Risk
The Longevity Adjustment
Do new mortality tables require updates to LDI strategies? What CIOs need to consider before 2018.
Reported by Dawn Reiss
“The good news is plan participants are living longer, the bad news is plans sponsors have to put up more collateral to make good on their promises,” said Sean McShea, president of Ryan Labs Asset Management.
That’s because prior to its Retirement Plan (RP)—2014 Mortality Tables update, the last time the SOA had updated the tables was 2000.
Life expectancy rates in 2000 for the average American were 74.3 years for men and 79.7 years for women, according to the Center for Disease Control. Fourteen years later, life expectancy rates hit an all-time high, 76.4 years for men and 81.2 years for women.
“It’s gone from being a surprise attack in the middle of the night to one where people in the industry have come to terms with the fact that mortality is going down and life extension is going on,” said Bob Smith, president and CIO of Sage Advisory.
But in 2015, U.S. life expectancy rates unexpectedly dipped slightly for the first time in decades. For example, the SOA RP-2014 mortality tables report, which doesn’t do overall average life expectancy rates but shows brackets based on decade averages, demonstrated that among males age 65, overall longevity rose 2.0 years from age 84.6 in 2000 to age 86.6 in 2014. For women age 65, overall longevity rose 2.4 years from age 86.4 in 2000 to age 88.8 in 2014. By comparison, the life expectancy for a 65-year-old male declined to 85.8 years under the MP-2016 scale, compared to 86.2 years using the MP-2015 scale. The life expectancy for a 65-year-old female is now 87.8 years under MP-2016, compared to 88.2 years based on the previous MP-2015 scale.
“It’s been a stop-start kind of process,” said Bob Smith, president and CIO of Sage Advisory, about the 2014 announcement by SOA and subsequent mortality improvement scales. “It’s gone from being a surprise attack in the middle of the night to one where people in the industry have come to terms with the fact that mortality is going down and life extension is going on.”
Pension sponsors and their investment staff can expect more of the same. After a nearly four year delay, the IRS will host a public hearing on proposed regulations on April 13, 2017, with an expectation that final regulations will be applied beginning January 1, 2018, according to the International Foundation of Employee Benefit Plans. An SOA spokesperson told CIO magazine the IRS is proposing to use mortality tables based on the RP-2014 tables, projected forward using the SOA MP-2016 mortality improvement scale.
These changes will likely lower balance-sheet funded status, resulting in a higher contribution requirement for most plans. That means pricier lump-sum payouts for participants who prefer a lump sum or an annuity, said Smith, who estimates liabilities will go up anywhere from 3% to 8%, depending on the structure and nature of the participant pool. Others, like John R. Markley, founder of Markley Actuarial Services, estimated pension plan liabilities for lump sum will creep even higher, between 4% and 10%. That also means higher variable-rate premiums to the Pension Benefits Guaranty Corporation (PBGC).
With many defined benefit plans already frozen, Markley said now is the time to consider terminating that plan since CIOs can offer lump sums to participants using the old mortality instead of the new mortality.
“CIOs for retirement plans should act quickly to minimize the impact of the new mortality tables in 2018,” Markley said.
Markley suggests CIOs consider offering a lump sum window where employees who have already been terminated employment are offered a lump sum based on already accrued benefits that are payable at age 65.
“If you do that in 2017 you can use the old morality tables, but if you do it in 2018 you’ll have to use the new mortality tables,” Markley said.
This doesn’t work in all situations. “If you are going to satisfy liability by buying an annuity, that new price is already factored into insurance rates,” Markley said. “But if you offer a lump sum, you will not have to factor in the new mortality rates if you do it this year.”
Still, not every CIO will want to terminate their frozen plans, especially if they don’t have enough money to make up the unfunded liability. “It’s important to look at how many dollars the plan is underfunded,” Markley said. “If it’s underfunded by a million dollars and the employer doesn’t have access to an additional million, terminating the plan isn’t an alternative you can use.”
Looking at LDI
Given that interest rates are rising and carry risk for most defined benefit plans, many CIOs are developing new liability-driven investing strategies.
“LDI is an issue for an ongoing plan,” Markley said. “You want to use it if your plan goes beyond 2018 to manage interest rate fluctuations.”
Markley suggests considering two approaches. CIOs should either match the expected payment of the plan with bonds on a year-by-year basis or calculate the duration of the plan and then invest in a mutual fund or fixed-income asset that matches the plan’s duration.
With the prospect of unfunded liabilities edging up under the new mortality tables assumptions (see table, page 48), many pension plans are struggling with how to hedge the new mortality assumptions and risk. “It’s like you’re paying for college based on a tuition rate that is $50,000 a year and then the tuition table is updated,” McShea said. “And then you try to calculate how much you are going to owe when you didn’t plan on it but you need to try hedge it out. It’s the same for mortality changes.”
“Increase in mortality was quite significant,” said Scott Hawkins a director of insurance research at Conning, citing a 3% to 7% increase in liabilities, depending on the plan.
Since the life of plan participants is extended, it’s forcing plan sponsors to pony up additional cash.
