Of all the issues confronting me, none is more crucial — or more complex — than funding. This is obviously true for those most immediately involved with defined benefit (DB) pension plans. Sponsors want to know what it will cost them to keep their pension promises, active members that the plan will be able to provide for them on retirement, and retirees that their monthly pension cheques will continue to arrive. But it is also true for others at one remove from the plan: pension regulators want to avoid plan failures, tax collectors the illicit sheltering of corporate income, union members the loss of much-valued and hard-won benefits, and investors and creditors the unanticipated “legacy” costs of maintaining a pension plan. And at the most general level, everyone concerned with the social and economic health of this province will want to have an accurate idea of how well-funded or otherwise are its occupational pension plans.
However, there seems to be little consensus about how to establish rules that will ensure that plans are “adequately” funded. This is partly because any given set of funding rules will have variable effects that depend on the design and benefit structure of plans; on the demographic profile of their members; and on the different points in the business cycle, the trajectory of the sponsor’s business fortunes, and the rise and fall of interest rates and equity prices at which “adequacy” is being tested. But it is also partly because stakeholders differ over what is “adequate” funding, not only among themselves and within each group, but, depending on the issue, from one stage in the life of a plan to the next.
This last observation deserves further consideration. Rules designed to produce greater security for retirees, for example, may drive up the sponsor’s costs or require earlier payment of contributions, thus creating resistance toward improving pension benefits, increasing other parts of the compensation package paid to active plan members, or even continuing to operate the plan. Rules designed to make the sponsor’s costs more “reasonable” or “realistic” may somewhat impair the security of the plan, but may also make the sponsor’s business more viable and, over the long run, improve the prospects of well-paid employment — and good pensions — for its workers. Rules designed to encourage higher levels of funding may encourage sponsors to pursue more aggressive — and riskier — investment strategies, which in the end makes plans less, rather than more, secure.
Nor, as I noted in Chapter Three, are stakeholders alone in experiencing ambivalence. Public policy makers also have difficulty in resolving funding issues. For example, my terms of reference instruct me to both ensure the “viability” of the DB pension system and to propose solutions that are “affordable.” To some extent, however, “viability” may involve a modest increase in expenditures — somewhat less “affordability” — both by sponsors to maintain higher levels of funding, and by government to promote better plans and plan governance and to more effectively regulate the system.
I underline this last point. If recommendations in this chapter concerning the reform of funding rules are to have any chance of successful implementation, they must be accompanied by significant investment in the enhancement of the regulatory machinery and by the reform of governance, proposed in Chapters Seven and Eight, respectively.
I make these points not merely to acknowledge how difficult it is to design “scientific” — or even “sensible” — funding rules that will serve everyone’s interests all the time, but to explain why I have defined my own ambition in this chapter much more modestly.
I begin with a baseline conviction: that however absolute the pension promise might appear to be, some degree of risk is inevitable in pension funding. From this conviction follow five propositions that inform my analysis and recommendations: (1) that risk assessment should be evidence-driven and not outcome-driven; (2) that in order to be evidence-driven, risk assessment should be as little influenced by wishful thinking as possible; (3) that pertinent assumptions should be made transparent, and relevant information should be disclosed; (4) that risks accepted when taking decisions affecting others should reside within narrower tolerances than risks knowingly assumed by people for themselves; and (5) that for all of these reasons, funding rules should be designed so as to force participants to make hard choices as explicitly, thoughtfully and responsibly as possible.
Bold language, perhaps — but as noted, modest ambitions. The conventional terminology used to talk about risk, the expert methodologies used to assess it, and the range of regulatory measures used to limit or accept it, are all too deeply entrenched for me to change or even challenge them to any great extent. However, where a case is made for change, I propose change in a way that accords so far as possible with the five propositions set out above.
This leads me to a final introductory observation. When change is introduced into a complex system such as the pension system, it must be done in a way that allows the stakeholders and the regulator to adjust to new requirements. New financial arrangements will have to be made, new analytical systems will have to be installed, new reporting and vetting procedures will have to be designed, and new rules and methodologies will have to be assimilated by all participants. I therefore emphasize in Chapter Ten that many — perhaps most — of my recommendations should be accompanied by transitional measures designed to bring them into effect as rapidly as possible, but not so rapidly as to destabilize pension plans, the employment relations they support or the regulatory regime that oversees them.
4.2 Theories of Pension Funding and Pension Policy
Conventional terminology, expert methodologies and regulatory measures are often explained or justified by “theories” of pension funding whose descriptive and prescriptive force is sometimes more evident to their authors than to other people.
Active member and retiree groups often characterize the pension fund as part of their total compensation package, as deferred wages for work already performed, and consequently, as property set aside for — and in that sense “owned by” — the beneficiaries. This theory of what pensions are and who owns the pension fund leads them to question any approach to funding issues that might reduce the accumulation of fund assets or render them unavailable for pension purposes. The most well-known and controversial claim — to those who hold contrary views — is that surplus funds that accumulate in the plan belong unconditionally to the beneficiaries. Some important court decisions and legislative developments that have given beneficiaries’ claims on pension funds precedence over those of the sponsor are cited by proponents of this theory as supporting their position.
Employers, on the other hand, begin with the theory that their obligation is no more and no less than to deliver specified benefits to those entitled to them. The purpose of a pension fund is to earmark and accumulate assets sufficient for that purpose. If the assets are insufficient, the employer must make up any deficiency. However, if the assets grow to a size in excess of requirements, the surplus — on this theory — logically belongs to the sponsor whose contributions have ultimately made growth possible. This explains why sponsors characterize the current funding rules as “asymmetrical:” they are required to cover all risks and fund any deficits, but do not get the unconditional use of the surplus on partial or full wind-up.
In his research paper prepared for the Commission, James Wooten analysed these arguments and found them both less than totally compelling. The deferred wage theory, he notes, does not seem to account for many aspects of the current rules, and most especially does not explain why the sponsor must augment the fund not only when the time arrives to make good on the pension promise but also during the life of the plan. At the same time, he points out that the asymmetry argument fails to address two key facts: that active plan members and retirees bear some of the risks associated with the plan, and that employers whose ongoing plans are in surplus generally have the right to take “contribution holidays.” Nor, he notes, is there any clear explanation under either theory of whether individuals whose employment is terminated prior to retirement should receive an amount in lieu of their pension, whether it is based on accruals to date or on future expectations as well. Further, Wooten describes the current surplus-sharing rules as “incoherent” and as creating sequential entitlements: employers may use the surplus in the plan during its lifetime, while employees may have access to it on partial or full wind-up. Finally, Wooten draws attention to U.S. tax rules, which discourage sponsors from claiming any surplus in the plan upon wind-up, and suggests that these rules contributed to the low funding levels at which U.S. plans were maintained, even during the 1990s when plans were achieving high investment returns.
Of course, even those who espouse different theories of pension funding acknowledge that rules adopted in deference to them may — and should — influence plan performance and public policy outcomes. I heard from many that more flexible rules, including greater employer access to surplus, would encourage sponsors to maintain existing plans, enhance them and even, perhaps, create new ones. Such rules, I was told, are more consistent with the realities of funding plans over the long term in inevitably volatile conditions. At the same time, I was told that flexible funding rules are unlikely to make much difference to employers who already have plans, or to those who might consider establishing them in the future. As I indicated in Chapter Three, much empirical evidence indeed suggests that the presence of unions, firm size and general labour market conditions — not funding rules — have been the prime determinants of coverage under our voluntary DB pension system. However, it does not follow that my only objective in reforming the funding rules should be to ratchet up the financial security of existing plans in order to protect active members and retirees. Not only do sponsors make some sensible arguments in support of their legitimate interests; it is of the essence of a voluntary system that those arguments should be acknowledged and interests accommodated, if possible.
A review of these competing “theories” of plan funding, of Wooten’s telling critique of them and of their possible implications for the future of the system suggests that while they may have some appeal, they do not lead inevitably, predictably or consistently to any particular set of funding rules. This conclusion makes it necessary for me to reiterate that my own attempt to clarify and recast the pension funding bargain proceeds not so much from a single grand theory of pension funding as from a determination to somehow strike a fair balance among conflicting theories, principles and interests.
I acknowledge, of course, that any attempt to strike a fair balance will affect existing rights, interests and expectations to some degree. However, in my recommendations in this and succeeding chapters, I have tried to do so only to the extent necessary to achieve fairer outcomes for the pension system as a whole. In general, I have preferred solutions that favour more plan members over those that favour fewer, solutions that enhance long-term system stability over those that produce occasional advantages for one party or the other, and those that make for clarity over those that contribute to ambiguity and uncertainty.
4.3 Measuring Funding
4.3.1 Factors affecting the measurement of funding
Three distinctive features of pension plans make the measurement of plan funding particularly difficult. First, the pension promise must typically be made good over decades. An employee who is, say, 30 in 2008 may retire in 2043 or 2048 and continue to draw his or her pension until 2068 or 2078 — or even longer. And because pension plans enrol new members all the time, the projected life of any plan is constantly being extended to take account of the promise made to its youngest recruits. Second, the factors that define the value of the pension promise — years of service and either a percent of salary (in final or career average plans) or a fixed dollar amount per unit of service (in flat benefit plans) — are not constant. They change because the makeup of the plan population changes, because life expectancy changes, because salaries and promised benefits change, and because business conditions and regulatory requirements change.
Third, the cost of paying for the pension promise is highly volatile. In principle, the plan’s assets must suffice over time to meet its evolving long-term liabilities — the future costs of honouring promises made to current employees and retirees. But those assets comprise three elements: annual contributions calculated actuarially every three years, special payments to make good any deficiencies experienced during previous periods, and returns earned on the investment of these contributions and special payments. So long as sensible assumptions have been made, and no unusual events have occurred, the first two of these three elements can be managed reasonably well. It is the third that generally causes the valuation of assets to fluctuate considerably. Pension funds invest in bonds, equities and other types of assets. If bond yields rise, if equity prices fall, if interest rates fluctuate, if the worth of other holdings wobbles, the value of the plan assets may either exceed or fall well short of what is needed to meet the obligations of the plan — or do both within a fairly short time span. Indeed, because the plan must provide departing employees with lump sums or annuities in some circumstances, the same events that cause asset values to fluctuate also affect the value of its liabilities.
4.3.2 The role of expert methodologies in measuring plan funding: actuarial and accounting practices
For all of these reasons, the valuation of pension plan assets and liabilities is a formidable challenge. However, it is a challenge that must be met at least every three years when, as required by Ontario’s Pension Benefits Act (PBA), the plan administrator files a valuation prepared in a manner “consistent with accepted actuarial practice and with the requirements of the Act and this Regulation.” These triennial valuations set the requirements for funding over the next three years, enable the sponsor to calculate its obligations to contribute to the pension fund, and provide the regulator with a benchmark to assess whether adequate contributions have been made. If a valuation discloses that a plan is funded at less than 80% of its liabilities (or less than 90% for larger plans), annual valuations are required until the plan’s funded status improves to the relevant threshold.
