Practical Considerations for Hedging the Liability Discount Rate: Managing the Credit Spread Hedge

Part 1: Managing the Overall Liability Hedge

Hedging Rates and Spreads Separately

With funded status surging to its highest level in years, many plan sponsors are sharpening the precision of their liability hedging assets (LHA). Some seek to hedge liabilities as a whole by managing LHA directly to liability cash flows or to a blend of market benchmarks approximating the liability characteristics. Others focus on the distinct investible components of the discount rate, namely Treasury yields and credit spreads, by establishing separate Treasury and credit sleeves within LHA.

The first approach may be preferable to some due to its clearer connection to the liability and operational simplicity. In contrast, the latter approach is more complex, but offers well-resourced plan sponsors (or their advisors or OCIOs) a number of advantages that could lead to better asset-liability outcomes:

• Flexibility for plan sponsors to express tactical views on rates and credit;
• A dedicated source of liquidity for benefit payments and “dry powder” (in the form of Treasuries) to allocate to the credit sleeve or rebalance into return-seeking assts;
• Potential to offset liability headwinds and generate alpha by including multiple complementary and/or differentiated credit strategies (rather than managing to a single, liability-based benchmark);
•  Simpler portfolio adjustments if liability cash flows change materially (for example, due to a lump sum window or a pension risk transfer); and, 
• Investible benchmarks and straightforward performance attribution.

Extending the flexibility embedded in this approach to the plan’s investment managers may also be prudent. As active market participants, managers are often able to take advantage of opportunities more quickly than the plan sponsor.

In the rest of the paper, we explore credit spread hedging within the two-sleeve framework. However, many of the same investment principles—active management, spread diversification, and broad guidelines—are applicable to both approaches.

Credit Spread Risk: A Challenge and Opportunity

Hedging liability interest rate risk is relatively straightforward, though there are some nuances, particularly for some cash balance plans. Liability duration and key rate durations can be matched effectively and efficiently thanks to the breadth, depth, and liquidity of physical Treasuries and Treasury derivatives markets.

On the other hand, hedging credit spread risk—a heightened concern in today’s low-spread environment—is much more challenging. Several factors contribute to these challenges, including:

• Often opaque, narrow, and uninvestible universe of corporate bonds comprising the AA discount rate (see callout); 
• High correlation between return-seeking assets and credit spread returns (see the MSCI ACWI Index column in Figure 3); 
• Downgrade headwind embedded in the liability discount rate²; 
• Limited utility of credit default swaps (given their relatively short tenors, basis risk, and limited liquidity); and, 
• Default risk, relative inefficiency, and transaction costs inherent in credit markets.

While it may be tempting to narrow the investment universe to more closely match the AA nature of the discount rate, this limitation could be counterproductive, as it could result in a low yield relative to the liabilities and a high level of concentration risk within the portfolio. Instead, expanding the universe to the entire investment-grade (IG) spectrum and potentially beyond, applying active management in credit, and effectively employing the dedicated Treasury sleeve can alleviate these challenges and help generate alpha.

Balancing Treasuries and Credit

While we address the benefits of a full IG approach in the next section, we first would like to illustrate that expanding the LHA’s investible universe does not necessarily increase tracking error to the liabilities. For example, we ran a ten-year back-test of a hypothetical, fully-funded plan discounted with the FTSE Above Median Pension Discount Curve, hedged with either (1) duration-matched A or higher credit or (2) a 75%/25% blend of duration-matched Treasuries and IG credit, and reflecting Bloomberg index returns.³ A higher (or lower) allocation to Treasuries may be appropriate if the discount rate is less (or more) aggressive, there is a larger (or smaller) return-seeking allocation, and/or the plan sponsor has a lower (or higher) risk tolerance.

As shown in Figure 1, both approaches end up at a funded status below 100%, which is indicative of the downgrade headwind and shows the importance of active management. The two-sleeve approach underperformed during the energy downgrade cycle in 2015-2016 (affecting primarily the BBB space). While not shown here, the A or higher approach underperformed in 2010-2011 following bank downgrades (affecting the A and AA universes). Despite these volatile periods, the average and maximum three-year tracking errors of the two approaches were substantially similar (see Figure 2).

Figure 1. Evolution of Funded Status for a Sample Plan

Source: Bloomberg Index Services, Dodge & Cox, FTSE Russell.

Figure 2. Summary Statistics for a Sample Plan

Source: Bloomberg Index Services, Dodge & Cox, FTSE Russell.

In addition to the investment managers actively managing individual securities, tactical management of the balance between the Treasury and credit sleeves by the plan sponsor can also improve results. For example, increasing the credit weight to 80% just for the three months of April, May, and June 2020 (in the immediate aftermath of the COVID-19-related spread spike in March 2020) would have added 0.5% to the ending funded status. Finally, matching liability key rate durations, via the Treasury portfolio or a Treasury futures overlay, can further reduce tracking error.

