After Construction, Who Gets the Gain? Refinancing Gulf Utility Assets

1 Define the refinancing decision and the contribution

An operating utility's refinancing proposal should identify the money that changes hands, the remaining obligations and the conditions under which shareholders may receive the resulting benefit. This paper examines a narrower decision: how a project company and its public purchaser can evaluate repricing, maturity extension and additional borrowing after construction, using one consistent cash-flow comparison. The proposed method combines contractual gain measurement with a completion funds-flow statement, coverage tests and a schedule of remaining debt. Each output answers a different approval question.

The analytical contribution is the reconciliation between these outputs. A projected gain can be positive while completion still requires a shareholder contribution. A distribution can be funded at closing while the revised structure fails a chosen operating stress. A tariff reduction can deliver a contractual present value while reducing the cash that supports debt service. The worked example calculates those consequences using declared assumptions. It supplies a reproducible comparison and identifies proposals requiring redesign; it supplies no observation about actual refinancing returns or market pricing.

The contract determines entitlement. The original financial model, approved amendments, financing documents and public-consent provisions determine which transactions qualify, which distributions enter the calculation and which gain is shareable. A general public-private-partnership methodology cannot establish rights in a particular power and water purchase agreement. Where the executed agreement is unavailable, a prospective adviser can outline the questions and calculate conditional cases. The adviser cannot represent an assumed sharing percentage, discount rate or consent power as an existing contractual right.

The worked case therefore uses a fictional utility and an expressly assumed sharing rule. Its purpose is to reveal economic dependencies before a negotiation. Three structures use the same operating assumptions and reference date. The model retains costs, reserve releases and future repayments in the shareholder comparison. Subsequent tests distinguish contractual present value from liquidity, covenant eligibility and public exposure. This structure permits a reader to reproduce the result and change the assumptions that drive a decision.

2 The public evidence and its contractual boundaries

The source base combines international contractual guidance, historical UK public-finance practice and corporate disclosures about Gulf utility assets. These categories answer different questions. Guidance supplies a framework for examining a transaction. A corporate disclosure establishes the transaction or accounting treatment described by its issuer. A project-specific conclusion about consent, sharing or distribution requires the applicable documents. The analysis below therefore treats public material as bounded evidence and identifies the additional records needed for an investment decision. Research access was completed on 9 October 2026.

The World Bank's 2019 Guidance on PPP Contractual Provisions, chapter 5, printed pages 93-100, distinguishes rescue, anticipated refinancing and value-realisation transactions. Sections 5.2.3-5.2.6 discuss consent, access to financing information, sharing alternatives and payment methods. The guidance recognises that contractual arrangements differ and offers equal or staged sharing as options. It supplies no general percentage for Gulf utility projects. Its sample drafting is a reference for contract analysis, subject to project and jurisdiction-specific adaptation. [1]

EPEC's guide, published on 16 April 2026, addresses refinancing on printed pages 197-198. It identifies debt pricing, maturity, gearing, reserves and guarantees as possible sources of change. It recommends addressing gain calculation, allocation, payment form, discount rates and termination consequences in the contract. Its discussion includes shareholder-gain valuation using net present value and payment through a lump sum or reduced future availability payments. The guide provides a public-authority perspective within European PPP guidance. [2]

For this paper's proposed diligence method, each public claim should be entered with its issuer, date, project, instrument and evidentiary scope. An announcement of completed debt replacement belongs in the transaction record. A disclosed accounting entry belongs in the accounting reconciliation. Neither entry should automatically populate the consent register or the shareholder-distribution schedule. Leaving an unsupported field unresolved preserves the distinction between the available evidence and the decision still to be made.

