- U.S. pension law is under revision, and FASB pension accounting reforms will take effect in 2007 and 2008.
- Forthcoming FASB changes will not meaningfully alter the determination of DB plans’ funded status, and so we do not believe the Phase I and II reforms will have a dramatic market impact. In the U.K., the FRS 17 accounting reforms do appear to have been a revolution in funded-status determination and reporting.
- While likely reforms to ERISA laws will allow more lenient calculations of plan balances than do the accounting standards, the new ERISA rules will have the force of law behind them.
- It is the government funding requirement that has driven extremely low long bond yields in the U.K. In the U.S., however, requisite “full-funding” under ERISA law will be to a more lenient liability evaluation, and this will leave most plans with substantial leeway in deciding whether to pursue an asset allocation other than immunization.
- Still, optimal pension plan management means optimizing plan surplus rather than optimizing total return, and plans should make sure their investment strategy is consistent with this perspective.
I. Introduction and Summary
A succession of reforms to the U.K. pension system is credited with driving bond yields there to extremely low levels. Yields on long gilts are under 4%, and yields on inflation-linked U.K. bonds are under 1%, with 50-year “linker” returns only 0.5% and the yield curve distinctly inverted.
Here in the U.S., just the anticipation of pension reform has been credited by some with the “conundrum” of declining U.S. bond yields alongside a 350 BP rise in short rates. With a pension reform bill set for imminent passage in Congress and FASB reforms due to be in place starting next year, it is natural to ask whether reform will push U.S. pension plans away from stocks toward bonds, as has happened in the U.K.
We analyze here the potential impact of coming U.S. pension accounting and regulatory reforms, with an eye toward determining whether U.S. markets are in for a “flight into long bonds.” We outline the differences between existing—FAS 87—U.S. accounting standards and those in the U.K., as well as the likely upcoming changes in FASB specifications. Similarly, we discuss the likely implications of forthcoming changes in U.S. pension law.
Our conclusion is that U.S. pension reforms are not likely to have nearly as sharp a bond-market impact as has been observed in Britain. The accounting-system changes here will not be as revolutionary as those which hit the U.K. with the introduction of the FRS 17 accounting standard in 2000. More importantly, while upcoming government regulations will require U.S. plans to achieve full-funding of pension liabilities, U.S. funding targets will be less stringent than those in the U.K., leaving U.S. plans with the leeway to continue to target moderately aggressive asset allocations if they choose to.
It could be that a more compelling reason for U.S. plans to seek heavier bond exposures in their pension asset allocations will come from the pure economics of “liability driven investing” (LDI), rather than from accounting or regulatory mandates. We’ll save exploration of this possibility for a coming report that will focus on “active LDI” strategies.
II. FAS 87 And Its Successors
Current U.S. pension accounting rules are based on FAS 87, which was introduced in 1985. In its day, FAS 87 was a revolutionary shift, requiring pension plans to determine their funded statuses based on independent, market-driven, evaluations of plan assets and liabilities. Prior rules allowed plans to “discount” future pension obligations (liabilities) at the same rate as that used to project income on pension assets. Similar rules were still in place in the U.K. until FRS 17 was imposed.
FAS 87 specifies that plan obligations be discounted using current market rates on high-grade corporate bonds, while plan assets are to be evaluated at fair market value. A plan’s funded status—surplus or deficit—is the difference between plan assets and plan liabilities. As liability valuation now moves inversely with interest rate swings, plan funded status moves so as well unless assets have interest-rate-sensitivity (duration) close to that of liabilities (thus the pressure to move assets into bonds).
Critics complain that FAS 87 does not require a plan’s corporate sponsor to report plan surpluses or deficits on its balance sheet. Instead, funded status is currently reported in the footnotes to financial reports. Plan deficits needn’t even be amortized on corporate books unless they fall outside a “corridor” of 10% of plan liabilities. In contrast, FRS 17, upon imposition in the U.K., requires immediate, full, on-balance-sheet reporting of all pension plan deficits and surpluses.
