Top considerations for (Corporate) UK Defined Benefit pension schemes in 2026 

As we say goodbye to a heady 2025 that saw AI take centre stage in driving the equity bull market even higher and a rally in corporate bonds that leaves spreads around their all-time tightest levels, we consider the key investment issues dominating the UK defined benefit pension scheme landscape in 2026.

1. AI concentration – not just in equities
2. The implications of lower expected future returns for equities and credit
3. The changing landscape of the risk transfer market
4. Thinking differently about risk

With modest allocations to equities, corporate DB pension schemes have had to pay limited attention to the impact of AI related positions in their portfolios. Where equity exposure does exist, schemes have enjoyed the benefits of the AI led equity bull run, often via passive or synthetic exposures. However, given the modest exposure at a strategic level, the related concentration issues within equities to tech and AI have not demanded the attention it once did.

In 2026, however, the AI juggernaut is coming to the bond market in a big way! In fact, it had started to approach in September 2025, but it really accelerates in 2026. An explosion of AI related new bond issuance, mainly to fund data centres, means pension fund trustees will need to carefully assess the implications and potential concentration risks in their portfolios.

The race for AI dominance driving debt issuance during 2026 is summarised below:

• JP Morgan estimates AI-linked companies will account for 14% of its investment grade index, surpassing U.S. banks as the dominant sector.
• In the asset back securities market, BofA estimates that data centres backed issuance is expected to reach $115 billion by the end of next year. This compares to total issuance of c.$457 billion in 2025 (up to November, according to SIFMA Research).
• Morgan Stanley estimates private credit markets could supply over half the $1.5 trillion needed for the data centre buildout until 2028

It is quite clear that the AI data centre build out will become a dominant part of the credit universe in 2026 and beyond, both in public and private markets and spanning investment grade and sub investment grade. Lower turnover ‘buy & maintain’ portfolios may be impacted less immediately, but new mandates that are ramping up exposure, passive portfolios and benchmark aware active investors will feel the impact. As AI exposure moves significantly into the bond markets, concentration risks will need to be considered at a total portfolio level, not just at an equity level.

Insurers, of course, also face the same challenges as pension schemes in terms of the wall of data centre debt issuance facing the market and could be bigger buyers in filling their new business books.

With data centres having little precedent on this scale other than within the infrastructure equity market, some will look to the recent past to better understand the risks to these companies in which we invest and the securities they issue. Whether the sector will experience something similar to the Technology Media Telecom boom of the early noughties, which saw companies laden with debt (and which went on to underperform expectations) participate in a bidding war to win mobile telephone licenses, is something we will have to wait and see. The Credit Default Swap market (i.e. insurance against a corporate bond defaulting) certainly seems to think there are significant risks given CDS spreads on AI related companies have been widening significantly (i.e. insurance against default is getting more expensive).

With equities at heightened levels, credit spreads at all time tights where should DB investors look to in order to eke out excess return over and above their discount rates?

Contractual income has long been the mantra for DB pension funds in seeking out dependable returns. However, this is getting increasingly hard to find and justify in a world where additional yield is so thin. Yet there are ways of protecting credit portfolios by reducing the sensitivity of a portfolio to credit spread widening whilst maintaining or improving yield; by tilting into shorter dated and more attractive areas of the credit market.

Many trustees will be undertaking investment strategy reviews with return assumptions which are materially different to those two or three years ago. Not surprisingly the results will be different. For those with strategic monitoring that incorporates expected return analysis, this may not come as a shock. For others the lower return environment will need careful consideration in relation to investment objectives and how existing portfolios need to evolve in order to achieve these.

Given long dated gilts are one of the few traditional assets that look relatively good value, increasing allocations to defensive fixed income and / or reviewing hedging levels and considering the role of hedging when surplus occurs could be an easy win.

The equity bull run and the benefits of AI may have much further to run and earnings growth has been supportive of the story so far. Data centres could turn out to offer attractive contractual income opportunities if structured right. So, we should not turn our back on credit and equity risk but it will be important to be measured when sizing positions and to consider rebalancing exposures to where they are expected to be rewarded. 

The usual rules of diversification will continue to apply – perhaps more than ever. There continues to be other asset classes that Mercer believe are attractive that can offer respite and safer havens in the form of better forward-looking risk-adjusted returns. Often these asset classes or strategies, like trade finance, hedge funds or frontier emerging market debt, are not as easily digestible in terms of execution, whether due to some aspect of complexity or because they pose a governance hurdle. The same can be said for the use of protection strategies. However, that should not stop investors considering all the options and looking to review and enhance their governance arrangements where it can result in more durable portfolios and better outcomes.

Two seismic events transformed the risk transfer market in 2025, the full effects of which will start to be seen in 2026. At the beginning of the year the rules regulating insurers (Solvency II UK) were loosened making it more attractive for insurers to invest in securitised assets and infrastructure projects. In the second half of the year, PIC and Just were acquired by insurers backed by the well know private credit manager, Apollo, and the infrastructure manager, Brookfield, respectively.

Without too much speculation, it would be fair to presume that these bulk purchase annuity providers will now look to maximise the return on their underlying portfolios by investing more in private market assets, where their parent companies specialise. The new flexibility afforded through regulation may help facilitate this, as may any reinsurance subject to PRA approval.

This will help these insurers offer more competitive pricing, reflecting the high returns (and possibly higher risks) of the underlying portfolios. This puts pressure on others in the market to do similar – noting Legal & General has already taken steps to protect its competitiveness through its partnership with Blackstone.

Arguably this underlies the risk of an irredeemable asset (i.e. a bulk purchase annuity) – it may not end up being managed by the same people or invested in the same way as when you bought it. These changing dynamics undoubtedly make trustees’ jobs more difficult in assessing the security of such policies versus other endgame / run-on options, albeit noting that all insurers balance sheets are well in excess of the PRA requirements. Nevertheless, we anticipate a growing interest in managing schemes to generate and run-on with a surplus.

Whilst turbulent waters may lie ahead, most corporate DB schemes have de-risked, have a surplus and as stochastically modelled, can expect to be able to withstand some headwinds and still be on track to firmly extinguish their liabilities by paying pensions.

Rather than nailing down investment risk to the nth degree on a fully funded position, we are helping more clients explore ways of enhancing returns to grow a buffer or surplus to either guard against unforeseen risks, or to benefit the sponsor and scheme members.  In many cases, taking a longer-term view and embracing some modest risk taking to enhance returns can actually result in a higher likelihood of avoiding a negative experience over the long-term compared to a very low risk investment strategy.

Taking a holistic view of a portfolio, a ‘total portfolio’ approach, and drilling down to exposures by sector, region, credit quality, issuer, tenure, etc, isn’t always as easy as it sounds, particularly for well-diversified portfolios across multiple investment managers.  But it can be revealing in highlighting unappreciated risks that traditional asset-liability modelling would not pick-up. 

Some of our clients have gone further and undertaken fire drills and deep-dive sessions scrutinising particular risks; operational, geopolitical and liquidity to name but a few. Looking to gain a better appreciation from a wider perspective of what could “break” the scheme and making adjustments, sometimes to governance structures, other times to asset strategies, can help make portfolios more robust.

Given the evolving investment and economic landscape, we expect 2026 to be an important year for reassessing investment strategies to realign with key objectives. We have found that strategy workshops, frequently paired with a review of investment beliefs, provide an effective forum for consolidating perspectives on investment strategy while also facilitating forward planning.

We look forward to discussing these topics and others with you as 2026 unfolds and hope this short paper has given you something to think about in the weeks and months ahead.