“Many frozen plans were especially surprised at the spike in liabilities,” he said. They were told by the actuaries that if they froze their plan there wouldn’t be any additional costs. But that didn’t happen because mortality rates improved and increased liabilities.
“Increase in mortality was quite significant,” said Scott Hawkins a director of insurance research at Conning, citing a 3% to 7% increase in liabilities, depending on the plan. “Plans were more underfunded than now, and there was a lot of anxiety over the impact of increasing liabilities so significantly. At the same time, there were some complaints about incomplete data and we needed more robust sets.”
As a result, Hawkins said the IRS delayed implementation to review the data and tables and come up with something new. “There will be new tables and plans can choose a static or customized table if they are big enough and if they have enough experience that matches their own plan,” he said. Given the outcry over the long interval between the mortality table changes, a SOA spokesperson told CIO magazine that going forward, “The SOA’s Retirement Plans Experience Committee has taken steps to regularly update the mortality improvement scale as new data becomes available.”
That’s where LDI strategies come in. As mortality rates improve, funded ratios drop, and the volatility of the funded ratio increases.
“In order to understand the impact of mortality tables on LDI, you first have to take a step back and take a look at the state of defined benefit pension plans in the US in 2016 going into 2017,” Hawkins said. “Broadly speaking, those plans are underfunded. The average plan has an average funding level of 80%.”
Life Expectancy Increases, 2000–2014
In 2014, the Society of Actuaries updated mortality tables to reflectincreases in life expectancies.
Age 25
Age 35
Age 45
Age 55
Age 65
Age 75
Age 85
Male
Female
Monthly deferred to 62 annuity due value
Source: This table is from the American Society of Pension Professionals & Actuaries based on data from the Society of Actuaries RP-2014 chart.
McShea suggested CIOs review three risk levers prior to the coming year: contribution strategies, asset allocation, and benefit management. Plan sponsors need to reset asset allocations plus improve contribution levels to reduce their deficit and repair funding strategies, he said.
Since many plans are underfunded, they have to be reconciled like a checkbook at the end of the year.
“Don’t blame asset allocation or money managers,” McShea said. “Blame the liabilities themselves from life expectancy. At the end of the day, pension actuaries need to [update] their mortality assumptions more frequently.”
Rich Sega, CIO at Conning, suggested looking at long-term liabilities, which are much longer than the longest assets. “It’s not just demographics. It’s not just pensions and LDI, it’s long-term care, structured settlements, joint coverages,” he said. “We’ve had this problem for a long time.”
It’s largely a global phenomenon, in fact, where companies must be nimble with their hedging strategies. As access to long assets such as long-dated corporate and government bonds with maturities beyond 20 years continue to tighten, in part because of political concerns, said Sega, central banks are trying to help their governments by funding on the shorter end of the curve to keep interest costs on government pensions lower.
“CIOs have to cope with that,” Sega said. “You may hold on to growth strategies longer than you would otherwise had you had a tactical target to keep ahead of this.”
One of the keys, he said, is employing a completion manager who can use the futures and interest-rate swaps market to achieve that duration when there’s nothing in the cash market to offset the liabilities. “You’ve got multiple managers, so it’s hard to get them to all sing from the same sheet of music. So that overlay and completion manager would be very useful in tying it all together and trying to get ahead of these market trends.”
Rising interest rates would help lower the liability and allow CIOs to reallocate cash to earn higher rates.
“It should help to close that gap that would otherwise have been widened by mortality extensions. If you had a sophisticated LDI model that would really help to bring the whole thing together,” Sega said.
Regardless, LDI is a very customized process that can’t be generalized, Smith said. “Companies all have the same kinds of problems and challenges, but I tell CIOs, you are a snowflake in a blizzard of challenges,” he said. “They all look the same, but up close, they are all uniquely different. These series of cash flows are never identical.”
Hike in Pension Liabilities with New Mortality Rates
| Pension Plan with Current Mortality |
Pension Plan with New Mortality (10% Increase) |
|
|---|---|---|
| Plan Liability | $11 million | $12.1 million |
| Plan Assets | $10 million | $10 million |
| Plan Unfunded Liability = (1.–2.) |
$1 million | $2.1 million |
| Increase in Unfunded Liablity |
$1.1 million or over 100% |
Source: Markley Actuarial Services and American Society of Pension Professionals & Actuaries
Other considerations
A lot of CIOs are also waiting to see what the PBGC does in response to the new mortality tables, since it will need to hike plan-sponsor premiums to provide insurance coverage. “That could cost,” Smith said.
While many will be tempted to increase their allocation to equities and reduce their duration of matched fixed-income strategies to close the funding gap, Smith advises against that. Although it may be enticing to try and take advantage of potentially higher returns to compensate for the higher liability assessments stemming from the mortality table adjustments, Smith said that could be detrimental in the long run.
“It’s a huge temptation at this point,” he said. “Managements are going to have to be disciplined and guarded about how much risk they take on, since the markets have had a great run, and many think it’s going to go on forever.”
Instead, he suggests remaining disciplined, continuing to de-risk and stay on the glide path. “It’s all about incrementally building an advantage here,” he said. “There’s a temptation that we will fall off the path and have to get back on that wagon. — Dawn Reiss