The plan balances reported under FRS 17 (and IAS 19 outside the U.K.) are calculated in identical fashion to those under FAS 87 specifications. In other words, the accounting reforms since FAS 87 have not altered the basic formula for funded status. They have only shortened or eliminated reporting delays.
FASB is currently working on reform of FAS 87. It is set to report its recommended Phase I reforms this Spring, with imposition of these likely in 2007 and Phase II scheduled a year later. Phase I is expected to follow FRS 17 in specifying immediate reflection of all swings in funded status directly on the balance sheet. It is also possible that Phase II FASB reforms will remove firms’ ability (existing under all current standards) to report expected returns on plan assets within operating income.
As with FRS 17 and IAS 19, the forthcoming FASB reforms will not change the determination of funded status. However, there is substantial empirical evidence that the delayed reporting currently allowed by FAS 87 does not matter for financial markets. A host of academic studies have found that U.S. financial markets are able to extract pension plan information from financial report footnotes and fully incorporate it into market valuations, so that financial markets are already informationally efficient with respect to funded status as it is currently reported in the U.S.2
So while the new protocols will make life easier for financial analysts, they are most unlikely to have much market impact. Therefore, they are unlikely, by themselves, to affect corporations’ pension strategies.3 Now, corporate treasurers are concerned about the possibility of entering all pension effects directly on their statements of operating income, with the resulting volatility in earnings. However, that reform does not appear likely at this time, as we discuss below.
Before moving on to regulatory changes, we’ll mention one specific of FAS 87 that is relevant for government regulatory (ERISA) considerations. FAS 87 created a construct called the Accumulated Benefit Obligation (ABO), which is based on future benefits already earned and evaluated at current wage levels. In contrast, the liability valuation used for determination of funded status is calculated via the Projected Unit Method, using expected wage levels as of retirement, and it is called the Projected Benefit Obligation (PBO). When a U.S. plan has insufficient assets to cover even its ABO, let alone its PBO, it is required to immediately report an additional minimum liability, generally in the amount of the unfunded ABO. With FRS 17, all plan deficits are recognized immediately, so there is no point in distinguishing between accumulated and projected obligations.
The ABO is akin to a “liquidation” value of the plan. It is the amount firms are already legally committed to, and so it is the funding target that federal pension law mandates, as we emphasize below. In contrast, the PBO measures the value of benefits already earned that the firm and plan intend to honor and that they should plan for in their pension management. However, it will accrue only as the firm keeps it plan open or otherwise chooses down the road to fulfill that commitment level.
III. Upcoming U.S. Government Regulatory Reforms
Accounting standards can specify how a corporation determines the funded status of its pension plan and how/whether that status is reported in financial statements. They can’t force a plan sponsor to heed the “signals” provided by accounting statements. If a corporation is content with the market implications of its pension plan strategy, it is free to proceed with that strategy, so far as accounting standards are concerned.
Changes in government pension regulations are something else again. These have the force of law behind them, and governments have the power to levy penalties on plans that are not in compliance.
In the U.K., a Minimum Funding Requirement was instituted in 1997 requiring plans to “fully fund” their pension liabilities: to maintain assets at least as large in market value as plan liabilities. The Pension Act of 2004 linked funding requirements to an FRS 17-like liability valuation (equivalent to a PBO), with penalties for non-compliance including higher premia for (required) pension insurance. Our judgment is that these funding requirements are the major driver of the sharp declines in U.K. long bond yields.
To see this, observe that a risky or aggressive investment strategy--henceforth referred to as an “active” asset allocation—makes sense only if a sponsor can follow investment levels below “full-funding” as per accounting standards. A $100 pension liability on a bond-equivalent basis can be fully serviced by an investment of something less than $100 in a successful, actively-managed portfolio. Furthermore, while a corporate sponsor is required to make up shortfalls if its plan assets under-perform, it is generally not able to extract funds from the plan in the event of stellar asset performance. Thus, a plan sponsor cannot “enjoy” the benefits of stellar plan asset performance unless it is able to invest less than the full, bond-equivalent value of plan liabilities.
