The Digital-Infrastructure Capital Stack: Debt, Mezzanine and Equity for AI

Introduction

A large digital-infrastructure asset, a data centre above all, is rarely funded by a single source of capital. It is funded by a stack: senior debt at the bottom, cheapest and safest, repaid first; common equity at the top, most expensive and most at risk, repaid last; and, in between, mezzanine debt and preferred equity that bridge the two. Each layer bears a different risk and earns a different return, and the way the layers are assembled, how much of each, on what terms, from which providers, determines the overall cost of capital, the return to the equity, and the resilience of the financing to adversity. Structuring the capital stack is therefore one of the central financial decisions in a digital-infrastructure investment, as consequential as the choice of asset or the securing of power, because the same asset can produce very different returns and risks depending on how it is financed.

This paper sets out the digital-infrastructure capital stack and how an investor or sponsor should structure it for Gulf data centres and digital infrastructure. It is written for the sponsor assembling the financing for a project, the equity investor judging its returns, and the lenders and mezzanine providers supplying the layers, all of whom must understand how the stack fits together and how each layer’s risk and return relate to the others. Its purpose is to make the structuring of the stack explicit: to show what each layer contributes and costs, how leverage lifts and risks the equity return, how the stack shifts as the project de-risks, and how the quality of the offtake lowers the cost of the whole. The central message is that the capital stack is not a given but a design choice, that the choice trades return against resilience through the level of leverage, and that the quality of the offtake, established in the companion papers as the driver of value and financeability, also lowers the cost of every layer and so the whole cost of capital.

The companion papers in this cluster make the strategic case for the buildout, set out the underwriting of the asset, and examine the power and land bottlenecks; this paper turns to the financing, the capital stack within which the equity and debt decisions of the others sit. Where the underwriting paper assesses the debt against the asset, this paper places that debt within the full stack of senior, mezzanine, preferred and common capital, and shows how the layers interact. The four together take an investor from the strategic case, through the asset and its constraints, to the structuring of its financing, which this paper supplies.

Results And Discussion

This section presents the framework in the order of the propositions: the layers and their cost (Proposition 1); the effect of leverage and the shift across phases (Propositions 2 and 3); the providers; and the role of offtake quality and the comparison of structures (Propositions 4 and 5).

4.0a Why the Same Asset Yields Different Returns

Before setting out the layers, it is worth establishing the proposition that motivates the whole paper: that the same asset, generating the same cash flow, can deliver very different returns to its equity depending on how it is financed. This follows directly from leverage. An asset funded entirely with equity returns to that equity exactly what the asset earns; the same asset funded partly with cheaper debt returns to its smaller equity the asset’s earnings less the debt service, spread over a smaller base, which magnifies the return when the asset earns more than the cost of the debt. So two investors buying identical assets can earn very different equity returns purely through their financing, and the structuring is therefore not a neutral wrapper but a determinant of the outcome. This is why the capital stack deserves deliberate design: it is one of the few levers, alongside the asset selection and the offtake, that materially changes the return, and unlike the asset and the offtake it is entirely within the financier’s control. The paper’s purpose is to show how to pull this lever well.

The same proposition carries a warning that runs through the analysis: that leverage magnifies losses as well as gains. The financing that lifts the equity return when the asset performs also deepens the loss when it does not, because the debt must be serviced regardless and the smaller equity bears the shortfall. So the structuring that can turn a sound asset into a strong investment can also turn it into a fragile one, and the difference lies in whether the leverage is matched to the cash flow’s certainty with a cushion, or stretched for the headline return. This dual nature of leverage, magnifying both outcomes, is why the analysis treats the structuring as a balance of return and resilience rather than a maximisation of return, and it is the single most important idea for the equity investor to hold: the financing that most flatters the base case may be the one that least survives the downside.

The Layers of the Stack

The stack is built from layers of differing risk and cost. Figure 1 sets out their share and cost.

Figure 1. The Digital-Infrastructure Capital Stack: Share and Cost

Indicative; senior debt is the largest and cheapest layer, common equity the smallest and costliest.

The figure supports Proposition 1. The stack runs from senior debt, the largest layer and the cheapest because it is safest and first to be repaid, through mezzanine debt and preferred equity, intermediate in both risk and cost, to common equity, the smallest layer in a leveraged structure but the costliest because it is last to be repaid and bears the residual risk. Each layer’s cost reflects its position: the safer and more senior the capital, the lower its required return, because it is more likely to be repaid and repaid first. The stack is thus an ordering of capital by risk and seniority, with the cost rising as the protection falls. The practical significance is that the blend of layers determines the overall cost of capital: a stack weighted toward cheap senior debt has a lower average cost than one weighted toward expensive equity, which is the basis for the leverage effect examined next. Understanding the cost and risk of each layer is the foundation of structuring the stack, because the structuring is precisely the choice of how much of each layer to use.

A practical implication of the risk-return ladder is that the layers must be priced and sized consistently, so that each provider earns a return appropriate to its actual risk and no layer is mispriced relative to its position. If the mezzanine is priced as if it were senior, its provider is undercompensated and will not supply it; if the senior debt is priced as if it were mezzanine, the structure is needlessly expensive. The sponsor assembling the stack must therefore understand the fair return for each layer given its seniority and the project’s risk, and price each accordingly, which is also how it accesses the right providers, since each provider expects the return appropriate to the layer it supplies. A coherent stack is one in which each layer is priced and sized consistently with its risk, and the sponsor’s task is to assemble such a coherent stack rather than to minimise the cost of any single layer at the expense of the others. The ladder is thus not only a menu for investors but a discipline for the sponsor, ensuring the whole stack hangs together.