Funding valuations, as noted, must conform to accepted actuarial practice — in effect, to the professional standards set by the Canadian Institute of Actuaries (CIA). All plans are required to submit to two methods of valuation: “going concern” and “solvency” valuations — but a plan must be funded according to whichever method produces the highest level of funding. If a valuation reveals a deficiency in the plan’s funding, that deficiency (along with normal cost contributions) must be paid into the fund (“amortized”) over a fixed period of years.
Going concern valuations have been required under Canadian legislation for some time. As the name suggests, such valuations assume that the plan will continue indefinitely and will predict how liabilities and assets are likely to accumulate in the future and thus identify so-called “current service costs” — the contributions required to enable the plan to pay for the liabilities it will predictably encounter over the next three years. In addition, a going concern comparison of the plan’s projected performance with its actual performance over the past three-year period can generate either a going concern surplus (an “actuarial gain”) if better than expected, or an unfunded liability (an “actuarial loss”) if the experience was worse or if new unfunded benefits have been added. Unfunded liabilities or losses must be amortized over 15 years; gains are available to reduce the amount to be paid by way of scheduled contributions (a contribution holiday).
Because going concern valuations focus on the long term, they must inevitably rely on assumptions about how the plan’s liabilities and assets will evolve in the future. Although the choice of assumptions has a considerable impact on the level of contributions that will be required from the sponsor, the PBA regulations and the CIA’s standards of practice allow more latitude to actuaries conducting going concern valuations than solvency valuations. Nonetheless, going concern assumptions are also generally required to contain a margin of conservatism. This margin is intended to incline plans toward surplus rather than deficit.
Solvency valuations— unlike going concern valuations — assume that the plan will be wound up immediately. This means that solvency valuations rest on a less conjectural foundation. This has both positive and negative aspects. On the one hand, plans need not take account of projected salary and benefit increases in the context of solvency valuations. On the other, assets and liabilities are valued by reference to current rather than conjectural conditions in financial markets, although they may be averaged or smoothed over a short time. This last point is of considerable significance. Under existing pension regulations, solvency valuations proceed on the assumption that annuities must be provided for all retirees and for those eligible to receive a pension “immediately.” Plan assets must therefore be sufficient to cover the cost of such annuities under current market conditions. Depending on long-term interest rates, which effectively set the price of annuities and lump sum payments out of a pension plan, this can be an onerous requirement. Further, if a solvency valuation reveals a deficiency in the plan’s funding, that deficiency (along with normal contributions) must be amortized over five years, rather than 15 years as with going concern valuations.
Solvency valuations often require plans to be funded at higher levels than they would be under going concern valuations. However, this outcome is by no means inevitable. Solvency valuations came into force in 1992 after a period when many plans were perceived to be under-funded. However, for about a decade thereafter, under very favourable market conditions, most plans were funded to going concern valuations. Then, from about 2000 onward, solvency valuations dominated funding requirements, with plan liabilities being driven upward as the cost of providing annuities and lump sum payments out of a plan increased steadily due to the decline in long-term interest rates.
These shifts over the past three decades have triggered a debate over the funding rules. With no change in its actual circumstances, a plan that was fully funded at one moment might swing into deficit the next (or vice versa), depending not only on actual changes in its assets and liabilities but also on the methodology of whichever valuation method happened to predominate in the economic circumstances of the moment. Moreover, a shift in the plan’s funded status might, in turn, precipitate a shift in the quantum and rate of payment of sponsor contributions and special payments.
Some stakeholders who were adversely affected by shifts and swings at particular moments in time proposed that their plans be exempt from whichever funding method appeared to be the source of their discomfort — most recently, solvency funding. On the one hand, they argued, solvency funding increases the volatility and overall cost of funding their plans; on the other, solvency funding is premised on the possibility that plans will wind up with insufficient assets, a fate that, for various reasons, is most unlikely to befall certain types of plans. Their concerns are understandable and are addressed elsewhere in this chapter. However, a study undertaken for the Commission by Brian FitzGerald suggests that the current funding rules actually do allow plans to maintain reasonable funding equilibrium over time, unless they encounter the extreme but atypical volatility that characterized the business and financial environment during the “perfect storm” years early in this decade. This seems to suggest that tweaking, rather than transforming, the present funding rules would be the wisest approach.
Finally, in addition to being actuarially valued for regulatory and funding purposes, pension funds are also valued so that their effects on the sponsor’s overall financial health can be accounted for. In recent years, internationally accepted accounting standards have adopted “mark to market” principles, which increasingly reflect the logic of solvency valuation. This tendency is likely to ensure that pension plans are treated on the sponsor’s balance sheet in ways that emphasize their cost and volatility, and thus to contribute to their growing unpopularity with corporate financial officers and investors.
In summary, estimating the adequacy of plan funding is intrinsically difficult in light of the three broad factors outlined in section 4.3.1. However, that difficulty is compounded by the fact that while actuarial and accounting rules may provide an accurate picture of the financial state of plans at a given moment in time, these rules can neither anticipate nor recapitulate their dynamic character.
4.4 Common Valuation Rules: Transparency and Timing
While each plan — indeed each type of plan — presents special problems of valuation, certain aspects of the valuation process are common to all plans. Specifically, all valuations ought to be as transparent as possible, and all plans ought to be valued not only at fixed intervals but also more frequently, if circumstances warrant.
The transparency of valuations, funding decisions and other operational matters is essential to any system of pension regulation. It provides regulators with what they need to know to prevent harm to plans or to the system, and to enforce the law if infractions occur. It reminds sponsors that funding decisions must be reasonable and based on sound analysis. It ensures that actuaries will provide analyses capable of standing up to third-party scrutiny. It reminds plan administrators that their decisions may be subjected to critical oversight by plan members as well as regulators. And most of all, it provides active members, retirees and their advisors and advocates with a window on the pension plan, which may be — in a practical, if not a strictly legal, sense — among a family’s largest investments.
However, valuations — in the opinion of many — are not at present sufficiently transparent. The issues that raise transparency concerns flow from regulations that:
- allow certain benefits to be provided without requiring them to be funded or included in the plan valuation;
- allow the spreading or smoothing of changes in certain variables in the funding formula;
- fail to require disclosure of some pertinent funding information; and
- rely on actuarial standards and practices that are opaque or imprecise in certain respects.
Each of these concerns warrants discussion.
Excluded benefits are mainly related to solvency valuations but have some application to going concern valuations. While all jurisdictions in Canada require both valuations, most do not allow benefit exclusions.
For example, while indexation paid to current retirees must be included, the cost of future indexation — whether on a formulaic basis or ad hoc — need not be included in either solvency or going concern valuations. While exclusion can be seen as an inducement to sponsors to provide this form of protection, indexation is generally an expensive benefit, so its exclusion may substantially undervalue the liabilities of a plan.
To take another example, plant closure benefits must normally be funded and included in a valuation. However, they may be excluded if they are provided pursuant to plan-specific arrangements in place prior to 1991. While this exclusion amounts to a form of “grandparenting” for a limited array of plans, it too represents a hidden factor contributing to their actual cost.
Benefit improvements — even expensive improvements related to past service — may be amortized over a period of five or 15 years. This means that although plan members are entitled to the additional benefits as from their inception, no contribution need be made immediately to cover the cost of providing the benefits. Indeed, such benefits may be introduced even though the plan is already less than fully funded. Although these are not benefit exclusions per se, they have the same effect for certain kinds of plans. In particular, a flat benefit or career average plan can avoid including anticipated benefit increases based on increasing salaries, whereas a final average earnings plan cannot.
Ontario regulations allow limited “smoothing” of asset values and discount rates in solvency valuations — that is, they allow a deferred recognition of gains and losses on investments, and a discount rate averaged over a period of time. CIA standards, applicable across Canada, allow smoothing on elements of the going concern valuation. However, the degree of smoothing allowed for solvency valuations in Ontario is greater than in other provinces. The 15-year amortization period of going concern unfunded liabilities is also arguably a form of smoothing, as is the five-year amortization period for solvency deficiencies.
There are good reasons for smoothing. It acknowledges the long-term nature of the obligation and avoids contributions being subjected to sudden and extreme changes. However, smoothing methodologies for going concern valuations are not carefully defined by actuarial practice, and the potential exists for changes in smoothing to be used not to respond to altered circumstances, but opportunistically to hide a funding problem. More importantly, smoothing can detract from clear understanding of a plan’s funded position if it is not fully explained in the valuation report.
Valuations should be transparent not only to the sponsor, the plan administrator and the CIA’s professional standard-setting and discipline bodies, but also to the regulator and to active plan members and retirees. One important matter not presently provided in an actuarial valuation is whether a contribution holiday — a reduction or suspension of normal cost payments — is being factored into the contribution schedule. The rules currently require a valuation to identify surplus in a plan, which would permit it to take a contribution holiday; however, the rules do not require disclosure of whether that holiday has actually been factored into the proposed three-year schedule of contributions. Quite apart from whether or when contribution holidays are appropriate, the fact that they are going to be taken should be transparent. Information about contribution holidays is essential for an understanding of plan funding, both for the regulator and for all plan participants, and should be provided in a document that is fully accessible to them.
Actuarial standards and practice
Many reports and calculations under the Act and regulations must be prepared by a Fellow of the CIA (a self-regulating, professional body) and in accordance with its standards and with accepted actuarial practice. This requirement makes the CIA and its members in effect part of the apparatus of pension regulation. Of course, the extent of their role depends largely on their willingness and ability to anticipate, respond to and reinforce changing regulatory strategies — greater transparency not least among them — and on the willingness of the regulator to pre-empt, defer to or supplement professional norms.
Happily, the CIA has recently changed its standards to improve the transparency of actuarial valuations. For example, all material assumptions must now be explained in a valuation. And it is likely that the new CIA standards will also require each assumption to be “independently reasonable,” rather than collectively producing a reasonable outcome, as at present. These two developments together would greatly enhance the transparency of valuations, and it would be helpful if the CIA were to promptly adopt the second as well as the first. However, if it is unable to do so, the government retains the power to require it by regulation.
The question of actuarial discretion is another area that affects the transparency of valuations, especially in connection with going concern valuations. Key issues include the selection of appropriate discount rates and mortality tables. In effect, how the actuary exercises professional discretion with regard to these matters determines the range of choice available to the sponsor in determining how much to contribute to maintain the plan’s funding. This situation exposes actuaries to subtle — even overt — pressures to exercise their discretion in a way that produces outcomes agreeable to the sponsor. In the end, of course, the sponsor makes the choice — but it is actuarial discretion that confers legitimacy on that choice.
