The weighted-average cost of capital is one of the most heavily used numbers in institutional finance and one of the most lightly examined. For a single-jurisdiction operating company financed in its home currency, a textbook WACC calculation will usually serve. For a cross-border deployment by a sovereign wealth fund, a state development bank, or a public pension scheme acquiring an asset two or three borders from its mandate, the same calculation is doing work it is not built to do.
The argument in this note is straightforward. The single-rate WACC template, applied unmodified to cross-border public-fund deployments, obscures the components that the investment committee actually needs to see. We set out the four components we have come to treat separately on these engagements (sovereign-rate base, jurisdiction risk premium, currency overlay, repatriation-tax drag), describe how each is sourced, and explain why a layered structure produces a more defensible appraisal even when the headline number is similar to what a single-rate calculation would have produced.
Where the conventional template strains
The standard WACC formula combines a cost of equity (typically CAPM-derived from a domestic risk-free rate, an equity risk premium and a beta) with an after-tax cost of debt, weighted by target capital structure. The formula is correct on its own terms. The difficulty is what each input is being asked to represent when the capital sits in one jurisdiction, the asset operates in a second, and the cash flows return through a third.
Three failures of representation recur in cross-border work. First, the risk-free rate is rarely as risk-free as the input field suggests: the deploying fund's own sovereign yield curve, the host country's curve, and the currency in which the cash flows are denominated may all be candidates, and they will disagree by hundreds of basis points. Second, the equity risk premium is jurisdiction-specific in ways that a global premium plus a country adjustment can fail to capture cleanly. Third, the cost of repatriating cash (withholding tax, tax-treaty mechanics, indirect transfer rules, foreign-exchange controls) is treated, where it is treated at all, as a deduction to free cash flow rather than as a component of the required return. The result is a single number that prices several distinct risks together, and a sensitivity layer that cannot pull them apart.
A layered structure: four components, sourced separately
We have found it more useful to construct the discount rate as a stack of four components, each sourced independently, each documented, and each available to the committee as a separate sensitivity. The headline WACC is then the output of the stack rather than the starting point of the conversation.
1. The sovereign-rate base
The first component is the long-dated sovereign yield in the currency the asset's cash flows are denominated in. For a Gulf sovereign wealth fund acquiring an Australian infrastructure asset, this is the AUD government curve at the asset's effective duration, not the US Treasury curve and not the deploying fund's home-jurisdiction curve. The choice is structural: the asset throws off Australian dollars, and the opportunity cost of holding those flows is the Australian sovereign yield. Where the asset has a significant non-AUD revenue line (a US-dollar export contract, for instance) we split the cash flow stream and apply the appropriate base to each tranche.
2. Jurisdiction-specific risk premium
The second component is the additional return required for holding equity exposure in the host jurisdiction relative to the sovereign curve we have just used. This is not a country risk score lifted from a published table. It is built from observable inputs: the spread between the host sovereign and a developed-market benchmark of comparable maturity, the equity-bond spread within the host market, and the sector-specific risk premium where the asset sits in a regulated sector with idiosyncratic policy exposure. On energy and infrastructure mandates, we treat regulatory revenue mechanics as a discrete sub-component rather than folding them into beta.
3. Currency overlay
The third component addresses the mismatch between the currency the asset earns in and the currency the deploying fund reports in. If the fund hedges the cash-flow stream to its reporting currency, the currency component should appear as the hedging cost (the forward differential, plus the cost of rolling the hedge over the appraisal horizon). If the fund does not hedge, the component should appear as the option-adjusted expected drag on long-horizon flows, with the volatility input drawn from realised cross-rate behaviour rather than implied at a single point in time. Either approach is defensible; what is not defensible is omitting the component because the model is already in the asset's currency.
4. Repatriation-tax drag
The fourth component captures the cost of moving cash from the asset back to the deploying fund's balance sheet. Withholding tax on dividends and interest, the treaty regime between the asset jurisdiction and the holding-company jurisdiction, indirect transfer rules, and any specific exemptions available to sovereign or public-fund vehicles will combine to produce an effective drag. Sovereign immunity from withholding tax exists in some jurisdictions and not in others, applies to some income types and not others, and is sometimes contingent on a written ruling that has to be applied for. The drag belongs in the required return, not buried in the free-cash-flow line, because it is structural to the deployment rather than operational to the asset.
What the layered structure changes in practice
The argument for this approach is not that it produces a different WACC. On a number of engagements the layered rate has come out within 25 basis points of a competently constructed single-rate calculation. The argument is that the layered rate produces a different conversation.
When a committee sees the discount rate decomposed into four components, the question shifts from "do we believe the WACC" to "which of these four components are we most exposed to, and what would change them". That is the question the committee is being paid to answer.
Three consequences follow. The first is that sensitivity work becomes more useful: stressing the host-jurisdiction premium by 200 basis points, while leaving the sovereign base unchanged, isolates a question (what if host policy deteriorates) that the committee can engage with. The second is that the repatriation drag becomes negotiable: where the structure can be modified to reduce withholding exposure, the financial value of doing so is visible in the rate rather than hidden in the cash flows. The third is that disagreement with internal teams, external advisers or co-investors becomes traceable: the committee can see whether a counterparty's lower headline rate reflects a different view on the sovereign base, the jurisdiction premium, the currency, or the tax drag, and respond to the specific point rather than to the aggregate.
Common objections and where they have force
Two objections to the layered approach are worth addressing directly. The first is that a layered structure double-counts: the jurisdiction premium and the currency component, for instance, may overlap if the currency reflects political risk that has already been priced in the sovereign spread. This is a real concern. The discipline we apply is to define each component against an observable input and to require that no component be moving on the same news as another. Where overlap is structural (in some emerging-market deployments it is), we collapse the two components explicitly and document the collapse.
The second objection is that the layered structure is not standard and will not match the rate a fund's auditors or external valuers use. This is correct: the layered rate is the input to the investment decision, not the input to the financial statement valuation. For mark-to-market or impairment work, the conventional discount rate may continue to apply. We treat them as separate exercises with separate purposes, and we are explicit with clients about which is which.
Closing position
Cross-border public-fund deployments are not the place to economise on the appraisal architecture. The risk that a single-rate WACC obscures is not numerical imprecision; it is that the committee approves the deployment without seeing which component of the required return is doing the work. Layering the rate, with each component sourced separately and documented, is the approach we have come to rely on to put the relevant choices in front of the people making them.
Cross-border deployment work is one of the engagements where the analytical architecture matters more than the headline output. We are willing to discuss the layered approach in the context of a specific mandate, under the standard confidentiality regime.