Target benefit plans - Game-changer or non-starter?
In our last Vision, we started a comparison of Defined Contribution (DC) plans versus Target Benefit Plans (TBPs).
This Vision reviews TBPs in greater detail. Whether one regards them from the perspective of employers or employees, our conclusion is that TBPs could prove to be the most promising development in Canadian pension plan design in decades. They should be of particular interest to public sector entities and larger private sector companies.
Governments will play a major role in determining whether TBPs are successful. The regulations that are still to be released in most jurisdictions will be of critical importance. Make TBPs too difficult to implement and few plan sponsors will be interested. This Vision describes both the opportunities and the pitfalls.
WHAT IS A TBP?
The TBP concept is not new. It has existed for many years in the form of negotiated contribution Multi-Employer Pension Plans (MEPPs), which are especially popular in certain industries. What is new is the notion of TBPs in the singleemployer environment.
In its most basic form, a TBP is a pension plan that aims to provide a defined benefit (DB), but with fixed contributions. To plan members, it is virtually indistinguishable from a regular DB plan except that accrued benefits are subject to reduction if the funding level falls below a given threshold. To avoid a reduction, TBPs are governed by more formal funding and benefit policies than one typically finds in defined benefit plans. In addition, conservative assumptions can be used in the setting of the target benefit with benefit improvements granted only if there is a significant funding surplus.
TBP assets are pooled for investment purposes and are managed in much the same way as in a DB plan although the TBP fund might be invested a little more conservatively to reduce the chances of a funding deficit.
One variation on the basic TBP allows for contributions to vary within a narrow range rather than being completely fixed. This is the idea behind the Shared Risk Pension Plan that was adopted recently by the New Brunswick Government (see the sidebar “The New Brunswick SRPP”).
REASONS TO CONSIDER TBPs
TBPs are promising because they eliminate the risk of rising costs inherent in DB plans while offering a better solution for most employees than most DC plans.
DB plans have long fallen out of favour in the private sector and are now under increasing attack in the public sector. Much has been written to explain why this is the case but for our purposes it suffices to say that the decline of DB plans is unlikely to be reversed. In the long run, the only DB plans that may survive are those sponsored by organizations with very deep pockets.
For their part, DC plans suffer from two fundamental defects. The first is the uncertainty of the benefits they generate, which complicates retirement planning (see Figure 1 for example). The second is that the members bear the risk, but all too often are not qualified to make their own investment decisions.
Figure 1: A STABLE INCOME STREAM IS HARD TO MAINTAIN
TBPs minimize many of these shortcomings. From the employer’s perspective, the following characteristics of a TBP are significant inducements to switch away from DB:
- The contribution level is fixed, with noobligation to contribute more even if funding proves inadequate.
- Accounting (in the ideal case) is based on contributions made, the same as in DC plans. Accounting in DB plans is an elaborate exercise that has produced some unpleasant surprises for employers in recent years. Pension expense has skyrocketed as bond yields have fallen and the situation has been exacerbated by investment losses during the financial crisis. This is the main reason DB plans became so unpalatable to the shareholders of private sector companies. The DC-type accounting that the accounting profession might decide to apply to TBP plans eliminates any such pension expense shocks.
- Pension Adjustment (PA) calculations under a TBP should be easy, being simply the contribution made (assuming the Canada Revenue Agency (CRA) agrees with this treatment). This simplifies plan administration and may also provide more Registered Retirement Savings Plan (RRSP) contribution room for employees.
- The target benefit in TBPs is monitored by means of going-concern funding valuations. Solvency valuations, which have been the source of such volatile funding in recent years as to threaten the very solvency of some companies, would not be required.
The sponsors of DC plans may also want to consider TBPs. The one big advantage they have over DC plans is that the assets are pooled and invested in much the same way as for DB plans. This means the TBP sponsor does not have to offer individual investment choice thus eliminating the onerous and sometimes futile task of trying to educate an entire workforce on investment basics.
While employees are unlikely to prefer TBPs over traditional DB plans, they will probably find them a better alternative than DC plans. There are some good reasons for employees to prefer TBPs:
- While the target benefit is less certain than in a DB plan, it is more certain than the income one can derive from a DC plan.
- The target benefit is paid for life so retirees do not have to worry about outliving their assets. They do have to worry about a potential benefit cut under a TBP but good funding and benefit policies reduce both the chances and the severity of such a cut.
