“NISA has always enthusiastically advocated for hibernation strategies for plan sponsors looking for an optimal cost vs. risk tradeoff as they wind down their pension liability,” said David Eichhorn, the firm’s CEO and Head of Investment Strategies. “Clearly, the cost associated with hibernating and maintaining a plan on the balance sheet is a key consideration when thinking about an annuity buyout because it effectively sets the hurdle rate,” he said, at Pensions & Investments’ Managing Pension Risk & Liabilities Virtual Series. Plan sponsors need to look at realistic costs for the strategy, review historical data to see if the hibernation strategy is sustainable and, finally, take a more informed view on longevity risk, he said, at the workshop titled ‘Decisions, decisions, decisions.’
HIBERNATION IS A COST-EFFECTIVE SOLUTION
Eichhorn referenced a recent piece by an insurer called ‘Hibernation’s Wake-up Call’, which he used as an illustration to unpack the costs associated with annuity buyouts versus hibernation strategies. The costs for hibernation in that piece are exaggerated or irrelevant, he said. The paper estimated that the cost of maintaining the liability on the balance sheet is 14%(1) higher than the stated GAAP liabilities and that an annuity purchase could save a plan sponsor 10%(1) relative to this value of the liability, “but that number is high versus what we’ve experienced running hibernation strategies for years,” Eichhorn said.
The insurer provided three line items as the key drivers for the estimated 10% savings via an annuity purchase, Eichhorn noted. The first line item, 4%(1) for administration and PBGC premiums, has variable rate premiums and fixed rate premiums. The former type make-up 85% of the costs assigned to PBGC premiums and are not relevant for this cost comparison, he said. “An annuity buyout is not what eliminates the PBGC variable rate premium. These are the premium paid on a deficit, and they can be eliminated by funding the plan. So, while variable rate premiums are an important consideration for a sponsor who has a choice to fund or not fund the plan, the annuity purchase is a separate decision.” Removing the irrelevant/inappropriate PBGC variable rate premiums line makes that total line item 0.5%.
“PBGC fixed-rate premiums are very relevant and are an additional cost as they're not factored in directly to the GAAP liability. However, for a typical monthly benefit of around $2,000, they're fairly a de minimis cost at four or five basis points per annum. In present value, for a 10-year duration liability, we estimate administrative and PBGC premium costs at 50(2) basis points to keep the plan on the books,” Eichhorn said. However, plan sponsors need to look carefully at the participant cohort that receives very small balance benefits as those can aggregate to get quite expensive in present value terms, and they could consider a small-balance transaction to eliminate that portion of the liability, he said.
The second line item, the cost of credit defaults and downgrades, is a genuine cost as it is not factored into the GAAP liability, but it is clearly overstated at 7% in the insurer’s hibernation estimate, Eichhorn said. “For the types of bonds we would suggest owning against the corporate liability—high-quality single-A and some BBB—we estimate the annualized cost of downgrades and default of about 50(2) basis points, or about 5% or so in present-value terms.” Also, downgrades and defaults are two different events, he said, noting that most downgraded bonds are still expected to mature. “If a plan sponsor is willing to be downgrade-tolerant and not sell at the arbitrary cutoff of BBB for every bond that's downgraded with the expectation that the vast majority will mature, that data changes quite a bit,” Eichhorn said. So, instead of 50 basis points, the annualized default loss drops to 5 to 25 basis points. This line item overall would then amount to about 150 basis points of cost that the GAAP liability doesn’t represent.
The third and last line item from the insurer’s estimate included 3%(1) for investment management fees which, Eichhorn said, is a very active fee level, higher than full active management of a hibernation strategy. “Probably the single most common error in thinking about costs to run a plan on the pension books, as opposed to a buyout, is that there is often no credit given for the alpha generated by active management,” he said. “So there's really two lanes you can take here: one, you can assume a passive fee, which we think amounts to about 5 basis points and has no alpha, and I’m not saying that’s the right route, it’s just for the comparison. Alternatively, look at the active fee versus alpha for a hibernation strategy that’s predominantly fixed income. The e-Vestment database from 2007 to mid-2020 shows long credit managers averaged 78(3) basis points over the benchmark. Adjust for conservatism, we get a net active management fee of negative 59 basis points.” That is, the expectation of future alpha reduces the value of the GAAP liability by roughly 3.4%(4).
