaking, the higher the discount rate that pension consultants use then the less money the pensions need in fund assets right now, in order to grow to cover future liabilities or pensioner claims.

“I would argue here that they didn’t go far enough. Imagine if the city were to revalue the liabilities using a 4.5 percent discount rate? They’d have a much more accurate valuation of how much in today’s dollars the plan has promised in pension benefits,” he said. “But they’d have to honestly recognize there is still a ton of debt in those plans.

“But with a payroll assumption that is off by 9 percent compounded annually, (unfunded) costs will not be level. They will soar.”

But even if the actuaries have impeccable mathematics or prescient projections, that still might not be enough to spare them from litigation.

Sharon Almonrode, partner at The Miller Law Firm PC in Rochester and formerly a litigator at Sullivan, Ward, Asher & Patton PC in Southfield, once won one of the largest court settlements of the past decade in 2007 at $110 million, for an actuarial malpractice lawsuitshe was handling against the former Watson Wyatt Worldwide Inc.

Almonrode and her former law firm represented the Ironworkers Local 25 Pension Fund in a 2004 lawsuit alleging that Watson Wyatt, now Towers Watson & Co., had caused the plan with about 5,000 participants to be underfunded due to faulty actuarial projections. She said late last week that some litigation turns on the quality of advice to the actuary’s client, more than a matter of math.

“When you litigate those cases, and something has gone wrong, you will tend to peel back and analyze every actuarial assumption. And you could argue the appropriateness of assumptions,” she said. “But the global claim in that case (Ironworkers) was the failure to properly advise. There was not a clear concrete case of a computational or assumption error. The error would have been in what was communicated and advised (to the client).”

Use of editorial content without permission is strictly prohibited. All rights Reserved 2016

www.crainsdetroit.com

Download

 

Leave a Reply

Your email address will not be published. Required fields are marked *