Whether — and if so, by what means and to what extent — actuarial discretion should be narrowed or structured is a matter of controversy. While narrowing or eliminating discretion would contribute to greater transparency in valuations (and incidentally, insulate the actuary from sponsor pressures), it would also prevent the actuary from capturing the individual characteristics of different plans.
A recent United Kingdom attempt to reduce or eliminate actuarial discretion is generally thought to have had perverse consequences, and that country has recently moved to restore actuarial discretion to something approaching its former scope — and perhaps even to widen it. The Commission’s research and a report recently published by the Financial Services Commission of Ontario (FSCO) both suggest that the actuarial profession in Canada is itself addressing the issue. Discount rates used in going concern valuations are gradually becoming more conservative and the CIA has indicated that it will be identifying the permissible maximum rates for going concern valuations. In addition, mortality tables have been updated substantially over the last few years for virtually all plans, making valuations more accurate. Indeed, a small number of large plans now use plan-specific, customized or modified mortality tables, which should improve the accuracy of valuations so long as CIA standards are flexible enough to permit this approach.
Flexibility, of course, does not imply a complete absence of concern, structure or monitoring. On the contrary, as noted, the CIA in recent years has itself been increasingly sensitive to the need to ensure transparency and to tighten up areas of professional practice that might give rise to inappropriate valuations. This approach, supplemented by constructive engagement with the regulator, stakeholders and other professionals, should continue to produce positive results. If not, the adoption of formal rules or regulations by or under the authority of the PBA remains an option.
Recommendation 4-1 — The Superintendent should work with the Canadian Institute of Actuaries to ensure that actuarial standards and practices continue to evolve in the direction of greater transparency and more structured discretion. For example, actuarial valuations should reveal the reasons behind the assumptions used in valuations to set discount rates and to select the mortality trends used to calculate plan liabilities. They should also reveal whether the sponsor intends to take a contribution holiday.
Recommendation 4-2 — The Superintendent should have the power to require that plans cease using assumptions that are unreasonable or that depart materially from accepted actuarial practice, and to order an independent valuation or peer review of a report, at the expense of the plan, if there are grounds to believe that the actuarial valuation misrepresents a material factor in its funding.
The changes in valuation rules, outlined above, will undoubtedly improve transparency — a good in itself; and some may also enhance the reliability of actuarial valuations, the autonomy of actuaries, the security of plans and the efficacy of regulation. But there is no denying that the cumulative effect of valuing all benefits, disallowing smoothing, structuring actuarial discretion and ordering additional valuations could mean a substantial shift in the timing of contributions or, indeed, an increase in the overall level of contributions required from sponsors. Since some of these practices are widespread and some less so, they will have varying effects on plan sponsors. Nonetheless, it seems quite likely that the measures I propose will exacerbate the volatility of funding for some sponsors, and lead many to resist even more strongly the introduction of new or enhanced benefits, including indexation. In the latter event, I suspect that many active member and retiree representatives will align themselves with sponsors, since they would rather have a contingent promise of larger benefits in the future than a greater certainty of receiving the lower benefits presently agreed.
For these reasons, so far as possible, the adverse consequences of increased cost and volatility should be avoided, minimized or offset. How can this be achieved? One possibility is that sponsors might be given an extended amortization period over which to fund the higher contributions expected of them. Another is that a methodology might be developed so that well-funded plans might be relieved of certain obligations to which under-funded plans are made subject. A third is that access to new or additional benefits may remain contingent until they are either fully funded or reduced, so that they are paid only to the extent to which they are funded. These and other solutions to the problem of increased cost and volatility obviously involve risks as well, especially the risk that the sponsor may encounter financial difficulties during the extended amortization period. However, in my judgment, the risks associated with longer amortization periods are lower than those generated by current rules and practices that discourage transparency and impair clear-minded decision-making. On balance, then, I favour enhancing transparency and acknowledging a degree of risk-taking because, as I confessed earlier, I believe that funding rules should be designed so as to force participants to make hard choices as explicitly, thoughtfully and responsibly as possible. Valuation and funding rules that tend to obscure relevant information should be minimized.
Recommendation 4-3 — Going concern valuations should no longer permit the exclusion of promised indexation benefits. Solvency valuations should no longer permit the use of smoothing practices or the exclusion of benefits. A special exception should be made for those plans that continue to provide plant closure benefits pursuant to a specific, long-standing commitment to continue their non-funded status.
Potential increases in sponsor contributions attributable to these enhanced transparency measures should be offset so far as possible by the extension of amortization periods, by selective relief from contribution increases for well-funded plans or by other means.
4.4.3 The timing of valuations and reporting
The current reporting rules require full actuarial valuations to be performed for each plan every three years on both a solvency and going concern basis. An additional nine months is allowed for the actual filing of the valuation, with further delays permitted if the regulator consents. Annual valuations are required for plans that have fallen below a certain level of funding, as set out in the regulations. All plans are also required to file annual financial statements, which contain a considerable amount of detail and provide some insight into their current level of funding.
It was put to me that the triennial norm — especially when extended due to statutory and discretionary delays in filing — allows too much time for a plan to shift from surplus to deficit, or to go from a modest deficit into steep decline. There is some suggestion, for example, that during the “perfect storm” of the early years of this decade, some plans took contribution holidays based on their last filed valuation despite an intervening sharp decline in their funded status. To prevent such inappropriate actions, and to bring developing problems into timely focus, many submissions to the Commission proposed that reporting should occur more frequently. While most European countries and Canadian jurisdictions require triennial valuations, the United Kingdom and Quebec require annual “mini-valuations,” in addition; the United States requires full valuations every year.
However, full annual valuations on the U.S. model represent a significant additional cost to sponsors, plans and, by extension, other stakeholders. Mini-valuations seem like a promising alternative, but since they are likely to present a partial, distorted or unclear picture of the plan’s finances, they may metamorphose into full valuations. (Perhaps for this reason, Quebec has yet to implement its mini-valuation requirements.)
An alternative approach is to require more frequent filings by plans that are — or are likely to be — in difficulty. Annual valuations are already required for Ontario plans whose triennial valuation indicates that they are funded below a threshold defined in the regulations. What is needed is a strategy for identifying plans whose funding deteriorates between triennial filings to the point at which the plan is at risk. The Office of the Superintendent of Financial Institutions (OSFI), the federal pension regulator, several other Canadian regulators and their U.K. counterpart all monitor plans more proactively than does Ontario: if they find cause for concern, they request interim valuations.
This approach has the merit of placing the extra burden of more frequent reporting on a limited group of suspect plans rather than on all plans. Since the value of an actuarial valuation declines as time passes, it has the additional virtue of providing the regulator with information that is as up-to-date as possible. However, the success of this approach depends upon the regulator having the capacity to identify economic conditions and other factors that may give rise to problems across the pension system, to develop plan-specific benchmarks and to monitor plans more closely when they begin to appear on its “radar screen.” The need for greatly expanded capacity in Ontario’s regulator is dealt with at greater length in Chapter Seven.
Recommendation 4-4 — The current requirement for an actuarial valuation every three years should be maintained. The time for filing the valuation after it is due should be reduced from nine to six months. Extensions should be given only in exceptional circumstances.
Recommendation 4-5 — Plans whose triennial valuation shows that their funding has fallen below a threshold to be specified by regulation should continue to be required to perform and file an annual valuation.
Recommendation 4-6 — The Superintendent should develop the capacity to monitor the pension system, and individual plans, more closely, and should have the power to order an interim valuation at any time if there are reasonable grounds to believe that a particular plan is at risk of failure.
Finally, it is important that plan sponsors and administrators and their professional advisors understand that promptness and accuracy in filing valuations and other reports are essential. While these standards are no doubt met in most cases, the inability of the regulator to deploy adequately trained staff to closely monitor filings may mean that some transgressions are going undetected. It is certainly not optimal that valuations and other reports submitted by plan administrators should, for the most part, be read superficially (or not at all) long after their submission. At the very least, those who file late or inaccurate documents should know that they stand a reasonable chance of being detected and sanctioned. The same approach should be taken with other participants whose tardiness or non-compliance with prescribed filing requirements delays timely reporting. I have in mind, for example, employers with multi-employer pension plans (MEPPs) who fail to provide the plan with a list of employees and the contributions made on their behalf. While the Superintendent and his staff are already armed with the necessary powers, and use them to some extent, I am recommending heightened vigilance to achieve higher and faster rates of compliance.
Recommendation 4-7 — The Superintendent should more aggressively discourage and more predictably sanction late filings, and develop a capacity to scrutinize filings to the extent necessary to improve the likelihood that inaccuracies will be detected.
4.5 One Size Does Not Fit All: Ensuring That Funding Rules Are Appropriate to Each Plan Design
The existing funding rules treat single-employer defined benefit plans as paradigmatic — as the plan model from which limited exceptions and departures have been permitted. This approach is problematic for at least two reasons.
First, while DB single-employer pension plans (SEPPs) number in the thousands and represent the most ubiquitous type of plan design by far, SEPPs cover only 31% of all workers enrolled in DB plans (and 41% of workers with any kind of pension coverage). In terms of membership, if not numbers of plans, MEPPs and/or jointly sponsored pension plans (JSPPs) clearly dominate Ontario’s DB pension system. Second, as Table 1 demonstrates, MEPPS and JSPPs display different institutional characteristics from SEPPs, including their arrangements for sponsorship, governance, contributions, benefits; their strategy for dealing with risks and Pension Benefits Guarantee Fund (PBGF) “insurance” coverage; their union involvement in establishing the plan; and, to a lesser extent, their location in either the public or the private sector.
|Percentage of workers enrolled||35%||34%||31%|
|Sponsorship||Joint with members (multi- or single employer)||Multi-employer sponsor or joint with members||Single employer|
|Governance||Joint||Members only or joint with sponsors||Sponsor only or some member involvement (rare)|
|Contributions||Contributory||Sponsor only or contributory||Sponsor only or contributory|
|Benefits||Fixed but reduced on wind-up if under funded||Target: adjust to available funding||Fixed|
|Risk distribution||Shared||Members||Sponsor or shared|
|Union involvement||All existing plans||Usual||One-third based on collective bargaining agreement|
|Sectoral location||Public or private (none presently private)||Public or private||Public or private|
The question is whether these differences in design necessitate or justify differences in funding rules.
A brief review of each design element suggests that this is indeed the case.
