Liability-driven investing: balancing risk and return in an uncertain world

By the end of the year, however, that funded status was down to 84% as the sharp decline in the stock market in the fourth quarter cut deep. Pension plan assets in the United States declined from $1.48 trillion at the end of 2017 to an estimated $1.33 trillion at the end of 2018.¹

Uncertainty continues to haunt markets. The U.S. Federal Reserve’s (Fed’s) U-turn on its expectations for further interest-rate increases this year compounded fears of slowing U.S. growth. In the wake of the Fed’s March announcement, the difference between the 3-month and 10-year key measure of the U.S. yield curve—watched by the Fed and seen as an indicator of a coming recession—inverted for the first time in over a decade.

At the same time, the difference between 2- and 10-year yields has dipped below 10 basis points for the first time this year. This is the primary indicator that investors watch because it's inverted before every recession since the Second World War.

The seesaw in markets and funded status since the tail end of last year has likely caused many plan sponsors to review their overall pension management strategy. The volatility in the fourth quarter of 2018—especially in December, which was one of the worst months for equities since the Great Recession—demonstrated how quickly conditions change.

This perhaps explains why U.S. pension plan sponsors have increasingly been paying attention to liability-driven investing (LDI)—an investment framework that uses a pension plan’s liabilities to replace public market indexes as the benchmark against which it measures investment performance.

Our aim with this paper is to share our views on how implementing an LDI framework could help plan sponsors of all sizes better manage funded status volatility and explain why we believe inaction could be costly. 

The underlying rationale for LDI is relatively straightforward: Financial markets move to their own rhythm, independent of a plan’s obligations. It’s therefore difficult for a plan that benchmarks its portfolio against a public market index to expect to match its specific liabilities in a consistent and reliable manner. Selecting an appropriate benchmark is critical for sponsors because the economic risk profile of a market index can be quite different from a pension plan’s liability risk profile. With this in mind, the LDI framework is built around the concept of liability matching.

When implementing LDI, a plan would typically split its portfolio into two parts:

• Hedging portfolio—A hedging portion that matches a plan’s liabilities, typically comprising fixed-income instruments as well as other income-generating assets
• Growth portfolio—A return-seeking or growth portion that could help narrow any gap in a plan’s funded status

Typically, a plan begins the de-risking process by determining the portion of liabilities that needs to be hedged, thereby establishing the hedge ratio, with the remaining assets invested in a diversified portfolio of return-generating growth assets. A dynamic asset allocation framework gives sponsors the flexibility to adjust their allocation between the hedging portfolio and the growth portfolio over time.

To help them decide precisely when they should adjust their allocation between the two portfolios, many plans have chosen to adopt a de-risking glide path. In its simplest form, the de-risking glide path maps out when a plan should increase allocation to its hedging portfolio as different conditions are met on its path toward fullyfunded status.

The three most common conditions, or triggers, used in a glide path are:

1   Funded ratio—As the funded ratio of a plan improves, sponsors are able to protect the gains they've made by reducing their exposure to growth assets and hedging more of their liabilities.

2   Interest-rate levels—Some plan sponsors are reluctant to invest in long-dated bonds or hedge a large portion of their liabilities in the current low rate environment. Instead, they set rate thresholds and increase their hedge levels once those rate levels have been met.

3   Time—Inaction could leave a plan exposed to even greater risk, so setting a timeframe can be a useful way of keeping expectations realistic. Plan sponsors may hope rates will go up or the funded ratio will improve, thanks to the plan’s asset mix, but if things don’t go quite as well as anticipated, at some point, it's best to move forward and begin the de-risking process.

A key advantage of a dynamic asset allocation framework is that it allows plans to lock in the gains extracted from the growth portfolio by transferring them to the hedging portfolio as they're realized. An additional potential benefit is that as sponsors make additional contributions to the plan, its funded ratio improves. This means any additional contributions will generally be invested less aggressively than they might be without a glide path in place.

Under a traditional LDI approach, which simply rebalances systematically to a fixed 60/40 asset mix, it’s left to the market to determine the amount of liabilities a plan wants to protect, which, generally speaking, isn’t ideal. Under a dynamic asset allocation framework, a plan starts by determining the proportion of the liabilities that it wants to hedge (the hedge ratio of a plan’s liabilities). This determines the right amount to invest in bonds. The remainder of the assets are then invested in a diversified portfolio of growth assets. In our view, this allows sponsors to better manage and reduce risks because they're able to control the hedge ratio of their liabilities a well as the exposure to growth assets.

Depending on the parameters agreed on when implementing the glide path, pension plans can also control the speed of de-risking and the sources of risk they're willing to accept during the interim.

By way of example, let’s take a pension plan with an initial funded ratio of 90% that's allocated 40% of its assets to its hedging portfolio and allocated the rest equally into two asset classes—U.S. stocks and international equities. Let’s also assume that the plan’s sponsor has adopted the dynamic asset allocation strategy and decided to increase the plan's liability hedge ratio by 2% for every 1% increase in its funded ratio.

According to our calculation, the plan would experience much lower funding ratio volatility if it had adopted the dynamic asset allocation framework. Over the course of 10 years, the difference between the highest and lowest points of the plan’s funding ratio would be around 15%, as opposed to more than 25% if it had gone with the traditional allocation strategy. This is because the dynamic asset allocation approach allows the plan to manage the level at which its liabilities are hedged; the plan’s managers won’t get a say under the traditional approach—the market decides.