3 Public-authority gain calculation and consent

HM Treasury's calculation note, issued within the historical SoPC 4 framework, requires pre- and post-refinancing models with identical assumptions apart from the refinancing effects. Sections 1.2-1.4 address actual investments, distributions, updated forecasts, the NPV comparison and threshold-return adjustments. Section 1.6 explains that final costs and rates require a post-closing calculation although consent and the calculation basis precede closing. The inspected PDF carries no explicit issue date; its government collection was published on 25 July 2013 and updated on 7 December 2018. [3]

The December 2012 Standardisation of PF 2 Contracts, chapter 28, is a historical UK document marked draft. Section 28.3 addresses consent and service, stability and termination-liability concerns. Section 28.5 describes positive and negative changes in projected distributions. Additional borrowing can increase an initial distribution and reduce later distributions through debt service. Section 28.6 addresses information access and review costs. These provisions illuminate questions for diligence; their applicability to any Gulf contract remains unestablished. [4]

HMRC's BIM 64370 describes UK PFI benefit-sharing through reduced unitary payments, lump sums, additional services or combinations. It also discusses purchaser permission and fact-dependent tax treatment. The manual was published on 22 November 2013 and displayed an update date of 4 August 2026 when inspected; the latter is a manual update date. This material supplies a UK administrative context and establishes no Gulf tax result or procurer entitlement. [5]

The proposed investment review should maintain a consent matrix alongside the financial model. Each row should identify the proposed change, affected document, required decision-maker, supporting clause, submission requirements and unresolved conditions. A reduction in financing cost, a maturity extension, a reserve release and a change in termination exposure warrant separate entries. This proposed matrix is an analytical control; the paper has no private Gulf agreement from which to populate actual approval requirements.

The same discipline applies to the comparison model. Operating improvements, revised availability assumptions and updated maintenance costs should be identified separately from the financing intervention. The reviewer should be able to trace every difference between the two cases. Any assumed public share needs a visible label, a stated calculation base and sensitivity testing. An unexplained percentage would conceal a material commercial choice.

4 Rabigh 3: completion evidence and accounting timing

ACWA Power's public Rabigh 3 announcement reports completed refinancing using a long-term senior secured project bond and replacement of the previous debt facility. It identifies participating institutions and acknowledges SWPC's cooperation. The inspected page displays no publication date, so it is treated as undated and accessed on 9 October 2026. It does not disclose the applicable consent provision, gain-sharing calculation, transaction costs or shareholder distribution. Its statements remain attributed issuer disclosures. [6]

The company's report for the three and nine months ended 30 September 2025 provides a separate observation. On page 8, section 2.3.1 describes lending arrangements in their final execution stages and approximately SAR 1.7 billion of loans to be refinanced. It reports September 2025 recycling of SAR 269.5 million from the cash-flow hedge reserve into profit or loss, including SAR 188.65 million attributable to ACWA Power, because the previously hedged risk was no longer expected on that loan. [7]

The case supports a proposed reconciliation with distinct columns for accounting recognition, settlement cash, financing proceeds and distributions. Dates should be attached to each event. The undated completion page leaves the precise completion date unresolved in this source set. A reader can assess the disclosed accounting event without assigning it an unsupported closing date. Determining sponsor value additionally requires the cash-flow schedule, settlement statements and applicable restrictions. This paper does not estimate that value from the reported accounting profit.

5 Barka: debt tenor and the unanswered post-contract question

The undated Barka announcement reports refinancing of USD 190.39 million and separate expansion financing of USD 43.1 million. ACWA states that the local-currency refinancing tenor could extend up to six years beyond the PWPA and the expansion facility up to 6.5 years; it also reports removal of a cash-sweep covenant. The inspected announcement has no displayed date. It is used as an undated issuer disclosure, accessed on 9 October 2026, without assigning current status or a subsequently realised outcome. [8]

For the proposed model, contracted and post-contract periods should have separate assumptions and outputs. A debt schedule extending beyond offtake expiry calls for an explicit residual-revenue scenario, operating-cost schedule and downside case. Renewal, alternative sales and asset realisation remain unresolved possibilities until supported. The public announcement provides no basis for selecting their probabilities. Accordingly, the case establishes a diligence question about the source of repayment after contract expiry. The paper's wider conclusions remain conditional on verified contract terms, financing documents and project performance.

