Target practice

In recent years, organizations have been reviewing the affordability and sustainability of their retirement arrangements in the context of low interest rates, market volatility, accounting reforms and aging demographics. Plan sponsors are searching for new ways to help manage the risks associated with their plans and contain rising costs.

The Appeal of TBPs
Target benefit plans (TBPs)—which effectively transfer much of the risk associated with traditional DB plans from the employer to the members while retaining many of the DB plan’s attractive features—could be part of the solution. Under this type of arrangement, benefits are communicated as a targeted amount but are not guaranteed.

One of the main reasons these plans are attractive to employers is that company contributions are typically either fixed or permitted to vary only within a narrow range. But there are other reasons why an organization might be interested in adopting a TBP.

For instance, because they offer many of the same characteristics as DB plans, TBPs can effectively support workforce management strategies, whether in unionized or non-unionized settings. In particular, they can be designed to include early retirement incentives tailored to meet the needs of a company’s workforce.

In addition, TBPs help spread the longevity and investment risks more broadly than DC plans do. Effectively delegating much or all of an employer’s fiduciary responsibilities to a board comprising plan member and employer representatives also has appeal for some employers.

Barriers to Adoption
However, there are a number of barriers to address before this model becomes a viable option for most plan sponsors. Indeed, there are significant challenges for any jurisdiction that is contemplating a single-employer TBP regime.

Although many provinces have indicated that they intend to adopt target benefit legislation, most have not yet introduced broad regulations to date.

And many of these jurisdictions have stated that, at least initially, the option will be made available only to multi-employer plans or to plans covering collectively bargained employee groups.

For a target benefit model to have widespread appeal to plan sponsors—particularly those in the private sector—it is essential that a past-service conversion mechanism, plus an exemption from solvency funding requirements, is provided for legacy defined benefits. Yet, at this point, there has been no strong indication that such conversions or exemptions will be permitted for single-employer plans outside of New Brunswick. In addition, for non-unionized workers, there needs to be a process to ensure adequate representation of their interests, both in establishing a TBP and in its ongoing operations.

Towers Watson’s 2013 Pension Risk Survey asked plan sponsors, if their jurisdiction allowed a DB plan to convert to a TBP, would they accept joint governance? (see chart, below). Only 34% said yes. And the yes responses were actually split, with 52% of public sector plan respondents in agreement, compared with 29% of private sector plan sponsors. (Private sector companies tend to shy away from joint governance, even though it results in their fiduciary duties being effectively delegated to an independent administrator.)

Clearly, many plan sponsors are on the fence about TBPs—likely because, with no final regulatory framework, there are just too many unknown variables and potential barriers right now.

The New Brunswick Solution
The shared risk model introduced by New Brunswick in May 2012 provides an innovative example of how the target benefit framework can be leveraged to address pension issues more broadly.

New Brunswick legislation permits existing DB plans to be converted to the shared risk model. Such conversions typically result in substantive changes to the underlying benefit promise. Specifically, future earnings-linked upgrades and post-retirement indexing become conditional upon affordability, and accrued benefits become subject to the same risks as future accruals. These changes can significantly reduce a plan’s funding liabilities on conversion. The reason is that the previously automatic inflation-protection provisions—which can be expensive—no longer have to be reflected until the conditional benefit increases are actually granted.

To ensure that these plans target a high degree of benefit security, robust risk management objectives are needed. Specifically, the regulations require that plans meet the following minimum risk management goals upon conversion:

• at least a 97.5% probability that base benefits will not have to be reduced in any year over the next 20 years (since 97.5% represents 39 times out of 40, this means that bad news regarding benefit reductions is anticipated in just one year out of every 40); and
• conditional indexing (inflation-protection) provisions that are expected to provide—on average, over the next 20 years—at least 75% of the value of the prior plan’s indexing provisions, as well as 75% of annual increases in the consumer price index (CPI) in respect of accrued benefits that had previously been based on a final average earnings formula.

These risk management objectives, which must be tested using stochastic asset/liability modelling (showing the full likely range of potential financial results over future decades), effectively result in building significant margins into the underlying contribution rules of the shared risk plan. Therefore, while this model can significantly bring down volatility in funding requirements and possibly reduce volatility in the annual accounting cost, it will not likely reduce the plan’s long-term cost. Also, stochastic projections involve complex mathematical modelling and can, therefore, be expensive to create, so these projections effectively increase the administrative complexity of operating the plan going forward.

