With a looming $13 billion debt payment coming due for the 55.97% funded Teachers’ Retirement System (TRS) pension plan, the Connecticut governor and treasurer are at odds. After Connecticut’s credit rating was downgraded four times last year, Treasurer Denise Nappier is emphasizing the covenant on the bond should not be broken. But Gov. Dannel Malloy and actuaries say the payments may be just too high for the cash-strapped state.
When the state decided to catch up on payments with a bond, the covenant to the bond required the Actuarially Determined Employer Contribution (ADEC) be paid for the life of the bond. It was a 25-year bond from 2007 and the pay-off date is 2032.
There are two options: Malloy wants to make TRS more like a state plan and reduce the assumed rate of return from 8% to 6.9% to align with capital market future assumptions, and then re-amortize the liability over a longer period of time. But just the change in the amortizing is a violation in the bond covenant. Facing an overall state budget shortfall of $244.6 million, Malloy is also proposing requiring towns to contribute to the TRS.
Meanwhile, Nappier is against violating the bond covenant and against re-amortization, fearing reputational damage will effect rates in the future when the state wants to borrow.
“The biggest issue that we always have is reputational risk—that sometimes causes the markets to go up and down,” Nappier told the viability board, “When you’re dealing with sophisticated investors and they look at a breach in the bond covenant, then every indenture agreement from here on is going to be suspect. And that causes the rates to go up, significantly.”
She is scheduling to go to the bond market at least three or four times in 2018. “The next time I go to the bond market, I’ll get hit,” she said. “You’d be surprised at how emotional the market is.”
In May, Fitch Ratings downgraded the state from AA- to A+, citing “reduced expectations for economic and revenue performance over the medium term” following a decline in personal income tax. In the same month, Moody’s reported Connecticut has the highest debt service costs of the 50 US states, although it declined to 13.3% from 14.3% in the previous year. Its net-tax supported debt (NTSD) is $6,505 compared to the median per capita of $1,006.
Nappier suggests waiting until 2025 to make any changes, pay out the bonds then, and move forward from there.
“There is money,” she told CIO. “It’s what bills are you going to pay, and I am saying you’ve got to pay this one. They need to pay this bill. It’s the bill that goes back to the 1940s.”
But the state is saying it can’t possibly pay the whole tab.
How it Happened
The plan’s first problems arose in 1939, when plan promises were made but not funded for both the state employees’ plan and TRS until the 1970s. “TRS has always used the less effective level-percent-of-payroll approach to calculate amortization payments. Additionally, a lax statutory funding schedule allowed TRS to underpay until 1992,” according to a 2015 report by the Center for Retirement Research at Boston College, commissioned by Connecticut.
The underfunding was partially corrected through a bond issuance in 2008, which had covenants mandating the plan’s unfunded actuarial accrued liability (UAAL) had to be fully funded by 2032, but shortly after the ink was dry on the agreement, the plan was battered, losing 17.3% in the financial crisis.
Last year, despite the teachers’ contribution rate increasing 1%, the pension’s funding levels still dropped from 59% to 55.97% with $13.1 billion in unfunded liabilities. The UAAL grew by $1.776 billion due to interest and decreased by $1.654 billion due to the amortization payments over the two-year period, according to the November 2017 Connecticut State Teachers’ Retirement System Actuarial Valuation.
What’s Next
About two months ago, The Teachers’ Retirement System Viability Commission was created by the state to explore the issue. The actuaries were hired, and a public hearing was scheduled but postponed due to an ice storm.
Yet on Feb. 7, (during the ice storm,) the viability board meeting was still held to allow the actuaries (who had traveled from Atlanta, Georgia) to fine tune their analysis before presenting their recommendations in late February.
Benchmarking Challenges
In benchmarking, the actuaries had their challenges, with a limited number of plans in which they could use for comparison. Since plan provisions varied, according to the actuaries, “they were not clean comparisons.” Since they could not change the benefits of the plan participants, the actuaries compared the plan against the latest tiers of benefits for other retirement systems, which, in some cases, provided fewer benefits or later retirement eligibility. The benchmarking was also limited to plans that were not covered by Social Security, which only left about a dozen plans to compare against.
Moving the Needle
In crunching their numbers and forming their projections, the actuaries found that if they changed benefits for future members, it didn’t “move the needle” enough to make a difference to the debt. It takes time, “decades,” to show a noticeable effect, they said. And the new hires are not going to contribute to the $13 billion liability. “There’s no unfunded liability on someone you hire tomorrow,” said actuary John Garrett of Cavanaugh MacDonald Consulting. Market-driven asset losses and low contributions contribute to the liability the most. Shifting from a five-year vesting plan to a 10-year one also had little effect. Teachers have one of the longest careers, at 35 years, with few enhanced benefits, and so, pulling back benefits also wasn’t applicable.
How Realistic Is It?
When modeling the plan, the actuaries said they found unrealistic expectations for the plan. In valuing the liabilities and cash flow, they used the current funding policy that the statutes mandate, including the amortization periods and the 8% rate of return assumption in the plan. When modeling the impact of investment volatility on the funding metrics, the actuaries used the treasurer’s office return expectations of a 7% geometric mean and a standard deviation of 11%, which is about the average for plans around the country.
“What that means is two-thirds of the time, we’re expecting standard returns between 18% on the high side, down to -4% on the low side,” said Garrett, which has historically been the plan’s norm. “If you follow the track up, with that assumption of 7% returns, we end up at a point where the required contributions are going to increase beyond a realistic amount of money that we can consider the state would put into the plan. There has to be a reality,” said Garrett.
The unrealistic expectations would include a contribution of up to 8% of the total revenue of the state. In a bad market environment, it would require the state put $4 billion into the plan.
“We’re going to hit a point in time between 2020 and 2032 that the challenge to the state of coming up with the full [amount] is likely to be more than they can afford,” said Garrett, and the plan is unlikely to be 100% funded.
The plan’s current amortization policy, where all is paid off by 2032, has a 70% likelihood of being breached by the constrained contribution, said actuary Larry Langer. He noted that now that plans have more equity exposure, actuaries are currently promoting the idea of layered amortization over longer periods of time, such as 15 years, to reduce volatility and so that contributions don’t need to spike after years of low performance. But the bond covenant likely won’t allow for changes in the contribution policies, he said.
The actuaries noted that despite the challenges in paying off the $13 billion past debt, the plan is still predicted to be solvent throughout a 50-year projection period, even through worst-case scenarios with constrained contributions.
Yet Nappier says violating the covenant is a big risk, not only to the state’s credit rating, but to the plan. If the covenants had not been in place, she emphasized, the plan would never have been funded. “They never had before,” she said. Also, the bond program was successful: the state has paid less in debt than it has earned in bond proceeds, which are up at least $142 million. “But it’s not about the debt, it’s about the discipline,” she said. Yet in keeping with either a 7% an 8% assumed rate of return, with the state making required contributions, the plan would still only be 48% likely to be 70% funded by 2025, according to actuarial projections. “That’s a little bit less than half likely that we’re going to get there,” said Garrett, who will return with his recommendations on Feb. 26.