A SEPP is sponsored by a single employer; the chances of that employer getting into financial difficulty to the prejudice of the plan are much greater than in the case of a MEPP (or most JSPPs), where multiple employers are available to sustain the plan, even if one goes under. Many MEPPs and most JSPPs are more able to spread risks and amortize costs over a larger member base than almost all SEPPs. Funding rules should therefore be designed to take account of the different risks inherent in each plan type.
SEPPs are almost always governed and administered by the sponsor acting unilaterally, although in principle nothing prevents a SEPP sponsor from agreeing to the participation of active and retiree members in governance procedures. If one of the purposes of the funding rules is to ensure that the interests of beneficiaries are properly safeguarded, that purpose becomes easier to achieve when the beneficiaries themselves have a significant voice in decision-making that affects them, as they do in the case of MEPPs and JSPPs. Of course, it becomes especially important that governance in these plans meets high standards.
By definition, SEPPs depend on sponsor contributions. So do JSPPs, but their active members share responsibility with the sponsor for maintaining a level of contributions sufficient to keep the fund solvent. MEPP sponsors, however, cannot be required to increase their contributions for the duration of the collective agreement by which they are fixed, even if the fund is in deficiency. Consequently, for MEPPs, and to a lesser extent JSPPs, contributions effectively determine benefits; the opposite is true for SEPPs. Funding rules might sensibly respond to this important distinction.
In DB SEPPs, benefits are by definition “defined” or fixed. Moreover, once accrued, they cannot be reduced. As noted, the obligation to provide these benefits defines the amount the sponsor must provide to keep the plan solvent. If the plan has insufficient assets to make good the pension promise, the sponsor must make good any deficiency by way of special payments amortized over a number of years. In MEPPs, however, the plan is committed only to providing a target benefit. If the target cannot be achieved with the available funds, benefits may be reduced (including accrued benefits and pensions already in pay). JSPPs are somewhere between the two: accrued benefits are normally regarded as fixed, but can be reduced if, upon being wound up, the plan turns out to be under-funded. In practice, moreover, their joint governance structure allows JSPPs to continually rebalance contributions and benefits, as required, by the plan’s funded status, and to deliver benefits on a contingent basis in the sense that their provision or magnitude depends on the availability of funding. A strong case can be made for having different funding rules for plans with defined benefits and those with target or contingent benefits.
SEPP sponsors claim that they bear the risks associated with plan funding. This claim is challenged by plan members who argue that in various ways, they too assume risks — especially those associated with under-funding and plan failure. Without purporting to resolve this conflict, it is at least clear that the SEPP situation differs from that which prevails in JSPPs where, by definition, risks are shared between the members and the sponsor, or in MEPPs, where they rest wholly on the members.
PBGF “insurance” coverage
SEPPs are obliged to pay annual premiums to the province’s PBGF, which guarantees payment of the promised benefits up to a maximum of $1,000 per month in the event that the plan is wound up with insufficient assets. MEPPs and JSPPs do not pay such premiums and their members are ineligible for coverage. Arguably, plans that partially “insure” their members might be treated less stringently than those that do not. On the other hand, plans that threaten to offload their losses to others in the “insurance” pool should be regulated more stringently than those that bear such losses themselves.
Most MEPP and JSPP members — but only one in three SEPP members — are enrolled in plans that originate in an agreement between their employer and a union or other representative body. If the purpose of funding rules is in part to ensure that the “pension bargain” is honoured, and if a representative organization can pressure or persuade the sponsor to reconfigure, reinforce or reinterpret that bargain so as to better protect the interests of its members, it may be reasonable to place somewhat less reliance on the regulatory regime in such circumstances. By contrast, where members lack effective representation, there is a particular need to subject the plan to rigorous regulatory oversight.
SEPPs are characteristically (but not exclusively) located in the private sector, as are some MEPPs and, potentially, JSPPs. Private sector SEPPs are therefore especially vulnerable to fluctuations in the sponsor’s fortunes and to the risk of the sponsor becoming insolvent. They are also more likely to face restructuring as corporate sponsors reconfigure themselves, merge with other corporations or simply decide to redesign or close their plans. Public sector plans — largely MEPPs and JSPPs — have faced divestments or reorganizations as well, but with somewhat less frequency.
The logic of “solvency funding” is that if a plan confronts immediate wind-up, its assets must be adequate to provide all members — active and retired — with lump sum payments or annuities, which will generate the equivalent of the pension they were promised. As I was frequently reminded by representatives from both SEPPs and MEPPs in the broader public sector, these plans are extremely unlikely to confront immediate wind-up, their sponsors will never become insolvent, and they should not be subject to funding rules that treat them as if they were exposed to private sector-type risks. On the other hand, several briefs to the Commission urged that funding rules should treat public sector plans on the same basis as their private sector counterparts.
I acknowledge, of course, that applying a standard set of funding rules to all plans might, on its face, seem more fair, more consistent with the rule of law and more efficient for the regulator to administer. I acknowledge, as well, that previous attempts to provide different plan types with special funding rules have sometimes produced highly unsatisfactory outcomes. The “too big to fail” regulation of 1992, for example, sheltered several major Ontario plans from the requirements of solvency funding; all but two have since teetered on the brink of failure. However, while mindful of arguments to the contrary, the extended analysis of how factors in plan design might influence funding rules is ultimately persuasive. The “one size does not fit all” principle can and should be applied to the design of funding rules — nor would such an approach introduce an anomaly into Canadian pension law. All Canadian jurisdictions presently provide different funding rules for SEPPs, MEPPs and/or public sector plans, and many have different rules for subcategories of MEPPs and for other special types of plans.
Recommendation 4-8 — MEPPs, JSPPs and SEPPs should have separate funding rules related to their distinctive characteristics. In general, MEPPs and JSPPs should be allowed more flexibility in funding, while SEPPs should be subject to stricter rules than other plans.
4.6 Multi-employer Pension Plans
MEPP representatives argued strongly that they should be subject only to going concern — not solvency — funding. The going concern approach, they contend, is more appropriate to the resilient, long-term character of their plans. Alberta and Ontario have recently excused MEPPs from meeting solvency funding requirements — in the latter case, only for so-called Specified Ontario Multi-employer Pension Plans, or SOMEPPs, which meet strict conditions, and only for the three-year period 2007 to 2010.
The argument for continuing these arrangements in Ontario in some form past 2010 is compelling, though not without its problems. As noted in section 4.5 above:
- MEPP sponsor contributions are usually fixed through collective bargaining rather than actuarially;
- because they are sponsored by a number of enterprises, MEPPs are relatively immune to sponsor failure unless, improbably, many sponsors fail at the same time, or the dominant sponsor fails;
- MEPPs can adjust benefits to accommodate changing economic assumptions and cope with a shortfall in funds, whether the plan is ongoing or being wound up; and
- MEPP active members (but not usually retirees) are influential and often control plan decision-making as a result of a statutory requirement that not less than 50% of a MEPP governing body must represent plan members.
All of these considerations argue for allowing MEPPs to subject themselves to the less rigorous discipline of going concern — rather than solvency — funding, if they wish to do so.
However, the case for exempting MEPPs from solvency funding is not without its difficulties. For one thing, going concern valuations as presently conducted allow for a significant range of discretionary decisions and, to some extent, may lack transparency. For another, they contemplate much longer amortization periods (15 years) than solvency valuations (five years) and thus permit plans to remain under-funded for a longer time than seems prudent. This could be problematic if there is a cyclical downturn in the MEPP’s sector or in financial markets. The SOMEPP regulation addresses this issue sensibly by requiring MEPPs to amortize fund deficiencies over 12 rather than 15 years, and by requiring benefit improvements that push the funded status of a plan below certain levels to be funded over eight years. Thirdly, the absence of a solvency valuation — whatever its defects — deprives trustees and administrators of important perspectives on the financial health of the plan. All of these issues were dealt with by the 2007 SOMEPP regulation.
A second set of concerns derives from the distinguishing feature of MEPPs: they address the issue of “full funding” rather differently than SEPPs. While surpluses for MEPPs may be used to fund contribution holidays or benefit improvements (as with SEPPs), unfunded liabilities in MEPPs are usually dealt with by reducing accrued or future benefits, including pensions already in pay — an option generally unavailable to SEPPs. Of course, if MEPP sponsors and their unions agree, benefit reduction can be avoided by renegotiating the sponsor contributions that are fixed in their current collective agreement; and in MEPPs that provide for member contributions, those contributions may be increased as well. But as a practical matter, these alternatives are often difficult to implement.
Consequently, benefit reduction is a real and present danger for most MEPPs and their members. Who will feel the greatest pain of benefit reduction is obviously a controversial issue, especially if it is not shared on a pro rata basis among retired, active and future plan members. Given that retirees have no voice in plan governance in many MEPPs, there is at least a possibility that their interests will not be protected. Moreover, since MEPPs are not “insured” by the province’s PBGF and do not wish to be, in those relatively rare cases where a MEPP actually does fail, beneficiaries cannot look to the PBGF for compensation. And finally, many MEPP members are apparently unaware that their pension benefits are not defined or fixed but are, instead, target benefits to be achieved, if possible, and reduced, if not. This is a serious shortcoming that could portend internal difficulties in the MEPP community, given that a number of MEPPs are currently funded at well below 100% of liabilities on a going concern basis.
These issues are addressed in greater detail in Chapter Six, which deals with plan failure, and in Chapter Eight, which deals with plan governance. However, to make a general point: if MEPPs are to be given a standing exemption from solvency funding, as they request and I propose below, they must be willing to do two things. First, they must acknowledge that they are accepting greater risks by abandoning solvency funding and ensure that their members are well aware of this fact. Second, they must initiate reforms in their governance arrangements that will ensure greater transparency in risk management, greater accountability by plan administrators, and greater influence by beneficiaries over decisions being made on their behalf in this new, riskier atmosphere.
Several issues of implementation must also be addressed. For example, the SOMEPP regulation provides relief from solvency funding only to MEPPs with more than 15 participating sponsors and, in other respects, attempts to ensure that only truly diversified MEPPs gain the benefit. While the details are negotiable, the basic concept is surely right: the obligation to provide solvency funding should be relaxed only for MEPPs whose design and other institutional features justify such relaxation. In the absence of solvency funding, and given the target nature of the benefits and fluctuations in asset values, it is difficult to determine the value of an active MEPP member’s future pension. All of these problems are capable of being resolved, but they require detailed consultations between MEPP representatives and those responsible for drafting legislation to implement this report, assuming its recommendations are accepted.
In general, the 2007 SOMEPP regulation provides a realistic basis for designing a more complete and permanent exemption of MEPPs from solvency funding. It has the added advantage of providing at least a brief window on how such an exemption might actually work, the difficulties that have to be resolved, and the positive possibilities for the growth of MEPPs that might ensue. The latter, as I explain in Chapter Nine, should be embraced as an important goal of Ontario’s pension policy.