Part 2: Managing the Credit Allocation
 

Broad Investment Grade is Attractive Relative to A or Higher

As we noted earlier, there are many benefits to expanding the investible beyond A and AA-rated corporate bonds, not the least of which is a higher yield that may be more closely aligned with the liability discount rate. Other key advantages include (Figure 3):

• Effective AA credit spread hedge. IG credit spread returns are highly correlated across quality levels, and the beta of broad Long Corporate returns to the narrow Long Corporate AA component is moderate, only 1.25. This means that a duration-matched portfolio of approximately 80% corporates/20% Treasuries would exhibit approximately the same credit spread risk (and, currently, also option-adjusted spread) as Long AA corporates.
• Greater diversification opportunity and lower concentration risk. The full IG Long Corporate universe is roughly twice the size of the Long Corporate A or higher universe, both by market value of bonds outstanding—$2.7 trillion vs. $1.3 trillion—and the number of issuers, 534 vs. 260. The considerable concentration of the higher-quality universe (see Figure 4), with Comcast alone comprising 4.2% of the Long Corporate A or higher universe, heightens issuer-specific risk for higher-quality strategies. 
• Well-compensated spread advantage. The 10-year average spread return of Long Corporate is 37% higher than that of Long Corporate A or higher (2.7% vs. 2.0%) but the volatility is only 19% higher (8.0% vs. 6.7%). From a credit loss perspective, the current annual 25 basis point⁴ (bp) spread advantage easily outweighs the average historical 5-year cumulative credit loss difference of 0.18% (0.54% for IG bonds compared to approximately 0.36% for A or higher bonds).⁵
• Lower downgrade potential. A bond rated A or higher is roughly 17% more likely to fall out of the A or higher universe over five years than an IG bond of being downgraded to below investment grade (BIG).⁵ This elevated downgrade risk for an A or higher strategy could lead to forced selling and additional trading costs. 
• Greater alpha potential. While difficult to quantify, the broader investment universe, additional inefficiencies in lower quality buckets, and the potential for split-rated bonds and cross-over trades in the BB/BBB space create more opportunities to add value than in the A or higher space.

As the fourth column of Figure 3 shows, adding non-corporate credit—i.e., taxable municipals and dollar-denominated sovereign and quasi-sovereign bonds—expands the investible universe by another $0.3 trillion and nearly 120 issuers. By virtue of diversification, it also improves risk-adjusted spread returns and lowers the spread beta to Long Corporate AA. Thus, while at first seemingly removed from AA corporates—and the liability discount rate—the inclusion of non-corporate credit could be beneficial.

Figure 3. Characteristics of Long Credit Indices (as of September 30, 2021) 

Source: Bloomberg Index Services, MSCI. Based on monthly observations over 10 years. For MSCI ACWI, the values shown are correlation and beta of MSCI ACWI (Net) return to Long Corporate AA spread return.

Figure 4. Top Issuers in Long Credit Indices (as of September 30, 2021)

Source: Bloomberg Index Services.

Expanding the Credit Universe Further

Including other credit-sensitive, non-corporate fixed income market segments within LHA can further diversify the credit spread hedge, increase yield, or both, though sizing is crucial to ensure that the hedge remains effective. As Figure 5 illustrates, spread returns of these market segments are highly correlated, but their spread levels and spread return betas (to Long Corporate AA) vary widely. Consequently, we believe these strategies may be best used either as tools within an active manager’s toolkit or as temporary, tactical allocations overseen by the plan sponsor, rather than as consistent, stand-alone allocations within LHA. That said, for open or underfunded plans, a dedicated “growth fixed income” component combining high yield (HY), bank loans, emerging market debt, and/or private credit among others, can be a meaningful and differentiated return driver within return-seeking assets.

Returning to LHA, to illustrate our preference for a tactical approach, we consider Agency MBS. Given their low spreads, low durations, and prepayment characteristics, Agency MBS are a relatively poor liability hedge. However, a plan sponsor wishing to express a credit spread hedge ratio underweight may elect to use Agency MBS rather than Treasuries, given similar safety and liquidity dynamics but positive spread carry. Taking an example from the Dodge & Cox playbook, a manager may replace a portion of the IG credit exposure with a modest allocation to Agency MBS bar-belled with select BIG issuers. This combination can achieve a comparable (or even higher) yield as the rest of the portfolio, but with a potentially greater opportunity for appreciation (from the fallen angels) and a higher degree of protection (from Agency MBS). On the other end of the spectrum, following periods of extreme spread-widening such as March 2020, a broad beta play in BIG bonds by the plan sponsor (or a manager) or a judicious allocation to select BIG issuers, the preferred Dodge & Cox approach, can be more advantageous than simply overweighting IG credit.