Our study shows that, over time, the plan’s hedge ratio should progress in a relatively smooth manner toward its targeted funding level under the dynamic asset allocation framework. By contrast, the plan’s hedge ratio would have simply moved in line with the market if it had adopted the traditional approach. The impact of this additional funded status volatility is increased volatility of contributions.

Global central banks have begun unwinding monetary policies that were introduced in the aftermath of the 2008 global financial crisis, and interest rates in developed markets have begun to nudge higher. This could translate into additional challenges for plan sponsors because, even though global yields remain low from a historical context, financial conditions have begun to tighten, and rising interest rates could hurt returns. This has led to concerns that plans could be forced to raise the risk profiles of their portfolios in the hope that higher-yielding assets can help meet their obligations.

One way of mitigating this risk is for sponsors to consider allocating a portion of their hedging portfolio to a diversified set of asset classes that are complementary in nature to their fixed-income portfolio. Assets that we believe could fit the bill include real estate, infrastructure, timber, agriculture, and absolute return strategies. Ideally, these assets should have a low correlation to those held in the growth portfolio. The goal is to create a diversified hedging portfolio that’s aligned with a plan’s obligations.

This approach, which we call LDI plus, is most effective when a mix of these assets makes up a segment of the hedging portfolio. This effectively replaces lower-yielding bonds with alternative assets, taking into account their income characteristics along with their correlation to liabilities. The end objective is to enhance returns while maintaining a good match with the plan’s liabilities.

While sitting alongside a mix of bonds in the hedging portfolio, the additional income these assets generate can be used to match more of the plan’s liabilities. Furthermore, including these income-generating real assets in the hedging portfolio can offer a better risk/reward proposition.

Introducing derivatives to an LDI strategy can also make a portfolio work harder. Derivatives are generally underused by all but the biggest pension plans due to perceived complexity or risk. While there can be risks involved in their use, these can be mitigated through good management. We believe pension plans can benefit from the proper use of these instruments; however, it's vital that plan sponsors ensure their investment policy sets out and caps the degree of leverage they're willing and able to use in advance. It's also critical that collateral and liquidity be properly managed.

A traditional LDI strategy uses physical securities in both the hedging and growth components, but this approach can limit a plan’s potential allocation capability. Let's say we have a pension plan with a traditional 60% growth and 40% hedging mix, which naturally results in 100% of its assets being invested. By using derivatives, however, the same plan could maintain its existing growth allocation while simultaneously increasing its exposure to hedging assets. The use of derivatives with varying degrees of leverage allows the plan to hedge a greater proportion of its liabilities and therefore reduce its overall risk exposure.

Leveraging the portfolio in this manner can also be beneficial from a return perspective. As long as the expected return on growth assets is higher than the short-term rates used to leverage the bond exposure, the excess expected return from the growth assets should enhance the performance of the hedging portfolio. If that's no longer the case, then exposure to growth assets should be sold, and leverage reduced or eliminated altogether, while preserving the bond exposure.

LDI was once seen as the preserve of larger pension plans because such strategies typically require robust support for operations and governance. However, as LDI strategies have grown in popularity, the way they’re implemented has become more sophisticated. For instance, collective investment solutions can efficiently address the liability-matching needs of small to midsize pension plans. These pooled products offer sponsors the ability to allocate between a series of funds designed for particular demographics and bond profiles to match the specific liability profile of their plan. They can provide a predictable and stable fund cash flow profile, are liquid—they’re subject to daily valuation—and are easily incorporated within a glide path framework.

The next logical step is LDI for defined contribution plans and individuals. A decumulation strategy can be used to create a kind of personalized defined benefit (DB) plan, where the liability to hedge is the individual’s retirement income.

This can be achieved through implementing a fixed-income portfolio that matches the minimum income a participant hopes to achieve in retirement. The remainder of the assets is then invested in a growth strategy to build on that income and ensure the assets last as long as needed. Again, this resembles a de-risking glide path for DB plans. As with a hedge ratio in a DB plan, gains in the growth strategy are automatically locked in and converted into fixed-income assets, thereby securing more income.

LDI has come a long way since the days of shifting an entire portfolio into bonds and is more than simply managing against a long bond index. Plan sponsors of all sizes can now choose from a range of de-risking solutions based on their individual needs. What's perhaps most important for sponsors is deciding to take that first step. It’s not unusual for plan sponsors to cite low interest rates as a reason to delay implementing de-risking strategies. Some sponsors have indicated that they'd prefer to wait for interest rates to rise further before acting—a view that, at first glance, makes some sense. However, by choosing to wait for interest rates to rise, sponsors would have—by default—chosen to embrace equity risk as well as interest-rate risk.

The implication here is that a plan’s funded status could deteriorate further if turbulence continues. In our view, while we may be hard pressed to define what could be considered the perfect moment to implement LDI or de-risking strategies, there's no wrong time to draw up a plan for the future.

1 "Funded status of largest U.S. corporate pension plans slipped in 2018, Willis Towers Watson analysis finds," Willis Towers Watson, January 2, 2019. 2 The traditional allocation strategy is rebalanced monthly to target allocation; the dynamic asset allocation strategy is rebalanced monthly to achieve a dynamic target hedge ratio.