Actuarial firms, bankers and insurancecompanies have been urging U.S. defined benefit plan sponsors to transfertheir liabilities to a third party. Transferring the total liability hasproven too onerous for most plan sponsors, so these institutions have beenpitching partial risk transfers (PRT) instead. However, the touted long-term cost savings and pension risk reduction may turn out to beunderwhelming, if not elusive.
Hibernation strategies, a lower-cost alternative to pension risk transfers,may lead to better outcomes for plan sponsors in the medium term – andeven put the plan sponsor in a better position for a more successful pensionrisk transfer in the future.
A common argument in support of PRTs revolves around the idea of reducingPension Benefit Guarantee Corporation (PBGC) premiums. The PBGC leviespremiums on single- employer plans on two fronts – a “perhead” premium (the fixed rate premium) and, for underfunded plans, avariable rate premium (VRP) based on the percentage of a plan’s unfundedliability.
There is some merit to the idea that PRTs can help plan sponsors reduce theirparticipant count and, as a result, also reduce their fixed rate PBGCpremiums. The fly in the ointment, however, is that while premiums arereduced, they may not fall as much as if the sponsor were to choose a strategyother than PRT.
PARTIAL PENSION RISK TRANSFERS: TWO SCENARIOS
Figure 1 shows a hypothetical plan with 10,000 participants and a 90% fundedratio ($900 million in assets and $1.0 billion in liabilities). The plansponsor is considering transferring the liability associated with smallbalances to an insurance company in an effort to reduce its liability value,participant count and PBGC premiums. For example, in exchange for providingannuity payments for $125 million of liabilities, the insurer will require asingle premium that exceeds the value of the liability transferred, as theinsurance company builds in conservatism margins, capital costs, profitloadings, etc. Assuming a 5% premium over the value of the projected benefitobligation (PBO) liability, our hypothetical plan sponsor would transfer$131.25 million of assets to annuitize $125 million of liabilities.
The funding ratio would fall given the plan’s initially underfundedposition and because a larger amount of assets than liabilities istransferred. Further, in fairness to participants who remain in the plan afterthe transaction, plan sponsors typically make a contribution to restore thefunding ratio to its pre-transaction level. In our example, the plan wouldneed to contribute $19 million to maintain the funding ratio at itspre-transaction level.
When all is said and done, it may appear as if the plan sponsor achieved itsobjective owing to a meaningful reduction in participant count (25% or 2,500individuals) and liability value (a 12.5% decline, or $125 million).
But how do we assess whether this constitutes a sufficient reward in light ofcosts and resources commitment incurred by the sponsor?
One could begin an assessment by quantifying the benefit in dollar terms andcomparing this to the “costs” incurred. In our example, the plansponsor achieved a $12.8 million reduction in its deficit and a $0.7 millionreduction in its annual PBGC premium. Overall, the plan gets advantages worth$13.5 million in exchange for a $19 million contribution to the plan. (Ourintent, of course, is not to compare these two numbers because the $19 millionoutlay is a one-time cost while some elements of the $13.5 million advantagewill be recurring).
The question remains, however: Would the sponsor achieve the maximum possiblebenefit for its $19 million commitment?
For an answer, consider a different way to use the $19 million. In scenario 2,the full $19 million is contributed to the plan without pursuing a PRT. AsFigure 2 shows, the plan sponsor would achieve
- A lower deficit and better funding ratio
- Roughly the same PBGC premium savings compared to the PRT scenario
All things considered, the plan would get a total benefit of $19.8 million ofdeficit reduction and premium savings (compared to only $13.5 million with thePRT transaction) for the same $19 million contribution and commitment ofresources (see Figure 2).
In scenario 2, the plan can achieve a better outcome without a PRTtransaction. While there are other situations where the PRT option may lead toa better outcome (e.g., with meaningfully overfunded plans or plans that aredeeply into the maximum per-head PBGC premium), it is important to recognize that PRT transactions donot necessarily lead to meaningful savings.
In many cases there is a better alternative. In scenario 1 (with PRT), forexample, the $19 million contribution is spent on reducing the least-costlyform of PBGC premium (fixed rate premiums are generally significantly lowerthan VRPs for underfunded plans) and funding the conservatism and profitmargins of the insurer see Figure 1).
In contrast, in scenario 2 (without PRT), the $19 million goes toward reducingthe most meaningful form of PBGC premium (the VRP) and counts completelytoward reducing the plan deficit instead of covering the insurer’smargins (see Figure 2).
While cost savings are helpful, the main objective of a pension risk transferis just that - transferring and eliminating risk from the plan sponsor’sbalance sheet. So before going down that path (and incurring significant costsand resource commitments), it is important to ensure that the transaction willindeed eliminate a significant amount of risk.
Let’s go back to our hypothetical plan and assume that the sponsor iscontemplating annuitizing its entire retiree population, which representsroughly 50% of the plan’s total liability. We will further assume thatthis plan has been on a de-risking glide path for several years and hasachieved a 75%/25% allocation between liability-driven investing (LDI) andreturn-seeking assets.