Recommendation 4-9 — Following consultation with Ontario’s multi-employer pension plans, special legislation and regulations should be developed relating to all aspects of their funding, regulation and governance. The basis for such legislation and regulations should be the Specified Ontario Multi-employer Pension Plan regulation of 2007. After five years, the practical effects of these arrangements should be assessed.
Recommendation 4-10 — Multi-employer pension plans should be required to fund only according to going concern valuations, but should continue to provide solvency valuations for the information of the regulator as well as their active and retired members.
4.7 Jointly Sponsored Pension Plans
As the analysis in section 4.5 suggests, JSPPs resemble MEPPs in many respects. Indeed, there is a considerable overlap between the two plan types in that all but one of the five JSPPs formed so far are MEPPs, and many MEPPs are also jointly sponsored and jointly governed. However, for regulatory purposes, the two are distinguished by several differences in the funding rules under which they presently operate. First, whereas JSPPs must be funded jointly by the sponsor and active plan members, MEPPs need not be funded in this fashion and are, indeed, often funded solely through sponsor contributions. Second, whereas MEPPs may reduce accrued benefits to meet funding deficiencies at any time, JSPPs may do so only upon being wound up. Third, whereas MEPPs are presently relieved from solvency funding if they qualify as SOMEPPs; JSPPs are not.
JSPPs, however, wish also to be relieved of solvency funding requirements. To be funded on both a going concern and a solvency basis provides a plan with a margin of safety. Consequently, to be funded only on a going concern basis in the present funding environment implies that the plan is willing and able to assume an additional element of risk. MEPPs are clearly well-positioned in this regard because they have a unique power to deal with a funding shortfall by reducing accrued benefits. The question is whether JSPPs have a comparable capacity to respond to the risk of under-funding. Clearly, they are able to increase contributions to make up funding deficiencies, and indeed have an enhanced ability to do so because they are jointly funded and jointly administered. Experience with JSPPs to date shows that this is more than a hypothetical response. Furthermore, their governance structure also allows them to establish benefits, such as indexation and early retirement on a contingent basis — to provide them if funding is available, and not to do so if the plan’s funded status does not permit.
But most importantly, the acceptance of risk is a foundation principle of JSPPs. If, on wind-up, they are unable to fully meet their obligations, they may reduce accrued benefits, including pensions in pay. Moreover, in such a situation their members have no recourse to the Pension Benefits Guarantee Fund. In these last two crucial respects they precisely resemble MEPPs and are very different from SEPPs.
On balance, then, I conclude that MEPPs and JSPPs share sufficient common characteristics that they ought to be treated alike for funding purposes.
Recommendation 4-11 — Jointly sponsored pension plans should be required to fund only according to going concern valuations on the same basis as Specified Ontario Multi-employer Pension Plans, but should continue to provide solvency valuations for the information of the regulator as well as their active and retired members. The comprehensive legislation and regulations governing the funding of multi-employer pension plans, to be developed pursuant to Recommendation 4-9, should apply, perhaps with appropriate modifications, to jointly sponsored pension plans.
4.8 Public Sector and Broader Public Sector Plans
Pension plans in the public sector are presently subject to both going concern and solvency valuations, just as if they were private sector plans. However, as a practical matter, going concern valuation, rather than solvency valuation, has tended to drive the funding requirements of many of these plans. In general, both public sector sponsors and unions representing their workers would prefer that this situation — which now depends on particular factors that influence individual plan valuations — be made a permanent feature of the funding rules as they apply in this sector. If all public sector plans were relieved of solvency funding, they contend, public sector sponsors and active plan members would generally be able to contribute at lower rates because the volatility of valuations associated with solvency funding would be reduced. This might be an attractive outcome for the plan participants — but on what basis might it be justified?
The ultimate rationale of proposals to relieve public sector plans from solvency funding is that they are unlikely to be wound up “tomorrow” — the central premise of solvency valuation. Moreover, proponents contend, even if these plans were to wind up, whether “tomorrow” or in the more distant future, there is no practical likelihood that their sponsors would be unwilling or unable to make good any funding deficiencies.
While somewhat overstated, this rationale for relieving public sector plans from solvency funding obviously has some grounding in reality. On the other hand, the long-term prospects of public sector pension plans are not necessarily enhanced by allowing a perception to develop that they provide better or less costly pensions than private sector DB plans, or that the government might subject itself to less onerous funding rules than private sector employers, or that the ultimate safeguard for public sector retirees is the virtual certainty that the taxpayers of Ontario will stand behind their pension plan.
More to the point, there are much better arguments in favour of relieving some public sector plans from solvency funding than the fact that they are “public.” Many of them are already MEPPs or JSPPs. Those that are not could (and for other reasons, likely should) join with other SEPPs to become MEPPs. Or if they wish to retain their individual autonomy, public sector plans could become JSPPs or JGTBPPs — a new type of plan described in the next section of this chapter. In any case, whether as MEPPs, JSPPs or JGTBPPs , they will at least be claiming relief from solvency funding on the same grounds as their private sector counterparts. These same approaches, I should note, are open to plans in the broader public sector seeking relief from solvency funding, of which universities may be the leading example.
4.9 Jointly Governed Target Benefit Pension Plans
In Chapter Eight I recommend the introduction of a new type of plan — the jointly governed target benefit pension plan (JGTBPP ) — with the following characteristics: it originates in a collective bargaining relationship; it provides for significant participation in governance by both active members and retirees; and it offers “target” benefits. My reasons for stipulating the first two characteristics can be found in Chapter Eight, and for the third, in my discussion of MEPPs and JSPPs above.
To recapitulate: pensions involve a certain element of risk; if people who are making decisions on their own behalf accept risk knowingly, and if they have at their disposal the means of mitigating risk or dealing with its consequences, they ought to be allowed some latitude to establish the kind of pension plan they prefer. I hasten to add that the acceptance of risk ought to be permitted only within fixed limits; that acceptance of risk by plan members ought to be informed and authentic; and that responses to risk ought to be limited to the spectrum of options — reduction of benefits, whether accrued or going forward, while the plan is ongoing or on wind-up, and contribution increases — that are already employed by either MEPPs and JSPPs or both.
I introduce JGTBPPs in the context of this discussion on funding in order to clarify that these proposed new plans will resemble MEPPs in that they will offer target benefits, and will resemble JSPPs in the sense that they will be jointly governed with enhanced capacity to adjust benefits and contributions. Accordingly, they should be funded in a similar fashion to MEPPs and JSPPs. They should also be subject to the same constraints as MEPPs and JSPPs, including restrictions on promising unfunded benefits and accelerated amortization requirements for plans that fall seriously short of being able to pay for the target benefits they have promised.
Recommendation 4-12 — Jointly governed target benefit pension plans that are based on an agreement between one or more sponsors and one or more unions, that have established explicit arrangements for joint governance, and that permit accrued benefit reduction in an ongoing plan in order to deal with funding deficiencies, should be funded in a manner similar to jointly sponsored pension plans, as provided in Recommendation 4-11.
In Chapter Five I deal with regulatory controls on transactions by which plans convert from DB to defined contribution (DC) plans. However, I regard conversion of broader public sector SEPPs to MEPP or JGTBPP status as being of an evolutionary character, rather than as a sharp break with the past. Accordingly, I favour a simplified conversion process in these situations, driven largely by a concern for transparency and democratic decision-making, but making clear that target benefits would apply only on a going-forward basis. For similar reasons, but without the complication of changing from defined to target benefits, conversion of broader public sector SEPPs to JSPPs should be made relatively quick and easy.
4.10 Single Employer Pension Plans
Though they enrol less than one-third of all DB plan members, SEPPs require more extensive analysis, receive more intensive oversight and are required to meet different funding expectations than other types of plans. Plausible reasons might explain this: few have governance structures that provide internal checks and balances to the sponsor’s unilateral administration of the plan; some are quite small and unable to develop, implement or afford sophisticated strategies for investment, member services and other functions; none can adjust accrued benefits to meet funding shortfalls; and all — even the largest — are closely tied to the fate of the single sponsor that created and funded them.
4.10.2 One valuation or two?
As noted above, SEPPs are currently required to prepare both going concern and solvency valuations and to fund in accordance with whichever requires the highest contributions. Most countries with employment-based private pensions require only a solvency valuation for funding purposes, although a few jurisdictions outside Canada require a second form of valuation for information purposes. Some stakeholders argue that Ontario should adhere to the majority position and require only a single approach to both valuation and funding, while others endorse the use of a second valuation for information — but not funding — purposes.
In the case of SEPPs, however, a strong case can be made for continuing the present arrangement. As the driver of pension costs, solvency valuations and going concern valuations tend to alternate as conditions in financial markets change. During much of the 1990s, when somewhat higher interest rates and strong investment returns were prevalent, solvency valuations were quite low and would have required much lower contribution levels. If going concern valuations had not been in place, requiring enhanced contributions, many plans would have been in much worse financial shape when funding requirements based on solvency valuation became the norm during and after the “perfect storm” of 2000. Thus, the requirement for dual valuations and enhanced funding may explain why DB plans across Canada weathered the “perfect storm” better than their counterparts in other countries.
Moreover, it is not at all clear which of the two funding models should be retained, if either going concern or solvency valuations were to be abandoned.
Solvency funding has been criticized because it requires plans to prepare for scenarios that may never come to pass (imminent wind-up) and to prepare to fund obligations that may never eventuate (especially the cost of annuitization and pensions that are initiated before normal retirement age). Moreover, under solvency valuations, funding is significantly influenced by long-term interest rates, but these rates may be inappropriate, unrealistic or simply unmanageable, thus making plans hostage to events over which they have no control. And finally, solvency valuations may make pension plans appear to corporate decision-makers as more costly and volatile than they need to be.
But solvency valuations have their defenders as well. As noted, the greatest risk faced by SEPPs is indeed that the sponsoring employer will wind up insolvent. Solvency valuations force sponsors to face up to this fact by focusing attention on the payout rates and annuity costs associated with plan wind-ups, and requiring that contributions be adjusted accordingly. Moreover, in Ontario, the wind-up of an insolvent plan represents a threat not only to the interests of active plan members and retirees, but to those of the PBGF to which they will look for “insurance” if the plan fails. By minimizing the risk of such claims, solvency funding protects not only individual plans and beneficiaries but, by extension, all plans covered by the PBGF — and, in that sense, the integrity and reputation of the entire DB system. Finally, every Canadian jurisdiction requires both solvency and going concern valuations, with funding set at the higher of the two. For Ontario to adopt a fundamentally different approach on a matter of such importance would require more compelling reasons than I have been able to muster.
Recommendation 4-13 — Single employer pension plans should continue to fund according to both going concern and solvency valuations.