“So, our response is regarding hibernation strategies, as opposed to a wake-up call, is ‘hit the snooze button!’
I'm an enormous fan of the snooze button, frankly,” said Eichhorn. With 50 basis points for administrative and fixed PBGC premiums, 5% for credit defaults and downgrades, and a net negative cost of -3.4%(4) on the actively managed portion of the hibernation portfolio, he estimated 2.1% cost of maintaining the plan on the balance sheet, which is far different from the +10% in the insurer estimate.
Eichhorn showed the historical performance of two portfolios – one a simple passive one and the other an actively managed one – over 5-25 years. “Over time, including costs and credit events, the hibernation strategy, even with passive management, keeps up and, in some cases, picks up some ground,” he said. “I’d argue this is validation that there aren’t significant hidden costs or, put differently, if you're 100% funded today and you get a hibernation strategy, you have every reason to expect it will remain 100% funded over time, of course with some modest variation.” Moreover, with the typical active manager alpha, the baseline hibernation strategy even does better, in some cases picking up 12% in funded status, he noted, adding that “To be very clear, I wouldn't want to manage expectations to say at a 95% starting point, you'll be able to pay every liability without a future contribution, but that at least is a reasonable expectation. Our poor man's back test, for lack of a better term, validates our cost assumptions and our analysis that says hibernation is sustainable.”
For plan sponsors considering an annuity buyout, a plan-specific mortality study is useful to get the right assumptions for the appropriate cohort, Eichhorn said. Looking at detailed data on longevity estimates by zip code shows significant variation due to socioeconomic factors, including industry, type of blue/white collar job, and lump sum availability. “There are zip codes within the US that have 21 years of different life expectancy,” he noted, adding that a wide range of plan participants can differ materially from the industry standard tables or the tables used for GAAP liability. “The longevity estimation process is the X-factor in our mind of reconciling the difference between what we describe as the cost of a buyout versus sometimes what’s reported in the market,” he said.
“To summarize, it’s all about getting the assumptions right when thinking about the cost of keeping a plan on the balance sheet versus the annuity buyout. We think if we do that appropriately the hibernation path is the cost-effective solution to manage the pension liability.”
(1) Source: http://pensionrisk.prudential.com/pdfs/hibernation.pdf.
(2) Includes PBGC, costs provided by PBGC, and administrative costs held constant at $40 per head.
(3) Reported fees are sourced from eVestment. For the purposes of this analysis, managers who did not report a fee was assigned a fee equal to that of the average reported fee amount. Calculations are based on a $1 billion portfolio. *Note: There are shortcomings to using databases, including limitations on inclusiveness and survivorship bias. The look-back period was chosen in order to include the effects of the credit crisis that began in 2008; additional periods are available on request. Benchmarks identified in the eVestment Analytics database are selected by the product’s investment manager and may not be indicative of the true product benchmark. The analysis is based on data from 12/31/2007 through 06/30/2020 from eVestment Analytics, Bloomberg, and Bloomberg Index Services Ltd. Neither NISA nor eVestment guarantee or warrant the accuracy, timeliness or completeness of the information provided by eVestment and are not responsible for any errors or omissions with respect to such information. Past performance is not a guarantee of future performance.
(4) Calculations are based on a $1 billion illustrative portfolio which has 60% allocated to an active Credit manager.
This sponsored advertorial is published by the P&I Conferences Group, a division of Pensions & Investments. The content is not produced by the editors of Pensions & Investments and www.pionline.com and does not represent the views of the publication or its parent company, Crain Communications Inc.