Stay informed with free updates
Simply sign up to the Technology sector myFT Digest — delivered directly to your inbox.
Key Takeaways:
- Systemic Interdependencies Amplify Risk:The burgeoning AI ecosystem is characterized by deep, often circular financial connections spanning equity, debt, and compute commitments. While fostering growth, this intricate web creates significant systemic risk, meaning a downturn or failure in one key player could trigger cascading effects across the entire value chain, impacting valuations and credit health.
- Circular Financing Raises Valuation & Credit Concerns:A substantial portion of AI financing involves ‘closed-loop’ transactions where capital flows between partners who are also customers, potentially inflating perceived demand and revenue. This practice complicates fundamental analysis, challenges traditional credit ratings, and could mask underlying financial vulnerabilities, raising questions about the true health and sustainability of some AI ventures and their long-term profitability.
- Heightened Due Diligence is Critical:Given the complex, interconnected nature of the AI market, investors and lenders must move beyond siloed analysis. Understanding the creditworthiness and operational resilience of any single firm now requires a holistic, system-wide view, factoring in supply chain dependencies, leverage profiles, and the quality of revenue generated through inter-company commitments to accurately assess risk and opportunity.
The financial connections weaving through the technology sector, particularly within the burgeoning Artificial Intelligence (AI) ecosystem, are not merely substantial; they are becoming increasingly intricate and, in certain respects, circular. While these deep interdependencies can accelerate innovation and market growth, they simultaneously introduce a level of systemic risk that warrants close scrutiny from investors, regulators, and market participants. In a world where corporate failures are an undeniable reality, the health of seemingly robust AI firms could be profoundly influenced by the stability of seemingly distant, yet financially intertwined, weaker links.
The prevailing narrative around AI has largely focused on its transformative potential, driving unprecedented capital inflows and soaring valuations across public and private markets. However, as the market matures and competitive pressures intensify, the underlying financial plumbing supporting this growth becomes paramount. The risk isn’t just about individual company performance; it’s about the potential for contagion in a highly coupled system, reminiscent of lessons learned from past credit crises or asset bubbles, where seemingly isolated issues rapidly cascade into broader market dislocations.
A new, timely paper by Craig Nicol, Oleg Melentyev, and Charlie Callan of Sona Asset Management, a prominent London-based credit hedge fund, offers a crucial lens through which to understand these dynamics. Their exhaustive effort to map the AI universe and meticulously catalogue the interconnections between major players reveals a web far more tangled than many might assume. The findings serve as a critical warning for those navigating the AI investment landscape, highlighting potential vulnerabilities that could ripple through global markets and challenge conventional risk assessment frameworks.
The authors generously shared their meticulously mined data, which we processed into a radial network chart. While perhaps not the most mobile-friendly visualization, it powerfully conveys the sheer scale and complexity of financial connections between various entities. These connections encompass a broad spectrum, including equity investments, debt financing, and crucial lease commitments for compute infrastructure – all vital components of capital allocation in this capital-intensive sector. Each node’s size in the chart is proportional to the total value of commitments it is reportedly attached to, offering a visual proxy for its centrality in the ecosystem and its potential systemic importance.
Some content could not load. Check your internet connection or browser settings.
For instance, the substantial green OpenAI blob vividly illustrates its central role, with vast commitments flowing both into and out of the entity. This reflects not just its technological leadership but its gravitational pull for capital. Conversely, smaller grey blobs, such as one representing Citadel, appear diminutive because specific commitment figures might not be publicly disclosed, even if their involvement in significant deals – like a reported $2bn Series C equity capital raise for neocloud provider Nscale – is known. This highlights a persistent challenge in market transparency within such a rapidly evolving and often privately funded sector, complicating the task for investors attempting to fully price risk.
The data compiled by Sona Asset Management is staggering: 176 deals totaling an estimated $3.6 trillion, involving 202 distinct entities, transacted over a three-and-a-half-year period leading up to August 2026. This figure likely understates the true value, given the undisclosed sums in numerous transactions. The sheer volume of capital flowing into and around the AI sector underscores the intense market interest and belief in its future, but also raises pressing questions about the sustainability of these valuations and the underlying health of the financial mechanisms supporting them, especially if market conditions tighten or growth expectations falter.
At the heart of this intricate web are foundational AI labs like OpenAI and Anthropic, both of which have accumulated colossal commitments. While their innovative prowess is undeniable, their current lack of positive (unadjusted) earnings positions them as crucial, yet potentially vulnerable, nodes. In a market increasingly sensitive to profitability and cash flow, particularly amidst rising interest rates, the financing structures supporting these ‘monoline AI labs’ warrant heightened attention. Their valuations are often predicated on future growth and market dominance, making them highly susceptible to shifts in investor sentiment or competitive landscape. A closer look at their financial commitments, facilitated by Sona’s data, reveals an even more complex picture for market analysts.
Some content could not load. Check your internet connection or browser settings.
The two Sankey diagrams detail the various commitments made to these labs and, critically, the commitments they have subsequently made to others. While often capital appears to flow from distinct sources, this isn’t universally the case. As the authors astutely observe, these relationships frequently exhibit characteristics of ‘circular financing’, where money effectively cycles within the ecosystem rather than representing purely external capital inflows.
Amazon has invested in OpenAI (~$50bn) while also securing a ~$100bn compute commitment from it; Microsoft’s ~$135bn ownership recapitalisation of OpenAI sits alongside OpenAI’s ~$250bn purchase commitment back to Microsoft Azure.
This dynamic extends further, creating a truly ‘closed-loop system’ that could have significant market implications:
NVIDIA has conditionally committed to invest up to ~$100bn in OpenAI (2025) with a further ~$30bn indicated in 2026; OpenAI in turn commits hundreds of billions of dollars to Oracle, Microsoft and Amazon for cloud and compute; these companies are among the largest buyers of NVIDIA silicon: the circle is complete.