Once it is accepted that both valuations should be maintained, the analysis must focus on whether there should be substantive changes to either valuation method.
4.10.3 Tweaking the solvency rules: the tension between affordability and benefit security
Earlier in this chapter I identified a number of concerns about certain aspects of solvency valuation that permitted some plan liabilities not to be funded, or to be funded only partially and gradually. Among the matters identified were the smoothing of changing asset values and discount rates; the exclusion from funding of some benefits such as future indexation, certain plant closure and early retirement benefits; and prospective benefit increases. If ensuring benefit security were the only factor to be considered, the current solvency rules ought indeed to be amended to ensure that all potential liabilities, including benefit improvements, are fully identified and funded not just fully, but promptly. My earlier recommendations tend in that direction.
In addition, I was told on several occasions that the key to improving solvency funding is to shorten amortization periods. Suggestions ranged from “as close to immediate funding as possible” to three years instead of the present five. While these suggestions are attractive in principle, and consistent with my own analysis, if rigorously implemented they might well bring in their wake higher costs for sponsors and greater sponsor resistance to benefit improvements; greater volatility in funding and more extreme perturbations on corporate balance sheets; and increased transaction and regulatory costs, occasioned by the need for more frequent valuation filings and regulatory reviews. For these reasons, I have not accepted this approach.
Instead, I am attracted by an alternative approach that would require not just full funding — 100% of what is presently required under a solvency valuation — but an additional “security margin,” or “provision for adverse deviation” (PfAD) as it is known in Quebec, where the idea is currently under development, and in other jurisdictions that have some experience with them, including the Netherlands and Switzerland. A security margin, or PfAD, would serve as a reserve against risks of all kinds. It might be calibrated to respond to the particular kinds of risks to which individual plans and plan types are most vulnerable; it might be modest or substantial in size; and it might be either optional or mandatory. Depending on the particular characteristics of a security margin adopted in Ontario — different PfAD models are under discussion across Canada and especially in Quebec — the additional funding costs to the sponsor might be modest or substantial. Accepting that extra costs would be involved, at least security margins, or PfADs, have the virtue of being easily calculable once a formula is decided, of being containable and predictable, and of possibly obviating the need for other detailed changes in solvency valuation rules.
The volatility in equities markets since 2000 has also heightened concerns over how well pension plans account for risks inherent in their investment strategies and, in particular, whether they are ensuring a match between their assets and liabilities. For example, a plan with a higher proportion of retirees drawing pensions than active members generating contributions might “match” the liabilities attributable to retirees with a greater investment in bond or bond-like instruments. The greater the degree to which fixed liabilities respecting retirees are matched with equities or other volatile investments, the greater the risk to the fund. Conceivably, solvency valuations could require sponsors with a significant asset–liability mismatch to either correct it or to pay extra contributions to the plan by way of an offsetting risk premium. This, indeed, is the intent of the PfAD proposal now under consideration in Quebec. On the other hand, a major drawback to asset–liability matching is that it is very expensive — perhaps prohibitively so for many plans. And it also has the potentially undesirable side effect of discouraging pension funds from investing in equities, to the detriment of capital markets.
Not all submissions concerning solvency funding involved changes that would enhance benefit security at a cost, large or small, to sponsors. On the contrary: a number of sponsor-side submissions insisted that the solvency rules should be made more flexible and that sponsors should be accorded some relief from solvency funding, either on an ongoing basis or under specified conditions. Many employer groups, and some pension professionals, emphasized the importance of flexibility in encouraging employers to persevere with their DB plans through an era of low equity prices and interest rates, and other adverse market conditions. Others suggested that more flexible funding rules might encourage sponsors to make more generous benefit improvements now in the hope and expectation that improving market conditions in the future will pay for them without the sponsor having to make additional contributions.
Several stakeholder submissions to the Commission favoured extending the five-year amortization period provided under current solvency valuation rules to 10 or 15 years in order to relieve sponsors from the funding pressures associated with the decline in long-term interest rates. The federal jurisdiction and Quebec have adopted this approach by providing temporary solvency relief to plans that comply with certain criteria through extensions of their amortization periods. And Ireland and the United Kingdom have moved toward greater funding flexibility, using somewhat different strategies.
All of these sponsor-side proposals designed to ease the rigours of solvency funding coalesced around the idea that if a concern for benefit security was allowed to overshadow a realistic appreciation of the need to maintain the affordability of DB plans, sponsors might react by closing or capping their plans. The result, I was warned, would be just the opposite of what was intended: a further weakening of the DB system. This is indeed an important reminder that my recommendations must strike a balance between the concerns of active members and retirees and those of sponsors, between benefit security and affordability, and between the present and future health of the DB system.
It strikes me that the fairest and most straightforward way to achieve that balance is to establish a security margin, or PfAD, earmarked for the explicit purpose of enhancing benefit security, but to do so at modest net incremental cost to sponsors. This would have the additional advantage of acknowledging the point made at the beginning of this chapter: because of normal short-term fluctuations in asset values and interest rates, and marginal changes in other factors, no plan is likely to be funded at precisely 100% at any given moment.
The formula used to determine the security margin should be as simple as possible in order to avoid generating significant compliance costs.
Recommendation 4-14 — Single employer pension plans should be required to maintain a security margin (or provision for adverse deviation) of 5% of solvency liabilities. This margin should be amortized over an eight-year period. The security margin should be deemed to be part of the plan surplus on wind-up, but not for other purposes.
In recognition of the enhanced security provided to plan members by this security margin, plans that are en route to achieving it should enjoy a longer amortization period.
Recommendation 4-15 — For plans that have achieved 95% of solvency funding, the normal amortization period for achieving the new required funding level, inclusive of the security margin, should be extended from five to eight years. For plans funded below 95%, the current amortization period of five years should continue to apply until such time as they become eligible for the extended amortization period.
4.10.4 The distribution of surplus on plan wind-up
Employers, active members and retirees have been engaged in conflicts over surplus use and distribution since at least the mid-1980s. The issue has sometimes been framed in theoretical terms to no conclusive effect, as I noted in section 4.2 above. It has been framed as a clash of legal regimes — the employment contract and plan documents embodying the intent of the sponsor, versus the law of trusts and the PBA, which protect the property rights asserted by plan members and retirees — with the latter securing a number of significant but not definitive victories. And it has been framed in political terms, most recently in 2002 when the government of the day was forced by strong labour and retiree opposition to withdraw legislation intended to resolve it. My sense — apparently shared and acted upon by most stakeholders — is that none of these approaches is likely to succeed, and that a pragmatic approach to surplus issues is the right one.
By way of context, the PBA provides that upon plan wind-up, the surplus should be dealt with in accordance with the terms of the plan documents. If those terms are not clear, the PBA creates a legislative presumption that the surplus should be distributed to the plan members. Subsequent regulations enacted under the PBA apparently permitted surplus to be addressed by a surplus-sharing agreement entered into between the sponsor and the union representing plan members, or, in the absence of a union, two-thirds of active members and a situation-specific proportion of retirees. However, as a result of the Tecsyn decision (2000), employers have had to establish their clear entitlement to surplus both under the plan documents and under a surplus-sharing agreement. This has meant, in effect, that sponsors have had to secure a court determination of their entitlement under the plan documents, often with the explicit consent of both active and retired members, in order to persuade the Superintendent to permit the distribution of surplus in accordance with the surplus-sharing agreement. This process seems unnecessarily cumbersome, time-consuming and expensive. It should be changed.
Three broad alternative approaches are possible. First, existing plans could be required to amend their documents to resolve the issue by clear language, as new plans are already required to do. However, this approach is likely to lead to disputes if members object to the sponsor’s proposed amendment. Second, a default rule could be laid down in the legislation overriding plan documents and assigning surplus to either the sponsor or the plan beneficiaries, or dividing it between them in fixed proportions and/or by priority. This approach would likely be controversial. Or third, legislation might mandate the distribution of surplus in accordance with a clear plan document or under a surplus-sharing agreement, subject to recourse to some dispute resolution process in the event of disagreement. Ontario stakeholders have become familiar with this third approach; they appear to favour it and it seems to work well, or at least it would if a more efficient dispute resolution process were available.
Recommendation 4-16 — If a single employer pension plan is in surplus on being wound up, the surplus should be distributed in accordance with the plan documents unless the parties agree, or the proposed Pension Tribunal of Ontario rules, that the documents are not clear. In the event of such an acknowledgement or ruling, the sponsor may propose a scheme for the distribution of surplus, which would take effect if approved in one of two ways:
- if plan members are not represented by a union, the proposal should be submitted to a vote by secret ballot of the plan members and retirees, and would take effect if approved by two-thirds of those voting; or
- if plan members are represented by a union or other organization, the sponsor should submit its proposal to representatives of the active members and retirees with a view to concluding a surplus distribution agreement.
If the sponsor and the representative negotiators cannot reach agreement, they should submit the matter for determination to a dispute resolution procedure of their own choosing. If they cannot agree on such a procedure, or if it does not resolve the matter within a reasonable time, any party may apply to the Superintendent to refer the matter to the Pension Tribunal of Ontario, which would then establish the terms of the surplus distribution agreement.
Any scheme approved by secret ballot, any surplus distribution agreement reached by representative negotiators, and any determination by the Tribunal or an agreed dispute resolution procedure would be final and binding on the Superintendent and on all persons claiming to be entitled.
4.10.5 The distribution of surplus in an ongoing plan
Ongoing plans may find themselves in a surplus position for many reasons. In accordance with Recommendation 4-14, which contemplates the establishment of a security margin in every plan, the first 5% of any excess should be set aside for that purpose; any excess over that margin should be considered surplus.
Like surplus in plans being wound up, surplus in ongoing plans has been the subject of frequent controversy. Plan members will naturally wish to see it retained in the plan to enhance or to index plan benefits, or to provide a further buffer — over and above the 5% security margin — in the event that the plan suffers reverses in the future. Sponsors may on occasion acquiesce in the views of active members and retirees, but often have other ambitions for the surplus. Sometimes they will want to withdraw the funds in order to use them for general business purposes or to pay plan expenses; more commonly, they will want to take a contribution holiday in order to treat the surplus as a credit against contributions they would otherwise be obliged to make.
Contribution holidays, which are lawful in the United States, the United Kingdom, and all Canadian jurisdictions, including Ontario, are clearly legitimate in appropriate circumstances so long as they do not compromise the solvency of the pension fund. Nonetheless, they are controversial. Anecdotal evidence provided to the Commission suggests that some sponsors took contribution holidays on the basis that their last triennial valuation permitted this, even though their funded status had since deteriorated to well under 100%. A research study prepared for the Commission by Jinyan Li notes that a limited review of federally regulated plans identified no “observable” relationship between funded status and contribution holidays, and that 45% of under-funded plans would not have been under-funded had they not taken contribution holidays.
