1H 2026 Pension Perspectives

Funded Status Drivers

In the first six months of 2026, investors navigated volatile capital markets, marked by the conflict in the Middle East, swings related to the AI build-out, a new Federal Reserve chair, and an evolving tariff policy. The markets continue to price in ongoing global growth, while beginning-of-year expectations for moderating inflation and near-term Federal Reserve cuts have been largely reversed or at least delayed.

Reflecting these sentiments, long-term Treasury yields rose 10 to 30 basis points (bps), and the yield curve flattened. Despite the conflict in the Middle East and record $1.2 trillion in year-to-date new issuance, long-term investment-grade (IG) credit spreads were 1 bp tighter at June 30 compared to prior year end. As a result, broad U.S. fixed income as well as long duration corporate bonds posted positive returns during the first six months of 2026. Global equities returned 11.25%, with particularly strong performance in emerging market equities.

In aggregate, large corporate defined benefit (DB) plans weathered the volatility well, aided by strong equity market returns and an uptick in discount rates. The Milliman 100 Pension Funding Index rose by 3.4 percentage points to 109.5%, its highest level since July 31, 2001.

Investment Positioning

We continue to believe that duration-neutral positioning may be appropriate for most plan sponsors. Given unabating inflation, a resilient labor market, and AI-fueled productivity growth, our baseline macro outlook calls for interest rates to remain high in the near term, likely not easing until 2028. However, there is a large degree of uncertainty, and interest rate volatility may be particularly elevated due to geopolitical risks, upcoming U.S. mid-term elections and the potential for a divided government, and less transparent Fed guidance under new Fed Chair Kevin Warsh.

Within Credit, we believe plan sponsors are well-served by being underweight credit risk relative to targets, selective within Credit, responsive to episodic widening, and open to non-corporate spread sectors, such as government-related bonds and high-quality securitized assets.

Despite record new issuance, index-level credit spreads are near their lowest levels in 25 years (see Figure 2). These levels are supported by broadly strong corporate fundamentals and attractive all-in yields, drawing a wide range of yield-based buyers. Consequently, spreads could remain range-bound as long as Treasury yields remain high and corporate fundamentals do not deteriorate materially. However, we continue to believe that over a longer time horizon, broad or sector-specific episodes of spread widening are likely, suggesting that price discipline, patience, and the ability to move quickly, both at the asset allocation and portfolio levels, could benefit investors.

In addition, year-to-date index-level stability in credit spreads masks increasing sector and quality dispersions, supporting a more selective approach within Credit (see Figure 3). The widening in the Technology and Communications sectors and the Aa space are at least partly a result of expansive AI-related new issuance; notably, the spread differential between long dated A- and Aa-rated corporates is currently less than 2 bps, compared to 11 bps at year end and 23 bps on average over the last 10 years (and 27 bps over the last 20 years).

Underweighting Credit in favor of Treasuries or avoiding higher-spread sectors or issuers could lead to a yield give-up relative to the liabilities or market benchmarks, but incorporating credit-adjacent assets (in place of Treasuries) and active management could offset this risk to some degree.

AI-Related New Issuance

As noted earlier, IG new issuance in the first half of 2026 was a record $1.2 trillion, with $220 billion of that being long-dated (over 10 years in maturity). Just four issuers—Alphabet, Amazon, Meta, and Oracle—accounted for $107 billion, or roughly 10%, of total IG new issuance and $45 billion, or roughly 20%, of long-dated IG new issuance. Nvidia, Salesforce, and SpaceX each also issued roughly $25 billion, including $6.0-$6.5 billion in long-dated debt. Not surprisingly, this new issuance is driving up issuer concentration, especially in the long Aa space from which most liability discount curves are constructed (see Figure 4).

Indeed, the three Aa-rated hyperscalers—Alphabet, Amazon, and Meta—now comprise 35% of the Bloomberg U.S. Long Corporate Aa Bond Index, up from 17% two years ago, 13% five years ago, and 0% ten years ago. This means that hyperscaler spread movements, and their effect on adjacent issuers and sectors, are likely to have an outsized impact on liability discount rates (relative to other market segments). While Aa issuer concentration is still below 2006 levels, when the Banking sector comprised 44% and Citigroup alone was 19.3% of the Bloomberg U.S. Long Corporate Aa Bond Index (see Figures 5 and 6), concentration could increase further, given expectations for additional massive future AI-related issuance.

The most concerning scenario would be a potential wave of downgrades out of the Aa space, like the one that affected banks in 2011. Should that occur, asset and liability returns could diverge widely, as downgraded issuers would fall out of the discount curve, driving discount rates lower (and liability values higher) but remain in portfolios, driving portfolio yields higher (and market values lower). We encourage plan sponsors to be mindful of the possible implications of AI-related issuance for both assets and liabilities.