6 Establish a controlled counterfactual and financial boundary

The counterfactual describes continuation of the existing financing at the refinancing date. It includes the remaining principal, debt service, reserve balance and future shareholder cash flows. Historical distributions are relevant to establishing the approved starting position and any contractual catch-up rule. They are excluded from this example's prospective gain because the comparison begins after those payments. Reintroducing a historical distribution into one case would create an artificial difference.

All financial values in the worked case are hypothetical USD millions. The fictional reference date is 1 January 2027. Fifteen annual operating periods remain, ending in 2041; payments occur at year end. Opening senior debt is 600, with a fixed assumed all-in rate of 6 percent and ten remaining annual level payments. A funded debt-service reserve of 45 is available at the reference date. It remains held in the counterfactual and is released at the end of year ten. The illustrative agreement permits distributions above an assumed legacy coverage threshold of 1.10x. That threshold describes this fictional agreement only.

Annual cash available for debt service, or CFADS, is 95. It is already after operating costs, project tax and planned lifecycle expenditure. No tax saving from the new financing is included. Output, tariff, costs and lifecycle spending are held constant across financing alternatives before a separate public-tariff-sharing option is introduced. The model assumes no inflation, currency mismatch, floating-rate reset, revenue growth, default, refinancing beyond the modelled transaction or terminal asset sale. These omissions define a transparent instructional case. An actual utility model would need the relevant additional schedules.

The refinancing alternatives use an assumed fixed all-in rate of 4.5 percent. Reprice retains debt of 600 and ten payments. Extend retains debt of 600 and uses thirteen payments. Cash out borrows 680 with thirteen payments. Assumed fees are 8, 10 and 12 respectively; each incurs hedge settlement cash of 3. The replacement reserve equals half of its annual debt-service payment and is released when that debt is repaid. The old 45 reserve is assumed legally releasable at closing. The model includes its release once and records the new reserve funding separately.

Shareable gain is calculated at an assumed 10 percent shareholder discount rate. The public share is assumed to be 50 percent of positive gain. New facilities are assessed against a proposed 1.25x distribution threshold. The stricter threshold is an analytical condition for the refinancing proposals. It is not substituted retrospectively into the original financing. Separate stress results show the effects of 10, 20 and 40 percent reductions in CFADS, including cases where the public share is delivered by an annual tariff reduction.

Every assumption above is intended to be replaced with transaction evidence before an actual approval. The opening balance requires a payoff confirmation; the reserve requires bank evidence and release conditions; financing economics require signed terms; CFADS requires a dated operating forecast; sharing requires the executed contractual mechanism. A table of assumptions records evidence status and units. A reviewer should reject a model input labelled verified when its only basis is the author's numerical example.

Table 1. Declared assumptions and documentary replacement requirements

Entire case is author-created. No input is represented as a market quote or an executed project term.

7 Derive debt service and retain all remaining payments

The level annual payment on debt D, annual rate r and n periods is D multiplied by r, divided by one minus the reciprocal of (1+r) raised to n. Interest in each year equals opening principal multiplied by r. Principal repayment equals the annual payment less interest. Closing principal equals opening principal less repayment. The model checks that all principal repaid equals original borrowing and that the final balance is zero, subject to negligible floating-point rounding.

The counterfactual annual payment is 81.52. Base coverage is 95 divided by that amount, giving 1.165x. It passes the fictional legacy 1.10x distribution threshold. Counterfactual available shareholder cash in years one through nine is 13.48 per year. Year ten also releases the 45 reserve. Years eleven through fifteen have no senior debt service; the assumed 95 CFADS is available subject to obligations already deducted from that cash measure. The full fifteen-year comparison retains those later cash flows.

Reprice's annual payment is 75.83 and base coverage is 1.253x. Its reserve is 37.91. Extend's payment is 61.97, reserve 30.98 and coverage 1.533x. Cash out's payment is 70.23, reserve 35.11 and coverage 1.353x. The tables use displayed rounding; model calculations retain full precision. Annual interest and principal schedules show how each structure produces these amounts. A single unexplained coverage ratio would leave the reader unable to reproduce the decision.