- Various studies indicate that plan administrators, with the advice of investment professionals, make better investment decisions than individual employees. TBPs therefore promise to stretch a dollar of contribution further than it would go in a DC plan.
- For a DC retiree to secure a stable monthly income, she would have to buy a life annuity. The insurance company expenses and anti-selection charges, as well as the very conservative investments underlying the annuity reserves tend to reduce the amount of payout. TBPs avoid all of those problems.
- Unions should appreciate the fact that TBPs promote solidarity. Unlike DC plans, all members in a TBP accrue the same pension for a given year of service.
PROBLEMS WITH TBPs
The biggest drawback to a TBP is that the payout can be reduced if the target benefit is less than 100% funded. This problem is more than theoretical as an estimated 25% of MEPPs have had to reduce benefits in the past decade.
As we will see, benefit reductions should be less likely in single-employer TBPs with the help of intelligent plan design features and the adoption of conservative investment, benefit and funding policies. The downside to conservatism, however, is lower target benefits and a greater transfer of wealth to the next generation, so some compromise is necessary. Hence, the prospect of benefit reductions in a TBP can never be entirely eliminated.
To appreciate the challenge of deriving a stable income from fixed contributions, consider Figure 1,1 which is taken from our last Vision. It shows that historically, a fixed rate of contributions would have generated retirement income that varies from a low of 15% of final average pay to as much as 55% depending on the year of retirement.
For the plan in Figure 1, the plan sponsor could have set the target benefit at 30% of final average pay and could probably have paid out the full basic pension consistently though there would have been some years when post-retirement indexing could not have been paid. Had the target been set at 40%, it certainly looks more attractive but it would mean forgoing indexing for prolonged periods and even basic benefits would have to be reduced in some years. Neither situation is perfect. The 30% target provides better security but in good times it may frustrate some retirees who feel they could have done better in a DC plan and who have no wish to be building up a reserve for the benefit of the next generation of plan members. On the other hand, a 40% target would lead to frequent disappointment.
Another challenge with TBPs is that employers may have to forgo plan provisions that incent certain behaviour, such as retiring earlier. Incentives come at a cost and members in a TBP will resent subsidizing others. Of course, subsidies exist in virtually every DB plan as well, but they do not tend to create problems there because the extra cost is perceived to be borne by the employer rather than coming at the expense of fellow plan members.
WHO WILL WANT A TBP?
Clearly, TBPs are not perfect vehicles but given the even greater problems with DB or DC plans, there will be many situations in which TBPs should be given serious consideration.
The most likely candidates for conversion to TBPs will be organizations that wish to, or need to, close their existing DB plans because of sustainability issues, but do not want to abandon their employees to the vagaries of DC plans. Some DC plan sponsors may also convert to TBPs. It is not easy to educate members on their DC investment options. We expect the best DC candidates will be larger organizations which had converted to DC fairly recently and still maintain a closed DB plan.2
In those situations, a DC culture might not have taken hold and the members might welcome a plan that resembles the DB plan they had just lost. Moreover, the plan administrator’s recent connection with DB history makes it easier to revert back to DB since most of the “machinery” is still in place or easily accessible.
The biggest stumbling block for DC plan sponsors to switch over to TBPs is simply inertia. If members are not complaining, then why bother? Employers who sponsor a DC plan have already stabilized their pension funding costs and the threat of litigation due to employees making poor investment choices may seem remote. While a TBP may seem like the better solution in theory, it might take considerable effort to convince employees this is the case and there will be at least a few “losers” to contend with—employees who would genuinely be better off under a DC plan.
Smaller employers, whether they currently maintain DB or DC plans, are also less likely to adopt a TBP. They might be daunted by the tasks of establishing formal funding and benefit policies and then maintaining a governance regime to administer those policies. If stochastic modelling is involved—a requirement in some jurisdictions—it is one additional complexity that smaller employers can do without.
THE NEW BRUNSWICK SRPP
The government of New Brunswick had a major problem. Its public sector plans had suffered large investment losses due to the financial crisis. Long term interest rates were falling which made pensions more expensive. And because of longer lifespans and earlier retirement, the province estimated that new retirees were receiving 11 more years of pension payments than retirees in the 1970s.
As in other provinces, the remedy until recently was to keep on increasing member and employer contributions. This was no longer sustainable, however, as taxpayers were not in a mood to pay higher taxes to bail out public sector plans. In addition, a large part of the rising deficits was attributable to retiree liabilities and since they were no longer contributing, it fell to active members to make up the difference which fostered resentment within the active population. A better long-term fix was needed.
