The idea that pension plan asset allocation should be tied to funded status is surprisingly young. It was not common practice before 2008. In April 2009, when we wrote a paper called “Liability-Responsive Asset Allocation,” it was (as far as we are aware) the first time that this approach was formally described in any detail. Just five years later, it’s taken as a given. Nobody is surprised when the financial statements of major corporations talk of “a broad global pension de-risking strategy” or note that a pension plan’s “interest rate hedge is dynamically increased as funded status improves.”1
Two developments together acted as the catalyst for this change in the industry. The first was that more and more pension plans began freezing new benefit accruals. For a frozen plan, investment returns become less important once full funding is achieved. If there’s already enough money in the plan and no new benefits are being accrued, then there’s not much to be done with any additional returns that might be earned. The obvious route of turning them over to the plan sponsor comes with a hefty tax bill attached.
The second development—less obvious, but equally important—has been the emergence of technology that makes it possible to monitor a plan’s funded status between the regular actuarial valuations. Even five years ago, estimating funded status on a daily basis seemed futuristic. Early versions of liability-responsive asset allocation (LRAA) tended to be based on monthly estimates. But today daily updates of funded status are standard practice.
There are some variations in how the industry applies the idea. Our hope that “LRAA” would become the standard terminology has not been realized; about a half-dozen different expressions are commonly used. But the basic idea of connecting asset allocation to funded status took hold quickly. In an industry that has traditionally been slow to adopt new ideas, this one found ready acceptance almost as soon as it was launched.
Why Companies Choose to Glide
Today many pension plans—especially frozen plans—use glide paths to define how their asset allocation should change as their funded status moves over time. If a plan sponsor maintained a static asset allocation, that decision would suggest that its goals for the plan’s assets will stay the same regardless of the plan’s funding position. But a plan sponsor’s goals are likely to change as the environment changes. The sponsor of a poorly funded plan may think of their goals in terms of the return needed to get to full funding, while the sponsor of a well-funded plan may think in terms of maintaining a steady funded status. This shift in focus does not happen at a single point in time, but rather happens gradually if funded status improves. That’s why organizations are adopting glide paths.
This is illustrated in Figure 1. For a plan in a strong funding position, there’s not enough potential benefit to justify taking the risk inherent in a return-oriented investment strategy. So there’s no reason to choose the upper right quadrant in our diagram. The bottom left, meanwhile, represents an underfunded position with a cautious investment strategy. The question a plan sponsor in that situation must ask is: How will the shortfall be made up? Pension plans have only two ways of raising money: contributions from the plan sponsor and returns earned on the existing assets. If the money is coming from plan sponsor contributions, it’s generally better to make those contributions sooner rather than later, as Pension Benefit Guaranty Corp. (PBGC) variable-rate premiums continue to rise. In fact, those premiums are quickly reaching a level where a plan sponsor would need to have an extraordinarily high cost of capital in order for delaying contributions to make sense2. So the bottom left quadrant of Figure 1 isn’t an attractive choice either.
Now that we’ve ruled out two of the quadrants in Figure 1, two choices remain. One is the top left quadrant, in which a plan faces a funding shortfall and so adopts a return-oriented strategy; this is where many plans are today. But where they want to be, and where they are headed, is the bottom right quadrant, with strong funding and low risk. That’s the targeted end game. Carefully crafted glide paths can set pension plans on the path to reach that goal.
1. These statements are taken from the most recent 10-Ks of Ford and United Technologies, respectively.
2. “Do PBGC premiums incent plan sponsors to borrow to fund their pension plans?” (Gannon, 2013, Russell Investments Practice Note).
LRAA Best Practices
When a pension plan’s funded status changes, the glide path can either lead to automatic action, or it can merely trigger a review of the plan’s asset allocation. The latter option has two drawbacks. First, it can lead to delays, leaving a plan with an inappropriate allocation for weeks or even months while the pension committee considers the authorization of action. Second, it leaves the door open to behavioral biases such as the chasing of returns. For example, an LRAA schedule may specify a change in asset allocation following either a run-up in the value of the stock market or an increase in interest rates. But the tendency of many committees in the face of such events is to want to let them run—why pull out of a strategy that is winning? Such a discussion confuses strategic with tactical considerations, as well as proving in many cases to be mistaken even from the tactical perspective.