Tenor extension changes the timing of principal repayment. Extend's annual payment falls relative to the counterfactual, but debt remains outstanding in years eleven through thirteen. The sponsor consequently receives greater early available cash and lower cash during those three later periods. Its replacement reserve is released three years later than the original reserve. A gain calculation that stops at the original debt maturity omits those consequences and exaggerates the benefit.

Additional borrowing increases completion cash and subsequent debt service together. Cash out cannot be modelled by adding 80 to sponsor proceeds while retaining Extend's 600 debt schedule. Its larger reserve and payment also change future shareholder cash. The example uses the full 680 schedule. Any repayment grace period, balloon or refinancing assumption would need its own explicit treatment; none is implied by these level-payment results.

Table 2. Financing terms and derived debt service

Hypothetical USD millions except rate, years and coverage. All calculations retain full precision.

8 Reconcile closing proceeds and shareholder present value

Available cash at closing before public sharing equals new debt, plus released old reserve, less old-debt repayment, fees, hedge settlement and the new reserve. Reprice produces negative 3.91; Extend produces positive 1.02; Cash out produces positive 74.89. A negative balance represents required sponsor funding before any public payment. It does not represent a sponsor distribution or a net refinancing loss by itself.

The gain combines that closing amount with the present value of subsequent changes in available shareholder cash. Each year's change includes the difference in debt service and any reserve-release timing. At the assumed 10 percent discount rate, Reprice's gain is 28.34, Extend's 53.39 and Cash out's 69.77. These values describe unrestricted available-cash economics before implementing an alternative public-payment route or any adverse-case covenant trapping. They are conditional model results.

The distinction matters at closing. Reprice's public share is 14.17; paying it immediately requires a total sponsor contribution of 18.08. Extend's share is 26.70; its immediate payment requires 25.68 of sponsor funding. Cash out's share is 34.88 and leaves a 40.00 sponsor distribution. The completion statement shows sources equal uses in all three cases by including the required contributions explicitly. It contains no balancing residual presented as unidentified financing.

The gain calculation deducts fees and hedge settlement through completion cash. They must not then be deducted a second time from the resulting NPV. The old reserve's future release remains in the counterfactual even though the refinancing releases that reserve at closing. The replacement reserve's eventual release remains in the revised case. Counting reserve release at closing while removing its later counterfactual receipt overstates the timing benefit.

A comparison based solely on discounted debt-service savings gives different results. After fees and hedge settlement, that proxy gives 23.98 for Reprice,47.75 for Extend and negative 12.94 for Cash out. The full shareholder comparison adds net new proceeds and reserve timing, yielding the gains above. Cash out supplies the clearest demonstration: a debt-cost-only proxy omits the incremental borrowing that funds the immediate distribution. The proxy cannot substitute for the integrated shareholder schedule when capital structure changes.

Table 3. Completion sources and uses, including immediate public payment

Sources are the first three rows; uses are the remaining six. Negative sponsor closing cash is shown as a contribution, never as a negative distribution.

Author calculations; all inputs hypothetical. Gain is before public sharing and covenant implementation; closing cash follows immediate public payment.

Table 4. Reconcile the debt-cost proxy to the available-cash valuation

Rows one and two reconcile to row three. Values use the assumed 10% discount rate. NPV does not assert an approved distribution.

9 Test discount-rate dependence and staged sharing

The assumed discount rate governs the relative value of accelerated and deferred cash. At 5 percent, the gains are 35.70 for Reprice,37.23 for Extend and 35.68 for Cash out. At 10 percent, the ranking becomes Cash out, Extend, Reprice. At 15 percent, the corresponding gains are 22.91, 58.10 and 86.51. These results follow from the same cash flows. They do not establish which rate the parties should adopt.

The ten-year repricing mostly creates savings over the original payment period. Extending debt retains borrowing later and transfers available shareholder cash towards earlier dates. Adding debt increases that transfer. A higher discount rate assigns less present value to the subsequent reduction in distributions. The calculation therefore makes the discount-rate agreement consequential: the preferred structure can change even when lender rates and operating assumptions remain constant.