By the same token, if a plan is fully-funded on an accounting basis, the sponsor will lose if plan assets henceforth under-perform, and it will not normally benefit if the assets henceforth out-perform, so it may as well reduce the risk of the plan by moving to a passive, all-bond asset allocation. Thus, a mandate that requires full-funding of plans according to accounting determinations effectively pushes plans toward a passive (all-bond) asset allocation.4 This is exactly what U.K. law requires, with mandatory “full-funding” of a PBO-equivalent liability valuation, and this is why we believe the U.K. pension regulations are the dominant factor pushing U.K. long bond yields lower.
In the U.S., regulatory reform is due by Spring. The final bill will be a compromise of House and Senate bills (H.R. 2830 and S. 1783, respectively) initiated late last year. Both bills require U.S. plans to achieve “fully-funded” status within seven years and maintain it thereafter, penalties for non-compliance including higher premia on PBGC pension insurance (0.9% per year of deficits). However, “full funding” under U.S. law will be of an ABO-equivalent liability valuation. So while accounting standards have required and will continue to require firms to report plan funded status based on future pension obligations evaluated at projected future salary levels (PBOs), U.S. government funding requirements will base liability valuations on current salary levels (ABOs). Both the House and Senate bills contain this feature.5
A U.S. plan with ABO sufficiently lower than PBO will be able to pursue an active asset allocation that “under funds” on an accounting basis (PBO) but that is still “fully funded” as per government mandates (ABO). If a company’s plan includes substantial numbers of younger, active6 participants who will not retire for many years, its ABO will generally be only a fraction of its PBO. Even for large U.S. corporate plans, with plans dominated by older workers, most ABOs look to be around 85% of PBOs.
While Congress may not have intended this, we view its ABO funding mandate as a constructive compromise between the competing needs for economic leeway and pension insurers’ concerns. Responsible U.S. plans will still have the leeway to pursue active pension allocations, but required full funding of ABOs will reduce the temptation for plans to take extravagant risks. U.K. plans, facing required “PBO” funding levels, don’t enjoy this flexibility.
IV. What Doesn’t Matter
We have focused here on one aspect of accounting and regulatory reforms that will be important for every defined-benefit pension plan: the minimum-funding requirements that will be in place here and abroad. Meanwhile, there are a number of other issues that have gotten lots of media attention, but that will turn out to be non-events for the bulk of well-managed, healthy plans.
For example, House-Senate conferees will haggle over whether to allow one- or three-year historical smoothing of the discount rate to be used in ERISA liability valuations. Recall that under FAS 87 accounting rules, U.S. firms have already been evaluating liabilities using current market yields since 1985 and that market evaluations appear already to fully reflect these. Historical smoothing under ERISA affects the timing--but not the eventual magnitude--of companies’ required plan contributions. This will not matter to a healthy firm with a reasonably healthy plan.
Similarly, some commentators have decried the feature in FAS 87 whereby companies can project high returns on assets and take these to operating income whether or not the returns materialize. While this feature will likely disappear with Phase II FASB reform, keep in mind that the FAS 87 funded statuses reported as footnotes in current practice already utilize actual asset returns, and these are already “discounted” into market cap. Greater “transparency” here is a desirable but non-crucial change.
Finally, many analysts claim that with immediate reporting of pension balances directly on financial statements, corporate treasurers will move to reduce active asset exposure in their plans, so as to avoid the resulting volatility in reported earnings and net worth. With respect to net worth, empirical evidence, again, indicates financial markets are already incorporating these swings into market valuations. Even if you doubt this reliance on “academic” findings, consider the following fact: the average S&P 500 company has pension plan assets equal to 12% of its market cap. Of the 363 S&P 500 firms with DB plans, the median S&P 500 DB plan had plan assets of only 8.7% of market cap. So even if plan assets were fully invested in equities--with their 15% annual standard deviation of return--that would contribute to annual fluctuations of at most 1.3% per year in corporate net worth. As for operating earnings, while FASB is likely to remove companies’ ability to report expected asset returns on operating statements, there is no indication that it will move to the opposite extreme of requiring entrance of realized asset returns there.7 Remember also that pension plan contributions do not enter operating statements.