Such arrangements, while strategically logical from a business development perspective, raise red flags in terms of financial transparency and systemic risk. When capital investments are coupled with future purchase commitments between the same entities, it can create an artificial impression of robust demand and financial strength. For investors, this makes it exceedingly difficult to discern genuine, market-driven revenue growth from internally generated, circular transactions. This opacity can distort valuations, making firms appear more financially stable or having higher revenue visibility than they might be in a truly open market scenario. It also raises questions about the quality of earnings and the true underlying cash flow generation, which are critical for long-term shareholder value.
Indeed, Sona’s analysis found that approximately 120 of the 176 transactions analyzed carried a ‘high circularity flag,’ indicating multiple, often looped, relationships between connected nodes. This high prevalence of circular financing resonates with broader market concerns. Just weeks ago, S&P Global Ratings highlighted the risk that ‘open models eat frontier labs’ lunch’ and, more pointedly, listed the presence of circular financing as one of its ‘known unknown’ risks to hyperscalers’ credit ratings. This directly impacts their cost of capital, potentially leading to wider credit spreads and greater scrutiny from bond investors. For equity investors, it implies a higher risk premium and necessitates a deeper dive into the quality of reported earnings and future revenue streams, as seemingly strong financial performance might be built on an internally reinforcing cycle rather than broad market demand.
While our charts provide one perspective, Sona’s own static visualization offers an equally compelling overview of the AI ecosystem’s ownership and financing web:
Sona’s model thoughtfully segments entities into four distinct orbits, providing a structured view of the complex interdependencies. The central orbit houses key AI players, memory, and foundry majors; the second ring encompasses other pivotal companies; the third identifies investor syndicates; and the outermost ring details investors and financiers. This layered approach helps market participants visualize not just the direct links, but also the ripple effects across different tiers of the ecosystem, from foundational technology providers to ultimate capital providers, informing a more nuanced assessment of risk and opportunity across the entire value chain.
Beyond direct financing, the fate of tech firms is also intricately linked through supply chain dependencies. Leveraging Bloomberg supplier data, Sona meticulously mapped the concentration and dependency of the AI supply chain by analyzing firms’ revenue reliance on one another. Unsurprisingly, chip firms largely sell to chip buyers. However, the analysis also revealed that certain companies’ revenue streams are almost existentially tied to the capital expenditure decisions of just one or two dominant players. This concentration of revenue dependence presents a significant operational and financial risk. A strategic shift, budget cut, or even a change in technology roadmap by a key client can severely impact a supplier’s top and bottom lines, affecting their stock performance, credit profile, and ability to meet debt obligations. Such vulnerabilities are particularly acute in periods of market uncertainty or technological transition.
Here’s a chart illustrating the top five companies’ revenue dependence on the AI ecoverse, demonstrating these critical linkages:
Some content could not load. Check your internet connection or browser settings.
Further enhancing this risk profile, Sona overlaid leverage – framed as total debt to EBITDA – onto their mapped network. This allows for a combined visualization of supply chain connections, overall debt burden, and financial risk. The resulting chart provides a powerful visual cue for identifying potential weak spots in the ecosystem, areas where financial strain could quickly become systemic:
In this visualization, larger bubbles indicate greater total debt, while a darker red hue signifies higher leverage. While most bubbles lean towards a healthier greenish tint, the prominent bright red of entities like Oracle and CoreWeave immediately draws attention. High leverage, especially when coupled with significant revenue dependency on a few key clients, magnifies financial risk. In an environment of fluctuating interest rates and potential economic slowdowns, companies with elevated debt burdens are more susceptible to refinancing risks, covenant breaches, and ultimately, distress, which could then cascade through their interconnected ecosystem, leading to wider credit spreads and investor flight from perceived risky assets.
To be clear, circular financing isn’t inherently illicit, nor is it unusual for chipmakers and cloud providers to derive substantial revenue from hyperscalers and AI labs. These interdependencies are a natural outcome of strategic partnerships and vertical integration attempts in a rapidly evolving industry. However, the profound and increasingly incestuous network of revenue, capital expenditure, and financing connections fundamentally alters the landscape for financial analysis. It means that assessing the creditworthiness, growth prospects, or even the intrinsic value of a single element within the AI system necessitates a comprehensive understanding of the entire, interconnected web. This holistic approach is no longer a luxury but a necessity for robust risk management and informed investment decisions, particularly as we acknowledge that, indeed, firms do sometimes fail.
Further reading:
— Are credit rating agencies getting fed up with hyperscalers? (FTAV)
— How big is the open-model threat to AI hyperscalers? (FTAV)
Market Impact
The intricate and often circular financial web within the AI ecosystem carries profound implications for global markets. Investors face heightened challenges in performing due diligence, as traditional metrics can be obscured by inter-company commitments that may inflate perceived revenues or valuations, leading to mispricing of assets. This calls for a sophisticated, network-centric approach to analysis, demanding a deeper understanding of capital flow quality and true market demand. For credit markets, the prevalence of circular financing and concentrated supply chain dependencies, particularly when combined with high leverage, increases systemic risk. The potential for a single point of failure – whether an AI lab struggling to monetize or a hyperscaler cutting capex due to external pressures – could trigger a contagion effect, leading to wider credit spreads, refinancing difficulties, and potential defaults across interconnected entities. Regulators, already grappling with antitrust concerns, must now also consider the financial stability implications of this closed-loop system, as it could pose risks to the broader financial system. The opacity inherent in these arrangements can contribute to market volatility, making the AI sector a crucial area of focus for prudential oversight and stress testing, as the interconnectedness amplifies both boom and bust cycles, influencing capital allocation decisions globally.