The solution involved setting up a task force that established the ground rules for a more sustainable pension system. Key principles in the process were transparency and equity. The new “pension deal” needed to be articulated clearly and had to ensure that all stakeholders were treated equitably.
Public consultations were held. As actuary to the province, Morneau Shepell worked closely with the taskforce in devising and stress testing a new plan. This involved numerous valuations on many different plan designs and dozens of Monte Carlo simulations, in addition to many meetings to analyse results and refine the solution. The process also involved an extraordinary amount of co-operation and goodwill amongst all stakeholders including the government and the unions. Since this plan was breaking new ground, the plan design process was more extensive than usual.
The end result was the creation of the Shared Risk Pension Plan (SRPP). The SRPP is similar in many respects to the regular TBP as described elsewhere in this paper but with some special provisions. Rather than being entirely fixed, employer and employee contributions can both fluctuate in a narrow range, between 8% and 10% of pay depending on the funded status of the plan. The funded status is determined using a special going-concern valuation that is known as a “funding policy valuation.” It takes into account not only today’s assets but also the extra contributions that are expected to be made over the 15 years following the valuation. Under the SRPP, two types of benefits are defined:
- Base benefit which is the basic target pension benefit, and
- Indexing of the base benefit to the CPI both pre- and post-retirement.
The indexing of benefits depends on the funded status of the plan. The funding policy established that only one fifth of the funding surplus can be used to provide indexation; this translates into a maximum indexing of 1% per annum if the plan’s funded ratio is 110%, and 2% if the ratio is 115%.
The plan is governed strictly in accordance with its special funding policy and investment policy. For example, a cut in benefits is triggered when the funded ratio is below 100% for two consecutive years. If a benefit reduction is triggered, the reduction will be made across the board, meaning it will apply equally to active members and retirees. Risk is managed using an annual monitoring process that requires an actuarial valuation, asset-liability modelling, and a review of both the investment and funding policy.
The range of contributions is calibrated to provide a very high level of security for basic benefits and a reasonably high level of security for ancillary benefits. The different levels of security are quantified using a Monte Carlo simulation.
Virtually all subsidies within the plan have been eliminated. The target pension is based on a career average earnings formula with accruals indexed (if funding is sufficient) to the change in CPI from the year of accrual until retirement. The target pension is payable unreduced at age 65 and is otherwise subject to a 5% reduction for each year that pension starts earlier than 65.
Because of the strong focus on the security of benefits, the target asset mix is more conservative than in the typical DB pension plan. About 55% of the assets are earmarked for fixed-income investments. As a result, the discount rate used in the funding policy valuation is a little lower than what is used in the going-concern valuation in most DB plans, but higher than the discount rate for a typical solvency valuation.
For the government, the SRPP was deemed a great success since it brought long-term cost stability. The plan members benefited from the creation of a more sustainable plan that still provides a very good level of retirement security. Moreover, the members have the assurance that the employer cannot take contribution holidays unless required under the Income Tax Act. All surpluses generated by the plan will be solely for the benefit of its members. Intergenerational equity has been greatly enhanced and the pooling of longevity and investment risk makes this a better solution than a DC plan.
IDEAL CONDITIONS FOR TBPs
While many plan sponsors will favour the TBP concept in principle, the evolving regulatory environment can strongly influence whether TBPs will flourish. In approximate order of decreasing importance, Table 1 shows the provisions we would hope to see incorporated in TBP legislation.
The jury is still out on whether the conditions for a successful launch of TBPs will be in place. The first condition, retrospective application, is especially in doubt. In our opinion, New Brunswick is a good model but other jurisdictions have not indicated yet whether they will follow New Brunswick’s lead.
IMPLICATIONS FOR PLAN DESIGN
While a TBP may look much like a traditional DB plan, the fixed contributions will have a subtle but important impact on plan design.
In particular, fixed contribution rates make for a compelling argument to eliminate subsidies within the plan. Subsidies are common in traditional DB plans of course but in a TBP, with its finite contributions, it becomes painfully clear that providing more value to one member leaves less for the remaining members.
