Connecticut should reduce the investment return assumptions of its two major state employee retirement plans, state budget Director Benjamin Barnes said Tuesday following an academic review of the state's pension funding.
"We should work with each fund's governing board to identify an appropriate and aggressive schedule for reducing their respective investment return assumptions," Barnes said in a letter to Gov. Dannel Malloy.
The State Employees Retirement System and the Teachers' Retirement System each assume an 8% rate assumption. TRS last week lowered its assumption from 8.5%.
Earlier Tuesday, the Center for Retirement Research at Boston College
"In my opinion, this scenario, while optimistic about investment returns, still presents the greatest long-term budget challenge facing the state," Barnes wrote Malloy. "However, if investment returns fall short of our 8-8.5% expectations, the future becomes alarming."
Should investment returns over the next 15 years parallel the previous 15 years – 5.25% per year on average – "we face balloon payments totaling $13 billion in order to fully fund our pensions in 2032," Barnes wrote. "This would be a catastrophic legacy to leave for our successors and our children."
According to the Boston College study by Jean-Pierre Aubry and Alicia Munnell, despite a "concerted effort" by state officials over the last decade, the funded status by SERS and TRS – the largest of six Connecticut retirement systems – declined by about 20 percentage points to 42% for SERS and 59% for TRS as of 2014, among the nation's lowest.
Pew Charitable Trusts considers 80% an acceptable threshold.
According to the BC study, the total unfunded actuarially accrued liability for the two systems combined was $25.7 billion -- $14.9 billion for SERS and $10.8 billion for TRS.
"As a result, in 2014, the state paid $1.8 billion to amortize the unfunded liability in both plans compared [with] about $400 million for benefits earned by current employees," Aubry and Munnell wrote. "And the state faces scheduled increases in amortization payments in order to fully extinguish the unfunded liability by 2032, according to the current plan."
Factors driving unfunded liabilities, they said, include legacy costs; inadequate contributions; actual performance less than what is assumed; and for SERS, poor actuarial experience.
Barnes, who called the BC study "especially helpful," also called for converting from level percent of payroll to level-dollar amortization, to eliminate back-weighting; establishing policies to convert amortization of unfunded liabilities from a closed period to a rolling one; controlling future pension costs by avoiding retirement incentives, contribution holidays or "other similar damaging practices;" and splitting SERS into two funds, one closed for Tier 1 retirees for whom most of the unfunded liability applies, and one open plan for active employees, mostly from Tiers 2 and 3.
