This is where our provocation begins. The scale of ambition embedded in the AI capex cycle requires a fuel supply that no single financing channel has ever delivered on its own. Hyperscaler balance sheets can fund part of it. Public debt and equity can fund another part. But the buildout will increasingly require scaled pools of capital such as asset-based finance, private investment grade, infrastructure capital and financing solutions that did not exist five years ago. At the same time, the buildout is also catalyzing innovation in the financing itself. New structures are emerging in real time as capital markets adapt to these assets, revenue streams, and risks. Even as we pen this note, leaders at OpenAI, Anthropic and other frontier labs have highlighted the importance of pacing in advanced model development, a reminder that the trajectory of AI investment may evolve over time.
To be clear, at KKR we are believers and investors in this ambition and believe the U.S. has a leading role to play. But at today’s scale and pace of expected investment, we think the availability of capital in the right form may become a live constraint. That is different from saying capital will run out. More broadly, the AI buildout reflects a uniquely powerful combination of technological leadership, deep capital markets, and an ecosystem capable of funding this innovation at scale, all of which reinforce our conviction in continuing to invest behind the theme. The question is not simply whether there is enough capital, but whether there is enough of the right kind of capital, arriving in the right form, at the right price, and with the right structural protections. In our view, the risk today is twofold: that capital becomes constrained, and that the discipline to deploy it fades.
The credit story compounds from there. Diversification is harder than it looks when the same short list of counterparties sit behind the equity book, the debt book, and increasingly the infrastructure supporting both. Power, chips, cooling, land, leases and financing often route back to the same handful of economic actors.
An allocator who believes they are diversified across strategies may ultimately be more diversified across wrappers than end issuers, sectors, or economic drivers. Historical correlations can also provide a false sense of comfort because the AI capex cycle is new. Exposures that appeared uncorrelated in the past may behave differently as they become increasingly dependent on the same underlying theme. Beneath them sits another layer of market plumbing, including securities lending, margin financing, guarantees, pledged collateral, and synthetic structures that can shape how risk transmits when markets move.
Compute as an investable asset class is still emerging, but the capital pattern at play is familiar. Capital tends to follow the strongest growth themes first. Institutionalizing infrastructure, pricing, ownership, and discipline tend to follow. Those are the mechanisms that determine whether a rush becomes a durable market.
This note takes a credit and capital markets lens to the AI buildout and asks three questions in sequence. Please note, the credit lens is undoubtedly different from the equity lens. Our focus is on how this large opportunity is financed, what supports repayment, and how the structure holds when the path changes.
The Rush
Is there enough of the right capital to finance the ambition being priced today?
The Field
Where does the interconnection actually reside, and what does that mean for the diversification investors believe they own?
The Discipline
What financing and underwriting discipline and pace can turn this rush into a durable market as the trajectory evolves, whether through regulation, capital constraints, infrastructure bottlenecks, or a natural slowing in the pace of the cycle?
Our affirmative case sits in the answer to the third question. The financing tools this buildout needs are ones the credit markets have spent more than a decade developing: scaled origination, structural seniority, covenants, collateral analysis, and specialized underwriting. Just as important is the ability to draw on expertise across corporate credit, asset-based finance, private equity, and real assets as the risks increasingly cut across traditional asset-class boundaries. The front-page story is how much compute gets built. The credit story is about who gets paid if expectations change.
EXHIBIT 1: Returns Fan Out Across Credit Markets in 2026