The Risk-Return Ladder

The layers form a ladder of risk and return. Figure 2 shows the required return by position.

Figure 2. Required Return by Position in the Stack

Indicative; required return rises with the riskiness and subordination of the position.

Implementation Considerations

Translating the framework into practice involves securing the offtake to lower the cost of capital, setting leverage to the cash flow, planning the phased financing, and matching layers to providers.

A foundational sequencing point, parallel to the other papers in this cluster, is that the financing should be planned before the project is committed, not arranged once construction is under way, because the financing feasibility is part of the project’s feasibility. A project whose economics work only with a leverage its cash flow cannot safely bear, or whose phased financing assumes a take-out that may not be available, is not in fact feasible, and the sponsor should establish the financeability, that a sound stack can be assembled across the phases, before committing. This means testing, at the outset, that the offtake will support the necessary debt, that the development capital can be raised and can survive a delay, and that the take-out can be arranged at stabilisation. A sponsor that commits to a project and then discovers it cannot be financed soundly has the worst of both worlds; one that establishes the financeability first commits only to projects whose financing it can assemble. Planning the financing as part of the project’s feasibility, before commitment, is the sequencing discipline the framework implies, and it mirrors the power-before-commitment discipline of the bottleneck paper.

A complementary first move, alongside securing the offtake, is to establish the project’s equity base before seeking the debt, because lenders advance against a project backed by committed, credible equity and are wary of one seeking to be financed almost entirely with debt. A substantial, committed equity contribution signals the sponsor’s confidence and alignment, absorbs the first losses that protect the lenders, and is in practice a condition of raising senior debt on good terms. The sponsor should therefore assemble its equity, its own and that of its co-investors, as the foundation on which the debt is then raised, rather than treating equity as the residual to be minimised. This ordering, a credible equity base first, then the debt it supports, both reflects how lenders actually assess a financing and ensures the structure has the loss-absorbing foundation that makes it sound. The equity is not merely the most expensive layer to be minimised; it is the foundation whose adequacy determines whether the cheaper layers can be raised at all.

Securing the Offtake to Lower the Cost of Capital

Because the offtake is the master lever on the cost of capital, the first structuring move is to secure the strongest, longest offtake the project can, since it lowers the cost of every layer and supports more leverage. A sponsor that secures an investment-grade tenant on a long lease before arranging the financing will find the whole stack cheaper and more available than one that approaches lenders and equity with a speculative or weak offtake. The offtake should therefore be pursued first, as the foundation of the financing, and the financing structured around the cash flow it provides. This sequencing, offtake then financing, ensures the stack is built on the strongest possible foundation.

A practical aid to setting leverage is to size the senior debt to the stabilised, stressed cash flow rather than to the base-case or development-phase figures, because the senior debt will be serviced from the stabilised cash flow under conditions that may include stress. Sizing the senior debt to a stressed stabilised coverage ensures that the largest, most senior layer is supported even if the cash flow disappoints, which protects the whole structure, since a senior default is the most damaging outcome. The junior layers, mezzanine and equity, can then bear the variability above this protected senior base, which is appropriate because they are compensated for it. This approach, sizing the senior debt conservatively to a stressed cash flow and letting the junior layers bear the variability, produces a structure whose foundation is secure and whose risk is borne by the layers paid to bear it, which is the sound ordering of risk through the stack. It is the practical method by which the leverage is matched to the cash flow with the cushion the analysis recommends.

Setting Leverage to the Cash Flow

Concluding Comments

A digital-infrastructure asset is funded by a stack of capital, and the way the stack is assembled determines the cost of capital, the equity return and the resilience of the financing. This paper has set out the layers of the stack, the risk-return ladder, the effect of leverage, the shift across phases, the providers of each layer, and the master role of offtake quality in lowering the whole cost of capital.

The evidence and analysis support five conclusions. First, each layer earns a return commensurate with its risk and seniority. Second, leverage lifts the equity return but raises its risk, so the optimal leverage balances the two. Third, the stack shifts from equity-heavy in development to debt-heavy at stabilisation as risk resolves. Fourth, the quality of the offtake lowers the cost of every layer and so the whole cost of capital. Fifth, the investor should structure the stack to its risk appetite and the project’s phase, matching layers to the best providers.

Before turning to the limitations, it is worth situating this paper within its cluster. The strategic companion establishes the opportunity and the primacy of offtake and power; the underwriting companion assesses the debt against the asset; the bottleneck companion examines the physical constraints; and this paper structures the financing within which those decisions sit. Together they take an investor from the strategic case, through the asset, its constraints and its underwriting, to the assembly of its capital stack. This paper’s contribution is to show that the financing is a deliberate design that materially affects the return and resilience, that it shifts across the project’s phases, and that the offtake the other papers prize is also the master lever on the cost of capital. The four papers together describe a complete approach to Gulf digital-infrastructure investment, from the decision to participate to the structuring of the financing, and the capital stack is the financial keystone that the others’ decisions rest upon.