As shown by comparing the first and second columns in Figure 3, this PRTtransaction does not achieve meaningful risk reduction. In fact, the fundingratio risk (in percentage terms) actually goes up (doubling from 4.6% to 9.3%)if we compare the pre-buyout and post-buyout bars. That is because theliability transferred is actually the least risky portion of the obligation.Retiree cash flows are fairly predictable, have lower sensitivity to changesin interest rates or credit spreads and therefore are relatively low risk.
By transferring part of its liability, the plan is left with only the riskiestpart, which drives funding ratio risk higher. To be fair, the higher risklevel is applied to a smaller liability figure in dollar terms; so, it isimportant to look at the difference in risk on a dollar basis. As shown in thetable beneath Figure 3, the funding ratio risk in dollar terms is relativelysimilar before and after the buyout ($41 million before versus $35 millionafter). However, the $6 million reduction is likely insignificant relative to thesponsor’s market capitalization or balance sheet size. In other words,the benefit of a smaller liability value is offset by the increase in risk (inpercentage terms) and the sponsor is left without any meaningfulrisk-reduction benefits.
In addition, the PRT drastically reduces the funding ratio (to 75% from 90%).This is because it costs the plan $105 to annuitize every $100 of benefits, anoutlay covered by transferring assets from a plan that owns only $90 againstthe $100 of liabilities being transferred. In short, by accepting a significantlylower funding ratio after the transaction and if no additional contribution ismade, the participants who remain in the plan would subsidize participantsbeing bought out. As noted above, to avoid this most plan sponsors often makea contribution at or around the time of the buyout to maintain the fundingratio at its pre-buyout level. In our example, the plan would need to make a$75 million contribution to maintain its funding ratio.
In the end (see third column in Figure 3), the plan sponsor will have investeda significant amount of corporate personnel and financial resources to executethe transaction and contributed $75 million to the plan in exchange for a meager $10 million reduction inrisk. In our view, this is an underwhelming result.
In a way, the transaction is analogous to an individual with a diet dominatedby vegetables and donuts who decides to cut down on vegetables to lose weight.Although the vegetables contribute a small amount of calories to the diet(just as the retiree liability contributes a small amount of risk), cuttingdown on veggies while still enjoying donuts is likely to have limitedweight-loss benefits.
It is true than in other circumstances the trade-off may be more appealing.But this again highlights the importance of doing a thorough cost-benefitanalysis before investing significant resources in a PRT transaction.
BUT WHY DO ALL THE COOL KIDS DO IT?
When providing customized analyses to clients, we often get asked: Why doeseverybody do it if it’s that unattractive?
Actually, PRTs are less prevalent than the many public discussions,conferences, articles and webcasts might lead you to believe. The truth isthat only a tiny fraction of the outstanding stock of liabilities getsannuitized every year. Figure 4 shows the annual volume of PRT transactions inthe U.S. market in recent years. Using the last five years, the annual average is around $16 billion dollars. This represents approximately 0.5% of the stock of liabilities outstanding. Obviously, not everybody is doing it.
IF NOT PRT, THEN WHAT?
Many plans may be better off with a hibernation strategy. Hibernationimplements a likely low-risk liability immunization program that deploys planassets primarily in liability-matching investments. Hibernation strategies canhelp achieve the objective of reducing funding ratio risk and volatility tolevels that are extremely low, while alleviating issues that have plaguedsponsors in the past – such as underfunding, contribution volatility,high PBGC premiums, heightened complexity and significant commitments ofinternal resources.
Moreover, hibernated plans – assuming they are frozen to new accruals orclosed with small remaining benefit accruals – will shrink significantlyover five to 10 years as benefits are paid.
At that point many sponsors would most likely be comfortable sticking with thehibernation strategy. However, if desired, a PRT could potentially be achievedat a meaningfully lower cost given the smaller size of the plan. Thehibernation portfolio is also likely to constitute a better annuity-in-kind(AIK) portfolio after several years of hibernation as it would conceivablyhold a considerable proportion of seasoned bonds that are often prized byinsurance companies. This could further reduce the buyout cost through a moresignificant AIK discount.
HIBERNATE NOW. CONSIDER PRT LATER
We believe that partial PRT transactions are often costly when one considersthe premium relative to the PBO liability as well as the ancillary costsinvolved in execution – and may provide very limited risk-reductionbenefits. In addition, while there may be some small reductions of the PBGCfixed rate premium, plans can often get more significant cost savings bycontributing a similar amount of capital to assets than would have been spenton the PRT transaction. This option can help reduce the more costly PBGCvariable rate premium and simultaneously improve the plan’s fundingratio – as opposed to worsening it, as is often the case with a PRTtransaction from an underfunded position. We believe that a better course ofaction is to hibernate the plan for several years before considering a risktransfer transaction or full termination of the plan.