Treasurers may be fretting about these possibilities, but they will likely find the reforms actually undertaken to be less daunting. For most “healthy” plans, volatility in plan assets relative to liabilities will drive only slight, “noise level” volatility in corporate capital positions. For the troubled plans, well, they are already in trouble.
Pension reform will be important. Our guess is that it will not be important for the commonly cited reasons. Actual determination of a plan’s funded status will not change from what it has been in the U.S. for twenty years, and this determination is the most important aspect of pension accounting.
In the U.K., pension accounting was transformed tremendously with the recent reforms. Under SSAP 24, the previous regime, a British plan’s funded status was determined using the same discount rate for assets (returns) as for liabilities (obligations). FRS 17 has imposed a funded-status determination identical to that of FAS 87, without any of the reporting delays FAS 87 currently allows. More importantly, the British government has mandated “full-funding” of the resulting, PBO-equivalent evaluation. So suddenly, British plans have been thrust into an environment much more stringent than what U.S. plans will face even after reform is complete, and they faced this accounting/regulatory environment without a sufficiently deep long-duration debt market already in place. The resulting “flight into bonds” is understandable in this context. The chart on Page 1 shows how U.K. long bond yields declines coincided with the introduction of the MFR, FRS 17, and, more recently, the Pension Act of 2004.
In contrast, while in the U.S., regulatory funding requirements can be expected to be the dominant aspect of pension reform here as well, these will leave most U.S. pension plans the leeway to pursue an “active” asset allocation, if they desire it. Therefore, any flight to bonds here should be far less extensive than that seen in the U.K. We may already have seen the brunt of it, with the current, flat-yield-curve “conundrum.”
Evaluating plan liabilities at a discount rate different from that for assets can alter a pension plan’s optimal strategy: from one of optimizing total return to one of optimizing plan surplus. This is a profoundly important point and one that we will address in a coming report. Similarly, how a plan should pursue surplus optimization depends on whether it currently is “under-“ or “fully-funded” as per accounting protocols, whether its plan is open or frozen, and whether its workforce is sufficiently young to allow it to reap the benefits of an active asset allocation. We will address these issues as well in reports to come.
Accounting Standards Board (2000), Financial Reporting Statement 17: Retirement Benefits.
Bianco, David, Michael Constantinou, and Austin Burkett (2005), “S&P 500 Pensions: Time’s Running Out,” UBS Investment Research S&P 500 Pension Situation Update
Financial Accounting Standards Board (1985), Statement of Financial Accounting Standards No. 87, Financial Accounting Foundation, Norwalk CT.
- For those unfamiliar with the pension accounting items listed here, a good explanation is given in Bianco, Constantinou, and Burkett (2005).
- Jin, Merton, and Bodie (2004) perform their own statistical analyses on these counts, as well as reporting similar findings on seven other, preceding studies on both equity and debt pricing. Not only do Jin et al. find that plan balances are reflected in corporate pricing, but also they find that the riskiness of plan assets (reported in ERISA Forms 5500) is fully reflected in companies’ market risks (betas).
- This assumes, of course, that corporate managers are aware that financial markets are already utilizing plan pension information and so have modified their pension plan operations in light of this. Managers would be remiss if they hadn’t already responded to these realities. Meanwhile, while we do not expect the accounting reforms to push plans toward bonds, the economics of reducing surplus variance under liability-driven investment practices could induce such bond exposure, as we will discuss next time.
- The sponsor of a fully-funded plan could still benefit from stellar future asset performance, because the future surpluses so generated would allow it to forego plan contributions when workers earn further benefits in the future. This potential benefit is of only limited utility against the disadvantages of “full-funding” described here. This is so even in the U.K., where benefit streams are seen to be longer-lasting than in the U.S.
- This discussion of pending U.S. legislation is based on our own readings of pending bills, as well as commentary and presentation from Gordon Latter of Merrill Lynch and David Bianco of UBS Warburg. The House and Senate bills refer to funding requirements for what they call “Accrued Benefit Obligations.” While the terminology is different, the functionality between these and ABOs is essentially the same.
- Pension terminology designates a person who is still working and who is under an open pension plan to be an “active” participant in that plan.
- FRS 17 reports these swings in a “below the line” STRGL statement (defined in Exhibit 2).