While subject to many of the same macro factors as U.S. corporate bonds, emerging market debt, commercial MBS, and non-USD corporate bonds often exhibit additional segment-specific and issuer-specific risk (such as country and currency risk, property-segment and geography risk, and differentiated issuer pricing across markets and currencies, respectively). As a result, these segments may be a fertile ground for generating excess returns, particularly through security selection, but at the same time, their hedging effectiveness may be limited. For example, currency risk is not directly correlated to the liability discount rate. This, along with unpredictable trading volumes in some markets, typically higher management fees, and other complexities, raises the bar for their inclusion within LHA and, even if included, may constrain the size of the allocation.

Figure 5. Characteristics of Non-Corporate Fixed Income Market Segments

Source: Bloomberg Index Services.

Finally, since bond durations within these market segments are often below typical liability durations or LHA benchmark durations, care must be taken to ensure that interest rate hedge ratios remain on target. This can be achieved via either the Treasury sleeve at the plan sponsor level or a Treasury futures overlay at the manager level.

Active Management: Credit Selection Is Critical

As noted earlier, the keys to successful credit spread hedging are generating sufficient yield, mitigating downgrade and default headwinds, and maintaining reasonable tracking error. Effective active management helps achieve these objectives and can generate credit alpha, thereby improving funded status, offsetting plan expenses, and guarding against adverse assumption changes and plan experience. Today’s historically low credit spreads as a whole and low dispersion of credit spreads across sectors make security selection critical to achieving these goals.

At Dodge & Cox, we construct and manage diversified, transparent portfolios with the goal of generating recurring alpha in a risk-controlled, liability-aware manner. We employ our internal team-based fundamental credit research process, draw on our out-of-benchmark expertise in non-corporate credit and structured products, and maintain the ability to respond quickly to evolving market and issuer dynamics. As shown in Figure 6, since its inception in 2009, the Dodge & Cox Long Credit composite performance has generated meaningful alpha while maintaining reasonable tracking error over rolling three-year periods. Historically, the vast majority of our outperformance is a result of security selection. We bring the same mindset to our clients who take a higher-quality approach, maintain blended market benchmarks for all their liability-hedging managers, or manage LHA directly to liability cash flows, subject to their specific investment guidelines.

Figure 6. Dodge & Cox Long Credit Composite Excess Return and Tracking Error to Bloomberg Long Credit Index

Source: Dodge & Cox, Bloomberg Index Services. For the purpose of showing characteristics of the strategy, tracking error for the Dodge & Cox Long Credit Composite is calculated gross of fees.

In Closing

As plan sponsors seek to hedge their pension liabilities more closely, a two-pronged approach targeting interest rate risk and credit spread risk may be optimal. Restricting portfolio credit quality to more closely match the AA (or not-so-AA) nature of the liabilities may actually lead to liability hedging portfolios that lag the liabilities, exhibit elevated concentration risk, and limit alpha potential. Plan sponsors can alleviate these challenges while effectively tracking their pension liabilities and increasing their investment flexibility by balancing an IG credit sleeve with a dedicated custom Treasury sleeve. Dodge & Cox has a demonstrated track record of using the entire IG universe, as well as out-of-benchmark exposures, to deliver credit alpha while carefully managing portfolio risk.

When Might a Higher Quality Strategy Make Sense?

We believe that that the majority of plan sponsors are well served by distinct Treasury and credit sleeves within their LHA, with weights and exact sleeve mandates determined by plan sponsor objectives at different points on their pension journey. However, certain situations may be better suited to a single-sleeve, higher credit-quality portfolio.

One example is frozen, highly overfunded plans in hibernation. As their capital base far exceeds the liabilities, these plans can remain healthy with less asset-liability risk, less alpha, and less active management. Consequently, assuming relatively stable liability cash flows (i.e., no erratic lump sums that might require unexpected liquidations), matching cash flows and holding to maturity could be an effective and low-fee way to defease a portion of the liabilities. Focusing on higher quality securities with a lower probability of defaults would be supportive of this “buy-and-hold,” benchmark-agnostic approach.

Due to its lower yield and alpha potential, a higher quality portfolio would likely reduce the cushion against adverse plan experience or assumption changes. However, given that a pension surplus is essentially “trapped” and cannot be withdrawn easily, some deterioration in funded status may be acceptable. Another implication of this approach is likely a lower expected return on assets and the resulting decrease in pension income, the materiality of which would depend on the size of the plan relative to the corporate balance sheet. Consequently, plans should decide to increase credit quality in the context of broader enterprise risk management.