Variations and Alternatives
While the New Brunswick approach is groundbreaking in many ways and has already been adopted by a number of plans in both the public and private sectors (such as the province’s plan for civil servants and the University of New Brunswick), there are some elements that other jurisdictions may want to consider adapting if they establish their own target benefit framework. In particular, alternative regulatory models that allow for greater flexibility by being less prescriptive in their risk management procedures—or that reduce administrative complexity—could further increase the appeal for plan sponsors.

For instance, another regime could allow the various stakeholders to negotiate either the degree of confidence that accrued benefits will not be reduced or the target level for future conditional indexing, rather than having these thresholds prescribed by regulation.

Alternatively, the underlying governance framework and funding mechanism could be modified more substantially by removing the mandated risk management goals altogether. For example, a regime could allow required contribution levels to be established based on a “best estimate” going-concern valuation, with an explicit PfAD (provision for adverse deviations)—sometimes known as a “going-concern-plus” regime.

Minimum PfADs and surplus spending thresholds could be prescribed based on the risk profile of the underlying plan, but, ideally, the plan’s stakeholders would be permitted to negotiate alternative thresholds within a formal funding policy. In addition, rules would also need to be established for what happens when a subsequent valuation reveals that the plan is in a deficit position (i.e., if the plan’s liabilities plus PfAD exceed the market value of its assets). Again, any regulatory requirements should not be unduly prescriptive but should instead allow the parties to define the order of any priority actions, including reducing benefits or increasing contributions, as part of the agreed-upon funding policy.

Of course, the New Brunswick model is effectively based on a fully integrated funding, benefit, investment and risk management framework. Modifying one aspect could have unintended consequences, such as reducing benefit security for members, increasing volatility in employer contributions or eliminating the potential opportunity for DC accounting treatment. Before advocating for a particular model, plan sponsors should assess the broader implications of such variations on their plan.

Alternatively, other sponsors have expressed a preference to maintain the guaranteed nature of a DB plan within a more flexible benefit and funding policy framework. For example, some situations, such as a large group annuity purchase, could warrant replacing CPI-based indexing (which is harder to hedge against) with a fixed rate indexing provision for which risk-reducing hedges are easier to set up. However, pension legislation generally prevents any changes that affect benefits earned in respect of past service.

In other cases, a sponsor may want to seek an exemption from solvency funding requirements where a more robust going-concern funding mechanism is adopted. Yet strict funding rules generally prevent regulators from approving alternative funding mechanisms, even if all relevant stakeholders have reached an agreement.

Communication is Critical
Under the New Brunswick rules, there is no requirement to obtain member consent or even consult with members in order to convert an existing plan. In addition, given the potential for accrued benefits to be reduced at some point, leading to possible lawsuits, legal immunity is granted under the shared risk plan legislation. Specifically, a plan administrator will not be liable under the New Brunswick legislation or regulations (provided the party has acted prudently). Also, all plan stakeholders are protected from legal action being instituted in relation to a breach of contract, trust, or any legal duty or obligation with respect to permissible changes in the nature of benefits upon conversion. Sponsors across Canada viewed these elements positively when they were announced, as they promote additional flexibility in plan design.

However, recent experiences illustrate the importance of proactive communication with all stakeholders. For example, a group of retired civil servants has formed the Pension Coalition NB, which has vigorously resisted any potential changes to their pensions.

Many New Brunswick sponsors that have converted their plans to date have indicated that full disclosure and open, transparent communication with all other plan stakeholders—including current pensioners—were key factors in their success. These attestations serve to remind plan sponsors that the adoption of a complex model that affects underlying benefit promises requires a thorough communication strategy from the outset. Ultimately, in order to improve pension coverage, policy-makers will need to provide new options that better support employer objectives and offer greater flexibility for plan sponsors to negotiate benefit and funding policies with their stakeholders. The emergence of new designs, such as TBPs and shared risk plans, is certainly encouraging.

The issue that remains is how to move TBPs along faster in the political agenda. An open discussion of the differences between relying on stochastic projections versus using going-concern-plus in the regulated funding and benefit adjustment rules would enhance the understanding of the critical issues involved.

Ian Markham is Canadian retirement innovation leader, and Dylan Moser is a senior consulting actuary based in Atlantic Canada with Towers Watson. ian.markham@towerswatson.

Get a PDF of this article.