This somewhat limited evidence suggests that contribution holidays ought at least to be made more transparent. Recommendations to this effect are found in this chapter, in Chapter Five and elsewhere. In addition, measures ought to be introduced to prevent the initiation or continuation of contribution holidays based on a previous triennial valuation when plans have, in the meantime, become seriously under-funded. Of course, these measures ought to take account of the possibility of rapid changes in a plan’s funded status and of the difficulty of estimating that status without subjecting the plan to a full valuation. They ought to be designed so as to trigger sensible action to preserve the fund and to discourage deliberate or wilfully negligent disregard for its security.
Controversy over the sponsor’s access to surplus in an ongoing plan extends to its use to pay plan expenses. In principle, I see no reason why the payment of plan expenses, including PBGF premiums, should not be permitted under the same circumstances and subject to the same conditions as contribution holidays.
Recommendation 4-17 — Plan sponsors should be entitled to reduce or omit their contributions to a plan in any year in which it is funded at 105% or more of its solvency liabilities. However if — based on benchmarks to be developed by the regulator — a plan administrator knows, or ought reasonably to know, that funding has fallen below 95%, the administrator should immediately notify the sponsor to resume contributions until the plan is again funded at 105% of solvency liabilities. The pension regulator should develop benchmarks based on the plan’s annual financial statements that will enable plan administrators to determine when contributions should be resumed.
If the regulator finds that a contribution holiday was improperly taken or continued, any contributions withheld from the plan should become immediately due and payable, together with interest, regardless of the plan’s present funded status, and the sponsor should be subject to an administrative fine of up to $1 million, or double the amount withheld during the improper contribution holiday, whichever is less. The improper use of plan surplus to pay the expenses of the plan, including PBGF premiums, should be treated in similar fashion.
The parties to a collective agreement should be free to negotiate other arrangements for the use of surplus in an ongoing plan. These arrangements should prevail notwithstanding those proposed in this recommendation or established in the plan documents.
Debate over the use of surplus in an ongoing plan has also surfaced in the context of proposals to use plan surplus to fund the merger, division or conversion of existing plans, or to subsidize the creation of a DC or hybrid plan. These issues are addressed in Chapter Five.
Finally, the sponsor’s right to simply withdraw surplus from an ongoing plan is an issue about which feelings run particularly high. However, it is a rare occurrence because, under current law, withdrawal requires the consent of every plan participant plus the maintenance in the plan of a buffer in excess of full funding of two years of current service costs, or 25% of plan liabilities, whichever is the greater. In general, I am reluctant to encourage the withdrawal of surplus from ongoing plans that may experience future difficulties from time to time. However, while I accept that the required buffer is an appropriate deterrent to reckless behaviour, I am also concerned that the present constraints on withdrawal are so severe that they may produce perverse results. On the one hand, sponsors may be reluctant to fund plans beyond legally permissible minimum levels if, as a practical matter, they are precluded from recovering any part of the surplus. On the other, the virtual inaccessibility of surplus may prompt sponsors to accomplish the same result by other means: to press for valuations that minimize their contributions, to take inappropriate contribution holidays or to overstate expenses to be paid out of the plan. I therefore propose some modification of the present rules on surplus withdrawal.
Recommendation 4-18 — Sponsors may apply to withdraw surplus from an ongoing plan pursuant to the procedures set out in Recommendation 4-16, provided that the plan remains funded subsequent to withdrawal at 125% of full solvency funding, or 105% of full solvency funding plus two years of current service costs, whichever is greater.
This recommendation involves two departures from the status quo. First, it lowers the required level of consent by beneficiaries (absent a union) from 100% to two-thirds of those voting. This requirement ensures that while a significant minority of members can still block a withdrawal, a few individuals cannot. Of course, if a union is present in the workplace, any decision concerning surplus withdrawal must be taken with due regard for its right to negotiate on behalf of all bargaining unit members.
Second, and more importantly, the recommendation establishes that the sponsor may withdraw surplus from an ongoing plan on the basis of clear plan documents. This possibility not only rests on the principle that clearly articulated rights ought to be respected, it is accompanied by a number of safeguards to protect plan members: the high threshold for withdrawal; the right of the union to use its bargaining power to negotiate superseding arrangements well in advance; the possibility of challenging the sponsor’s claim that the plan documents are clear; and the improved negotiation and dispute resolution process provided in recommendation 4-16.
To my mind, the possibility of properly controlled surplus withdrawal provides sponsors with a carrot — a small carrot — to complement the substantial stick of enhanced funding requirements and tighter controls on contribution holidays. I believe that it will give sponsors some confidence that they can fund well over the minimum required by law to the general benefit of all plan members, and that it will prejudice only the very few plan members who might otherwise be able to extract some surplus from the sponsor as the price of agreeing to withdrawal of surplus in the rare circumstances that a plan finds itself funded at above 125%.
4.11 Annuity Rates
A number of stakeholders focused on annuity rates as a problematic aspect of solvency valuations. A key element driving solvency valuations is the price that plans might have to pay to purchase annuities for their active members and retirees in the event of a wind-up. When long-term interest rates are low, as they are now, assumed annuity costs used in solvency valuations are high. Moreover, they are kept high by the fact that under current regulations, annuities may be purchased only from Canadian insurance companies. Since relatively few companies are available to provide them, the supply of annuities is limited and the cost of annuities is higher than it might otherwise be under fully competitive market conditions. The situation would become even more difficult if a number of large pension plans were actually required to purchase them at the same time. The net result, I was advised, is that upward pressures on solvency valuations attributable to annuity costs make it cheaper to simply keep plans open and in operation than to annuitize benefits and close them. This situation may serve neither sponsors nor beneficiaries over the long term.
Several suggestions were advanced for avoiding or reducing the cost of annuities. These included creating a publicly funded non-profit annuity pool for pensions; removing the requirement of annuitization on wind-up and allowing for lump sums to be paid and invested as the active or retired member wishes; and allowing a private or public agency to handle stranded pensions, thus obviating the need for them to be annuitized and, presumably, reducing both the demand for annuities and their price. Such changes, it is argued, would make pension funding less volatile and more affordable by detaching funding valuation from annuity rates. In Chapter Five I address these concerns by recommending the creation of an agency to handle stranded pensions.
Recommendation 4-19 — Ontario should investigate strategies for reducing the cost of annuities and the influence of the annuities market.
As is well known, inflation, even at modest rates, can erode the real value of pensions. In order to protect retirees from this harm, some plans provide that pensions will be adjusted according to a formula, usually agreed in the course of collective bargaining; others provide that the sponsor in its discretion may adjust the pensions from time to time; and still others are simply silent on the issue.
Representatives of active members and retirees have argued that sponsors should not have access to surplus unless adequate provision has first been made for indexation. They note that surplus arises, in part, due to growth in the nominal value of plan assets, some portion of which may be attributable to inflation, which, ironically, at the same time undermines the purchasing power of pensions. On this basis, as well as on other grounds, they claim to be entitled to some or all of a surplus in order to fund improvements in the plan — including indexation.
This issue of indexation was carefully canvassed by the Friedland Task Force, appointed in 1986 following periods of runaway inflation in the mid-1970s and early 1980s. The Friedland Task Force recommended a formula by which some proportion of the increase in asset values would be used to provide indexation — and in fact, legislation to this effect was enacted, though never proclaimed in force or implemented by regulation.
Employers do not accept this analysis. They see surplus as arising from investment growth, which may be attributable to a superior investment strategy, and from higher-than-needed sponsor contributions — especially in recent times, when significant payments have been required to meet solvency rules. They argue that, at a minimum, surplus attributable to either of these causes ought not to be appropriated to provide inflation protection, and that at a maximum, if inflation protection is neither expressly provided in the pension bargain nor funded on an ongoing basis, it ought not to be imposed ex post facto.
This latter argument may have legal merit, but like the other arguments against indexation, it does not address an important social problem. There can be no doubt that even modest inflation over the long term generates difficulties for all persons on fixed incomes, including retirees. And there can be no doubt that for people retiring today, the long term is likely to be longer than it used to be. However, providing this enhanced income certainty for active and retiring plan members would translate into significantly increased costs and uncertainty for plan sponsors. Sensible retirement planning and pension design should provide for indexation, even if it means reducing the initial value of pensions in order to maintain their purchasing power into the future. But should sensible planning and design be required by law — either prospectively or retrospectively? Given the cost and risk implications, and their potential destabilizing effects on the whole DB pension system, this is not a step to be taken lightly. Nor is the problem confined to the DB system. Recipients of DC pensions and individuals who save for retirement in other ways also confront a likely decline in their standard of living over the long term due to inflation. So do DB members who happen to be enrolled in plans that are not in surplus.
A minimal or initial response to the problem of indexation is to make it more visible and urgent by forcing sponsors and unions — if not individual plan members — to face up to the long-term consequences of not providing it. This can be accomplished in part by providing active members and retirees with timely and accurate information.
Recommendation 4-20 — Every plan should contain a clause stating explicitly what provision, if any, has been made for the indexation of benefits and for the funding of indexation. Each triennial valuation and each annual statement provided to the regulator, active plan members and retirees should provide the same information.
This approach has several advantages. It alerts beneficiaries to an important positive or negative feature of the pension bargain; it potentially brings pressure to bear on their union, if they have one (some three-quarters of all DB members do), in order to address the indexation issue through collective bargaining; it forces the negotiating parties to confront the hard choice between higher current pensions and pensions whose value will endure over the long term; and it is more consistent with the voluntary character of Ontario’s pension system than the other solutions proposed. Indeed, some plans have begun to address indexation by various contingent benefit formulae. For example, indexation may depend on fund performance and/or the plan’s ability to pay — innovative approaches that should receive further study.
At best, however, this reliance on disclosure and experimentation would provide gradual movement toward some sort of solution for most plans and beneficiaries. Unfortunately, it has several serious defects. First, it would not help plan members who have no union representation. Second, unions and their members may be reluctant to sacrifice current gains for future indexation, especially if the benefits of indexation accrue to retirees who no longer have a voice in union deliberations and cannot participate in collective action to help the union achieve its bargaining goals. And third, it does not deal with situations of unusually high and rapid inflation, such as the one that led to the appointment of the Friedland Task Force. This last defect, at least, can and should be overcome. Ontario should not — as it did in the 1970s and 1980s — find itself in the position of being unable to deal with the effects of an “inflation emergency” on the pension system.
Recommendation 4-21 — The government should proclaim in force the provisions of the Pension Benefits Act that allow it to require that pensions be inflation-adjusted in accordance with a formula to be prescribed. That formula should be restricted to “inflation emergencies.”