Monetizing DB Surplus to Fund DC Benefits

Shifting to strategic considerations, the Strengthening Benefit Plans Act of 2025, which would allow for the funding of DC benefits from DB plan surplus assets, appears to have low prospects of advancement in Congress. In the meantime, funding DC benefits from the DB surplus requires either plan termination (and the establishment of a qualified replacement plan) or re-opening of the DB plan and shifting employer DC benefits to the DB plan. For sponsors considering the latter, a market-return cash balance design could be especially attractive as it would reduce plan sponsor investment risk and offer employees a lump sum benefit and market return they may be familiar with in the DC plan.

PRT Market Evolving

In the PRT space, recent trends suggest growing interest in buy-ins and a potential slow-down in buy-outs, especially among large plan sponsors. In 2025, PRT activity totaled $48.6 billion with $31.2 billion in buy-outs and $17.5 billion in buy-ins (see Figure 7). While the aggregate volume is similar to the prior three years, buy-in premiums surged nearly five-fold compared to the prior three-year average, and the number of buy-in deals doubled. Buy-ins are typically an interim step toward buy-outs, but since they are rarely disclosed in company annual reports or insurer press releases, it is difficult to ascertain drivers of this apparent surge.

Buy-outs and overall PRT activity slowed down in the first quarter of 2026, totaling just $3.8 billion compared to $7.1 billion in the first quarter of 2025 and $14.6 billion in the first quarter of 2024. We do not believe this trend is necessarily indicative of a widespread change in plan sponsors’ approach to PRT. Instead, plan sponsors may not feel the same sense of urgency to execute PRTs in the context of strong and stable funded status, equity market volatility, expectations of higher interest rates, and unresolved PRT lawsuits.

The buy-out market continues to be dominated by smaller transactions. We believe this may be at least partly due to PRT-related lawsuits. Of the 10 major PRT cases, half have been dismissed with the other half either surviving a motion to dismiss or awaiting a final ruling on such a motion. In 2026, the government weighed in as well. The DOL filed an amicus brief supporting the plan sponsors, arguing that (a) retirees should have no standing to sue as they have not suffered any harm (yet) and (b) different fiduciaries may reasonably make different choices with respect to the “safest annuity possible” under its guidance. As expected, the PBGC re-affirmed its view that PBGC protection ceases once a PRT occurs (for obligations that are part of a PRT), even if the PRT insurer fails in the future.

Market-Return Cash Balance Design

Market-return cash balance (MRCB) plans have been gaining traction by offering a DC-like benefit within a DB framework, while significantly reducing, or even eliminating, plan sponsor investment risk. For example, four major U.S. airlines incepted MRCBs for their pilots in 2023 and 2024. A proposed amendment to U.S. accounting standards could make MRCB design even more attractive by aligning the accounting treatment of these plans with their economic reality.

In a MRCB plan, the benefit is defined as a hypothetical account balance, credited annually with a pay-based credit (such as 5% of pay) and an interest credit equal to the return on plan assets (or a similar market-based rate). This crediting rate differs from the uninvestable rates common in legacy cash balance designs, such as a fixed rate or the 30-year Treasury yield. Since interest rate credits reflect actual return on plan assets, if contributions are equal to the pay credits, then plan assets are equal to the sum of the hypothetical account balances, meaning that the plan is effectively fully funded (subject to certain nuances) from an economic perspective.

However, under current accounting standards, most plan sponsors calculate the plan liability by projecting forward hypothetical account balances to retirement dates with the expected return on assets (EROA) and discounting back to present value with a high-quality bond curve. Since the EROA is typically higher than the discount rate, the resulting liability is larger than the sum of the hypothetical account balances, creating “artificial” underfunding and negating the key “market return” feature of MRCBs.

In June, the Financial Accounting Standards Board (FASB) proposed setting the discount rate equal to the EROA for MRCBs.³ Thus, under the project forward/discount back methodology where the EROA equals the discount rate, the plan’s liability and the sum of the hypothetical account balances would be equal, matching the economic perspective. Certain nuances, like the preservation of capital rule⁴ and participants’ annuity elections, can be generally addressed in straightforward ways. However, other considerations, such as ERISA funding requirements and PBGC premiums, which still rely on the traditional methodology, remain. Comments on the proposal are due by August 10, with the FASB final decision and possible effective date expected in subsequent months.

As always, we welcome the opportunity to speak with you or your advisor about our pension risk management solutions as you progress along your pension journey.