Man Institute 2026 H2 Credit Outlook

The Gulf conflict has understandably dominated headlines in recent months, but credit markets have had another story to tell. Artificial intelligence (AI) and hyperscaler bond issuance is booming in public markets, while in private credit, a series of high-profile defaults has placed AI and software deals under the microscope. In light of these developments, we are devoting this section of the credit outlook to AI and its growing footprint across public and private credit markets. The regular sector outlooks follow below. We also offer a snapshot of how we are using AI to advance our credit capabilities.

The pace of bond issuance in the AI and hyperscaler space year to date has been nothing short of staggering – and expectations only point to an acceleration in the coming quarters.

Source: Dealogic, Pitchbook | LCD, Morgan Stanley Research forecasts, as of May 2026. As a point of comparison, internet and tech issuance during the dot-com era peaked at US$85 billion in 2001, representing an average of 14.5% of total issuance across 2000–2001. The current cycle threatens to dwarf that period: at the high end, Morgan Stanley estimates total AI and hyperscaler-related issuance could reach US$400 billion in investment grade (IG) and a further US$65 billion in high yield and loans, potentially accounting for nearly 20% of total issuance. Bear in mind that, at the start of 2015, the hyperscalers (Amazon, Microsoft, Meta, Alphabet and Oracle) represented less than 1% of the market.

Source: Dealogic, Pitchbook | LCD, Morgan Stanley Research forecasts, as of May 2026.

Credit markets are already feeling the weight of that supply. Year to date, four of the five hyperscalers have underperformed the broader market on a spread basis.1 The scale of AI-related capital expenditure (capex) is also impacting US GDP, where economists forecast a contribution of around 1% in 2026,2 potentially papering over cracks elsewhere. Despite spread widening, we do not believe levels adequately compensate investors for the risks given the supply overhang and, in our view, growing concerns around competition, buildout delays fuelled by labour shortages, increased public backlash and regulatory intervention.

While supply alone is not necessarily sufficient to trigger a broad sell-off, the sheer scale of forthcoming paper risks pushing spreads wider across the credit complex. The substitution effect is a key transmission mechanism: as investors are required to absorb ever-larger volumes of higher quality AI-related supply (and Treasuries to finance fiscal deficits), compensation in the form of wider spreads across lower quality issuers becomes increasingly necessary to clear the market. While many of these deals benefit from robust structural protections including amortisation features, construction guarantees, and hyperscaler lease backstops, the enthusiasm for the sector has encouraged significant overreach. We believe investor caution is warranted given the execution risk inherent in such a rapid buildout: capital intensity has not yet peaked, off-balance sheet leverage is rising, and sustainability of returns on AI investment remains unproven, particularly for neo cloud and data centre operators further down the credit quality spectrum. We are particularly concerned about issuance in the high yield and leveraged loan markets, where many borrowers remain firmly free-cash-flow negative.

Source: JP Morgan, as of 9 June 2026.

Sector vulnerabilities are not new to credit markets. Recall the mini cycles in internet and telecoms in the early 2000s, European sovereigns and banks in 2012, energy in 2015-2016 and real estate/cyclicals in 2022. Stress rarely discriminates within vulnerable sectors, but this might create opportunities to buy the best houses in the worst neighbourhoods. For now, we have been focusing on software companies with stronger ties to healthcare and government projects, where margins – while under pressure – remain defensible and refinancing within maturity timeframes appears achievable. Despite all-in spreads remaining expensive, we are seeing signs of dispersion, which we expect to create opportunities.

Source: ICE BofA, Bloomberg, Man Group analysis, as of 30 April 2026.

The dispersion visible in public markets is arguably more pronounced in private credit. A series of high-profile defaults in software and software as a service (SaaS) has placed the sector under the microscope amid comparisons to the dot-com era and growing fears that infrastructure investments could be leapfrogged by newer technologies. Retail outflows, influenced by these concerns, have been portrayed in the media as the beginning of the end of private credit. We prefer to characterise these events as growing pains for the asset class, though credit quality and disciplined manager selection will remain paramount. When assessing the tech and software portion of the market, which today represents roughly 20% of the middle market direct lending universe,3 we believe a nuanced approach is important. The key distinction is between businesses that are deploying AI to enhance and entrench their competitive position versus those whose core product is being displaced by it. Mission-critical enterprise software with deep customer integrations, high switching costs, and strong pricing power is a very different credit proposition than a point-solution, commoditised SaaS product. Not all software will face existential disruption, and indiscriminate avoidance of the sector is as much a mistake as indiscriminate concentration within it. The question is whether the borrower is ahead of the curve on AI adoption – using it to widen its moat – or behind it, facing margin pressure and customer attrition.

Public market credit investors in the AI space face an uncomfortable asymmetry: they bear meaningful exposure to execution risk and buildout delays, yet receive none of the equity upside if the boom plays out as bulls expect. That mispricing of risk is coiling a spring: the greater the enthusiasm today, the more violent the eventual correction is likely to be. With that said, we are not advocating avoidance. What we believe is required is rigorous credit selection across public and private markets, a clear-eyed view of which borrowers are ahead of the AI curve, and a portfolio that is diversified enough not to be held hostage to the AI story playing out as expected. For investors seeking broader diversification away from the boom, European and emerging market credit might offer an increasingly compelling alternative. 1. As of 8 June 2026. 2. Source: Morgan Stanley, as of 31 May 2026. 3. Source: Pitchbook – “Peek under the Hood: An analysis of private credit loans in top public BDCs”, 28 April 2026. No paper on AI would be complete without some perspective on how it is changing our day-to-day credit analysis. Below, we describe one of the many tools we have built to improve the quality and efficiency of our research process. Business development companies present a particular analytical challenge. Their portfolios are typically large, heterogeneous and reported with a lag, making it difficult to identify stress before it surfaces in headline numbers. Our AI-driven monitoring framework aims to address this by tracking payment-in-kind (PIK) activity, distinguishing loans structured as PIK from inception from those that toggle mid-life, with the latter signalling deterioration. It also aims to detect synthetic PIK, where companies quietly defer cash interest in ways that standard filings do not capture. In addition, a dedicated module scores every portfolio company on AI disruption exposure and adaptive capacity, identifying which lenders carry the most concentrated risk to AI-vulnerable industries. Combined with additional signals across sector classification, loan performance and mark migration, the framework seeks to provide an earlier read on distress than public filings alone would allow.

At Man Group, we have no house view. Portfolio managers are free to execute their strategies as they see fit within pre-agreed risk limits. With that in mind, click below to expand the outlooks for the second half of 2026 from our different credit teams.