Consider early retirement incentives, for example. It is not unusual for DB pension plans to allow employees to retire earlier than age 65 without having to reduce their pension. A 60 & 30 condition, for example, is fairly common.3 Starting a fixed amount of pension at age 60 is worth about 40% more than starting the same pension at 65. If the employee and employer are both contributing 6% of earnings to fund the target benefit, for example, the member who retires at age 60 receives an employer benefit worth about 8% of pay versus just 4% for the member who retires at age 65.
For the same reason, a TBP that provides an earnings-related benefit is more likely to base it on actual earnings in each year of the employee’s career (indexed perhaps) rather than on final average earnings. Otherwise, a fast-tracker, who probably fared much better financially over her career in any event, will be paying a smaller share of her ultimate pension than someone whose raises were more modest (see the sidebar, “The Fast-Tracker Problem”). Once again, this happens in a traditional final-earnings DB plan, but there is the sense that the extra value given to the fast-tracker is being funded by the employer, not fellow employees.
An indexing provision, albeit on a contingent basis, is almost a must in a TBP. “Indexing” has two meanings in a TBP. During the accumulation phase, it means increasing accrued pensions to keep pace with an outside index such as the Consumer Price Index (CPI) or Average Industrial Wage (AWI) indices. This is how the CPP works. During the retirement phase, indexing involves increasing pensions in payment based on part or all of the change in CPI since the last such increase. If a TBP is not sufficiently well funded to provide indexing in a given year, the plan can probably still pay the full basic benefit. Contingent indexing therefore provides another ayer of protection for the basic pension. Forgoing indexation is not painless, of course, but it is not as painful as cutting basic pensions. If all MEPPs had a contingent indexing provision, fewer of them would have had to reduce benefits.
Designing a TBP is also an excellent time to reconsider the appropriate retirement income target. As we have noted in past issues of Vision, it may not be necessary to set the retirement income target (from all sources) as high as 70% of final average pay to replace 100% of regular pre-retirement consumption. For many years now, the main preoccupation of DB plan sponsors has been dealing with rising pension costs so plan design in general, and the basic benefit formula in particular, have been secondary considerations. That may change with TBPs.
THE FAST-TRACKER PROBLEM
Fay is a fast-tracker. Thanks to a number of promotions, her pay has been rising by 6% a year, every year during her career. Sally’s pay rises more slowly, at 3% a year. Even though they start at the same pay, by retirement Fay’s pay is nearly 3 times as much. They both contribute 6% of pay each year. In a career-average earnings plan where each year’s accrual is indexed to inflation, both Fay and Sally end up paying about 50% of the cost of their pension (actually 52% and 48%). In a final average earnings plan that has the same overall cost, Fay pays only 40% of the cost of her pension while Sally pays 60%. It will be hard to justify a TBP where the higher values given to all the fast-tracking Fays are at the expense of the Sallys with their slower-rising pay.
Table 1 (see pdf)
IDEAL CONDITIONS FOR TBPs
- Retrospective application should be permitted – Employers in DB plans need to be allowed to adopt TBPs retrospectively, which means that both the past and future service benefits of active members should be capable of being modified to conform to TBP rules. The alternative, in which past service entitlements are preserved in DB form, makes TBPs less attractive to any plan sponsor. Not only would two-part arrangements be more difficult to administer and communicate, the TBP advantage of fixed contributions would take many years to be realized.
LIKELY GOVERNMENT RESPONSE- So far, only New Brunswick allows TBP rules to be applied retrospectively. Proposed legislation in other jurisdictions has been ominously silent on the matter. Unless this is rectified, it may prove to be a key deterrent to the ultimate success of TBPs.
- An unimpeded right to reduce benefits – The plan administrator needs to be able to reduce benefits if a deficit arises. This seems to be obvious but under existing MEPP rules, this is not permitted currently in certain jurisdictions, such as Quebec.
- LIKELY GOVERNMENT RESPONSE - It is expected that all the major jurisdictions, including Quebec, will recognize the need to reduce benefits in underfunded TBPs.
- Easy accounting – TBPs should be accounted for the same as for DC plans, meaning that the employer’s pension expense would be equal to the contribution made. To ensure DC-type accounting, plan sponsors might have to avoid certain plan design features, such as guaranteeing a minimum return of the member’s contributions with interest.
- LIKELY GOVERNMENT RESPONSE - At this stage, it is possible that some jurisdictions may require a minimum refund of a member’s contributions with interest on the termination of employment in which case we need to understand how it affects accounting treatment.
- Minimize the requirement to maintain funding surpluses – Some administrators will want to maintain a high reserve in order to avoid having to reduce benefits. Others will prefer lower reserves to promote inter-generational equity.