Indexation is a classic example of how difficult it is to address long-term issues of fundamental importance to Ontario’s pension system in the context of debates over more immediate, practical and controversial policy issues. Discussion of indexation is essential, and I hope that the Pension Champion, whose establishment I recommend in Chapter Ten, will ensure that this issue engages the attention of stakeholders and the government.
4.13 Letters of Credit and Asset Pledges
Many sponsor representatives proposed that using fully secure alternative forms of assets could facilitate the funding of plans under certain circumstances. One such asset is the letter of credit. If a sponsor is credit-worthy, it could obtain a letter of credit from a bank — in effect, a promise to pay the fund an agreed sum if the sponsor defaults on a scheduled contribution payment. The letter of credit could be made “callable”: if the sponsor is unable to renew the letter of credit because it cannot pay the bank charges or because its credit rating has deteriorated, the bank becomes legally obliged to pay the full value of the letter of credit into the fund. The federal jurisdiction, Quebec, Alberta and British Columbia now allow letters of credit on different terms up to different maximum amounts and for different periods of time. The United Kingdom imposes few specific limitations on letters of credit but requires that the plan trustees be satisfied with the security provided. During the relatively brief period since their inception, letters of credit do not appear to have generated difficulties in any of the jurisdictions that have adopted them.
However, several objections have been expressed to the use of letters of credit: that they will not generate investment growth for the pension fund; that the existence of a letter of credit in favour of the pension fund might prejudice a union’s position in negotiations resulting from a sponsor’s insolvency; that they may be used to defer much-needed payments into the fund; and that they would be unnecessary if surplus rules were tightened up.
While these objections are understandable, given the relative novelty of letters of credit in the pension context, I do not ultimately find them persuasive. On the contrary, I am convinced that letters of credit can be drafted in such a way as to allay any concerns. They strike me as potentially useful tools that would allow corporate sponsors to retain their capital for business purposes for finite periods of time while fully protecting the financial interests of the fund. Indeed, since banks will provide a letter of credit only to a credit-worthy company, their willingness to provide one following a credit assessment of their client — the sponsor — constitutes a strong signal that the sponsor is, in fact, solvent. Conversely, if the bank decides to call the letter of credit, this would amount to a warning to the pension regulator that both the sponsor and the plan may be in difficulty.
Recommendation 4-22 — Irrevocable letters of credit should be permitted as security for a fixed proportion of contributions owing to a plan, and for a maximum period of time, provided they are enforceable by the plan and immune from inclusion in the sponsor’s estate in the event of insolvency. The Superintendent should have no power to relieve against these requirements either before or after the fact.
After five years, experience with letters of credit should be reviewed by the regulator. If no difficulties are found, they should be made available as a permanent feature of pension funding in Ontario.
The United Kingdom also permits the use of asset pledges as security for debts owed to pension funds under specified circumstances. Pledgeable assets might include land, bonds or other easily marketable goods. To be acceptable for this purpose, assets would have to be independently, objectively and conservatively valued and be made fully available to the pension fund in the event the sponsor defaults on payment.
Recommendation 4-23 — Ontario ought to investigate the possibility of permitting the use of asset pledges to provide security for unpaid contributions to pension funds, and to define the purposes for which, and the conditions under which, such pledges might be used. If asset pledges seem useful for sponsors, safe for pension plans and capable of being overseen by the regulator, their use ought to be allowed for an initial period of five years, subject to renewal on a permanent basis if experience warrants.
4.14 The Influence of Federal Law on the Funding of Pensions
While my mandate requires me to report to the Minister of Finance for Ontario, and while it clearly contemplates that I will confine myself to matters within provincial jurisdiction, I heard from a number of stakeholders that improvements in funding practices and in the financial viability of occupational pensions in Ontario depend in some measure on changes to federal law. In the analysis and recommendations that follow, I therefore propose that the Honourable Minister make representations to his federal counterpart, seeking such changes. I have good reason to believe that these changes would be widely welcomed by members of the pension community across the country.
4.14.2 The Income Tax Act
The federal Income Tax Act (ITA) and the rules and regulations made under it have significant implications for pension plans. These rules limit the maximum benefits that can be paid out of, and the maximum contributions that can be paid into, a pension plan in order for it to qualify for favourable treatment under the ITA. I was told on numerous occasions that both limits are too low.
The tax-incented benefit levels in Canada are a fraction of benefit levels in other countries, making DB plans less attractive to employees with above-average levels of income, and therefore, perhaps, less attractive to corporations generally. The contribution limits are 110% of liabilities, in most circumstances. It is widely accepted that these limits have had a negative effect on funding over time, as they prevent a reasonable amount of surplus from accumulating in good economic times to carry the plan through more difficult periods. In addition, in 1991, the ITA was amended to remove the preferential tax treatment previously afforded to DB pension plans. Some observers suggest that, given the superior qualities and public policy benefits of DBs, tax policy should once again encourage DB plans and those with similar characteristics.
While such proposals inevitably involve tax expenditures, these are to an extent offset by the reduced claims made by DB retirees on government income support programs and the greater expenditures they make (and the multiplier effects they generate) when they spend their pensions on goods and services. These are good public policy reasons to encourage DB and like plans.
Recommendation 4-24 — The Ontario government should endeavour to persuade the federal government to increase benefit and contribution levels for registered pension plans under the Income Tax Act, and to consider policies that encourage participation by workers and employers in DB plans or their functional equivalents.
4.14.3 The impact of the Pension Benefits Standards Act investment rules on pension plan investment strategies
The investment strategy of a pension plan is a critical aspect of its funding. Valuations, contributions and ultimately, benefit security, are all significantly affected by a plan’s investment profile and performance. Moreover, as outlined in Chapter Two, the cumulative effect of pension plan investments constitutes a powerful influence on Ontario‘s capital markets and economy.
The investments of registered pension plans in almost all provinces are governed by investment rules enacted under the federal Pension Benefits Standards Act (PBSA) and then incorporated by reference into provincial pension regulations. These rules prohibit pension plans from directly or indirectly:
- holding more than 30% of the voting shares of a corporation;
- holding more than 5% of the book value of a single Canadian (but not foreign) real estate company, or 15% of a group of Canadian (but not foreign) real estate properties;
- holding more than 25% of the book value of Canadian (but not foreign) resource companies;
- holding more than 10% of the shares of any particular company or group of companies, except through a segregated, mutual or pooled fund that complies with certain requirements;
- lending money to a related party, including the plan sponsor; and
- holding shares of a related party, including the plan sponsor, unless obtained on a public stock exchange or as needed for the operation of the plan.
By contrast, and layered on top of these specific requirements, the Ontario’s PBA sets a general standard for investment decisions: a plan administrator is required to “exercise care, skill and diligence” and to demonstrate “the prudence of an ordinary person holding the property of another” in the investment of the pension fund.
Both to the Commission and in other forums, a number of major Ontario pension plans have raised significant concerns about the federal investment rules. In their view — and in that of other observers, as well — some of the federal rules, such as those that seem to favour foreign over domestic investments, are misconceived. Some rules are inconsistent with the desire of major plans to become active rather than passive investors; and some — more by accident than design — seem to inhibit attempts to generate optimal returns for the benefit of pension plans and their members. In addition, the federal rules are seen to diminish the potential contribution of these important institutional investors to capital markets. Nor are major plans alone disadvantaged. If, as I recommend in Chapter Nine, small pension funds were provided with the means to pool their resources to achieve better investment outcomes at lower costs, the federal rules would affect them adversely as well. Finally, it is widely believed that plans find ways to do what the investment rules prevent. If so, not only are the rules subverted, but so too is general respect for the law. These are all good reasons to change the rules, but the federal government has so far declined to do so.
Would it suffice for the federal government to simply repeal its investment rules, or for Ontario to declare that it is no longer bound by them? I think not. Ontario’s “prudent person” rule, as it stands, is very vague. More explicit guidance is needed: the meaning of “prudence” should not be left so broad that it has no normative power, nor should it have to be determined after the fact in expensive litigation. On the contrary, clear rules ought to lay down at least the main principles governing investment decisions. For example, plans should be prevented from “putting all their eggs in one basket” or, as in the federal rules, from investing in the sponsoring corporation (unless through a public exchange and within the diversification limits), or from engaging in other forms of self-dealing. These general rules might then be refined — perhaps through information bulletins or interpretative rulings — so that they speak to potential problem areas such as employer-administered SEPPs or union-administered MEPPs, which may be tempted to invest their funds with a view to advancing their collective bargaining or job creation strategies, rather than maximizing the pension plan’s investment returns.
The aim should be to enable funds to pursue prudent and profitable investment strategies commensurate with their resources and capacities; to prevent investment practices that might place pension plans at unacceptable risk; to avoid plans having to engage in convoluted or unseemly manoeuvres in order to undertake normal, responsible marketplace behaviour; and to accomplish all this with a minimum of direct, bureaucratic control.
Ideally, there should be Canada-wide consensus on a new set of investment rules, which suggests that the federal government ought to take the lead in formulating them. In the absence of such action, Ontario can and should adopt its own investment rules, which might continue to incorporate by reference, or to simply reproduce, some of the federal rules, omitting only those to which it objects. Or, as suggested above, the province might build on its “prudent person” rule. Finally, a reconstructed pension regulator, as proposed in Chapter Seven, might provide greater guidance to plan administrators on what it regards as inappropriate investment behaviour, both in the form of information bulletins and by way of advance rulings to plans embarking on novel investment strategies.
Pension plans must also take steps to adapt to a future in which they may be given greater latitude than they presently enjoy. Those steps must begin with a serious assessment of the implications of any transition from passive to active investment practices, and they must involve greater capacity for fund management and risk assessment. And of course, there must be transparency in the adoption of new policies and in accountability to plan beneficiaries for their success or failure. I return to this issue in Chapter Eight, where I deal with the so-called 30% rule that limits the ability of pension plans to become actively involved in the management of enterprises in which they have a significant or dominant ownership position.
Recommendation 4-25 — The Ontario government should endeavour to persuade the federal government to reform the federal investment rules and, in particular, to remove or amend particular quantitative restrictions that no longer make sense, such as those involving prohibitions on Canadian, but not foreign, investments.
However, if the federal government does not do so within a reasonable time frame, the Ontario government should cease to rely on the federal regulations and establish its own investment rules, tracking the federal rules only to the extent that doing so is deemed good public policy in Ontario.
Finally, interest in socially responsible investment (SRI) has been evolving over the last decades to the point where many pension plans have investment policies that consider SRI issues. A number of European states now require that corporate boards and bodies like pension trustees disclose their SRI policies, if any. It seems to me these are not only positive, but likely inevitable, directions for investment practice, and should be acknowledged. This issue is addressed in greater detail in Chapter Eight.