For that reason, we regard it as a best practice to make the glide path the policy, rather than a guideline. Bear in mind that one of the reasons the LRAA approach has found such wide acceptance is that it meets, head-on, the common view that 2007 was a missed opportunity for many pension plans. Some plan sponsors feel that if they find themselves in as good a position again, they want to lock it in. The temptation to let the good times run left many plans with what turned out—in hindsight—to be too much risk.
Once a plan sponsor decides to adopt a glide path, an endpoint must be defined, establishing the funded status at which the glide path will stop and what the asset allocation should look like at that time. Figure 2, below, shows a sample de-risking glide path in which the plan reaches its lowest-risk allocation at a funded status of 110 percent with 20 percent of assets in return-seeking investments. These numbers may vary based on the specific circumstances of the pension plan and the corporation, but they should be decided in advance (albeit subject to periodic review).
Another basic decision in glide path design lies in the question of whether asset allocation shifts are a one-way street. All LRAA schedules involve de-risking when the plan’s funded status improves, but plan sponsors must decide whether their policy should also require increasing investment risk when funded status declines. In theory, a case can be made for either approach. In practice, the one-way street is most popular. That’s largely because a one-way street avoids moving away from the desired end goal; many plan sponsors find it acceptable to continue to take a risk that is already in place but would be less comfortable with adding that same risk back in once it has been removed.
Glide paths also vary in their complexity. The most common approach is to base allocation on funded status alone, but glide paths can incorporate other variables as well. Some have trigger points based on interest rates, in addition to funded status. Others have a calendar component. The motivations for such variations might be tactical (as in the case of interest rates) or strategic (as in the case of calendars). In either case, the plan sponsor ought to clearly articulate the rationale for the approach.
An LRAA policy should specify with more granularity the breakdown of assets within the categories “return-seeking” and “liability-hedging.” All glide paths specify the balance between return-seeking assets, such as equities, and liability-driven investing (LDI)/hedging assets, such as long-duration corporate bonds, depending on the plan’s funded status level. But planning needs to go further.
As a plan moves down its glide path and the size of the return-seeking portfolio shrinks, diversification of these assets may become harder to achieve, so the make-up of the portfolio may need to vary. Asset illiquidity may become more of an issue over time. Similarly, within the LDI portfolio, there may be little value in fine-tuning the hedge to the liabilities when the plan is at the return-seeking end of the glide path and allocation to LDI is a small portion of the total portfolio. At this early stage, the most important goal of the LDI investments is to respond to interest rate changes—i.e., to increase the duration of the portfolio as interest rates rise. But when the hedging portfolio accounts for a larger proportion of the plan’s total assets, the plan sponsor should consider more closely tying the hedge to the liabilities, focusing on credit spreads and key rate durations (KRDs), for example.
This evolution of the return-seeking and LDI/hedging portfolios may be formally built into the LRAA schedule (easing future implementation) or may be left implicit, as decisions for the plan sponsor to take on when the change to the overall balance of assets occurs (allowing fine-tuning of the strategy). Whatever approach an organization takes to tackling these second-order decisions, they should not hold up the adoption of a glide path strategy.
There are many other ways in which de-risking practices vary, from the use of tactical rebalancing ranges to the use of different measures of funded status, such as the present value of future benefits (PVFB) rather than projected benefit obligation (PBO) for a plan that is closed but not frozen. Clear best practices have emerged in certain areas, while other areas continue to evolve.
However, what is clear across the board is that more and more defined benefit pension plans are adopting the de-risking model. The overall exposure of this investor group is on a path—formally documented in their investment policies—that will see them move out of riskier assets such as equities and into assets that match their liabilities, such as long duration bonds. The exact timing of the move will depend on both interest rates and the stock market, but the direction has been set.
Bob Collie is chief research strategist in Russell Investments’ Americas institutional business. Collie is responsible for the strategic advice delivered to the various parts of Russell’s institutional client base, as well as for working with the manager research team, product groups, and other research efforts across Russell. He is also the lead author of the Fiduciary Matters blog.
James Gannon is director, asset allocation and risk management, in Russell’s Americas institutional business. Gannon oversees the examination of pension plans from an actuarial perspective, and he helps advisory and investment outsourcing clients reach an effective asset allocation decision by bringing together actuarial liabilities and available asset classes in a risk-reward framework.