- LIKELY GOVERNMENT RESPONSE - It is expected that most jurisdictions will impose funding requirements skewed toward building a reserve but not necessarily an excessive reserve.
- Employers should be able to implement a TBP in a non-union environment. It is counter-productive to allow TBPs only in collective bargaining situations. Such a requirement would force more non-union employees into pure DC plans. Representation by members and retirees on the governance body for the TBP should be sufficient to protect members’ interests.
- LIKELY GOVERNMENT RESPONSE - Ontario introduced legislation in 2011 that would allow TBPs only in collectively bargained situations. Other jurisdictions are expected to allow TBPs without union representation.
- All funding and entitlements under TBPs should be based on going-concern assumptions – One of the attractions of a TBP is the elimination of solvency valuations and the volatile funding they create. The same actuarial basis that is used to fund a TBP should also be the basis for the lump sum entitlements of members who leave before retirement.
- LIKELY GOVERNMENT RESPONSE - It is expected that all jurisdictions will stipulate that TBPs be funded on a going-concern basis. In some cases, a solvency valuation may be required for disclosure purposes but not to determine funding requirements.
Except in New Brunswick, TBPs cannot be implemented yet. In some provinces, such as Quebec, legislation still has to be drafted. In other provinces, such as Ontario, Nova Scotia, Alberta, and British Columbia, legislation has been drafted but not enacted into law yet and it will still be some time before regulations are released.
Nevertheless, any plan sponsors contemplating TBPs can start conducting a feasibility study now. Such a study would help them assess whether a TBP is appropriate in their situation and if so, what regulatory and other conditions are conducive or a successful implementation. Even before a TBP is implemented, DB sponsors may wish to modify the various subsidies that are built into their existing DB plans. This will not only pave the way for the eventual implementation of a TBP, it may lead to a fairer and more sustainable DB plan even if a TBP is never adopted.
Another reason for organizations to start a dialogue with their actuary now is that it will help them to understand if a TBP is right for them. Moreover, there may still be time to influence government policy, which can make a major difference in the successful launch of TBPs.
By the glacial standards of Canadian pension reform, the introduction of Target Benefit Plans on a national scale can now be considered imminent. We endorse TBPs in spite of their limitations. In particular, TBPs are unlikely to give rise to new pension plans in workplaces that lack one now. If the goal is to augment overall pension coverage in Canada, TBPs will not help much, if at all. Pension coverage will improve significantly only if (a) PRPPs or some other form of occupational pensions are made mandatory or (b) C/QPP is enhanced which is increasingly likely.
Within the sphere of employers who provide pension coverage now, TBPs will be an important new plan design that may ultimately replace DB plans in both the private and public sectors.
Saskatchewan — The Act already permits TBPs in collectively bargained situations.
Ontario — Securing Pension Benefits Now and in the Future Act receives royal assent. It includes provisions regarding TBPs in collectively bargained situations.
Nova Scotia — The new Pension Benefits Act, which includes TBP provisions, receives royal assent but is not proclaimed into force.
British Columbia — The Pension Benefits Standards Act that makes provision for TBPs receives royal assent but has not been proclaimed into force.
New Brunswick — An act to Amend the Pension Benefits Act comes into force. It allows for SRPPs.
Alberta — The Employment Pensions Act that makes provision for TBPs receives royal assent but has not been proclaimed into force.
Quebec — An act to provide for the establishment of TBPs for certain pulp and paper companies comes into force.
Alberta — Finance publishes a pension paper with proposed rules for collectively bargained MEPPs that wish to convert to TBPs.
Ontario — Ontario’s budget indicates plans to consult to establish single-employer TBPs.
Quebec — Draft regulations are published that would allow TBPs in certain pulp and paper companies.
1 Figure 1 is based on combined employer/employee contributions of 8% of pay made over 30-year periods starting with 1938-1967 and ending with 1983-2012. The asset mix is a constant 60-40 throughout each 30-year period. In estimating the DC income, it is assumed that the DC account balance would be used to buy an indexed annuity at the interest and mortality rates that prevailed at the time of retirement.
2 A “closed plan” is one that accepts no new members. Existing members might either have a frozen DB benefit or may be grandfathered in the legacy DB provisions for the
rest of their working lives.
3 Age 60 and having 30 years of service.
For the complete article download: Vision Nov. 2013 (PDF)