One proposed sharing alternative applies 25 percent to the first 20 of gain and 60 percent above 20. It is a fictional negotiation illustration, separate from the equal-sharing case. For a gain G above 20, its public amount is 5 plus 0.60 multiplied by (G minus 20). Marginal bands avoid an immediate jump in the public amount at the threshold. Reprice, Extend and Cash out produce public amounts 10.00, 25.03 and 34.86 respectively after rounding. Their associated sponsor closing cash must be recalculated rather than borrowed from the equal-share completion statements.

Source-specific sharing would require a consistent allocation of interactions. Lower pricing changes the annual payment and the replacement reserve; higher debt changes proceeds, principal, interest and reserves. Separately assigning the entire reserve release to one category while also retaining it in another category's shareholder model counts value twice. A transparent alternative is to run a specified sequence of incremental financing changes and show the attribution's dependence on ordering. The parties would still need to agree that sequence and reconcile the allocated components to total gain.

The paper uses equal sharing for the principal case to keep the financing comparisons traceable. It does not claim that equal sharing is a Gulf default or that a staged rule is superior. A contractual calculation may contain equity-return catch-up, exemptions, different discount rates or costs excluded from sharing. Each would require an additional schedule. A gain generated by assumptions outside the agreed mechanism remains outside a contractual claim until its treatment is resolved.

Author calculations from the same hypothetical cash flows. Sensitivity rates are assumptions, not recommended contractual rates.

Table 5. Equal and marginal-band sharing alternatives

Both rules are author-created negotiating illustrations. Thresholds apply as marginal bands. Neither rule is asserted to govern a Gulf agreement.

10 Compare tariff delivery with cash payment

An assumed constant annual tariff reduction can deliver the public amount over fifteen year-end payments. Dividing the public share by the fifteen-year 10 percent annuity factor gives reductions of 1.86 for Reprice,3.51 for Extend and 4.59 for Cash out. Their present values equal the corresponding upfront public shares at the same rate and dates. The tariff branch replaces the upfront payment; retaining both would charge the public benefit twice.

The tariff reduction also reduces operating cash available for debt service. Reprice's base coverage falls from 1.253x to 1.228x and fails the proposed new 1.25x distribution condition. That option is rejected in its submitted form. Extend's base coverage becomes 1.476x and Cash out's 1.287x. Each still requires stress testing before a distribution decision. A present-value equality supplies no conclusion about the sufficiency of coverage.

Budget valuation can use a different rate from the contract's sharing calculation. At a hypothetical 6 percent public discount rate, the three tariff streams have values 18.09, 34.09 and 44.54 respectively. These exceed the contractual public-share amounts valued at 10 percent. A report should identify which valuation answers which question. It should not describe the streams as universally equivalent to the closing payment merely because one calculation uses a common rate.

Missed payments, performance deductions and tariff indexation can change delivery. The constant annual reduction here is nominal and fixed. No collection default or contingent relief is assumed. If an actual amendment reduces only a variable tariff component, its value will depend on the corresponding operating quantities. The tariff schedule would need to carry those quantities, deductions and dates into both gain valuation and debt coverage. The example's annuity calculation cannot certify a differently structured amendment.

Table 6. Public benefit through a constant fifteen-year tariff reduction

The tariff branch replaces the immediate public payment. Hypothetical full delivery is assumed. Passing base coverage supplies no adverse-case approval.

11 Derive stress outcomes and the borrowing constraint

Under a 20 percent CFADS reduction, annual cash falls from 95 to 76. Without tariff sharing, Reprice's coverage is 1.002x, Extend's 1.226x and Cash out's 1.082x. All fail the proposed 1.25x distribution threshold. Cash after debt service remains positive in each, albeit narrowly for Reprice. The model therefore identifies a lock-up condition before an immediate annual debt-service cash deficit. A facility's actual covenant, reserve and cure provisions would determine the consequence.

Retaining fixed tariff reductions strengthens the constraint. Coverage under the same stress becomes 0.978x,1.170x and 1.017x respectively. Reprice has an annual pre-reserve cash shortfall of 1.69. Extend and Cash out retain positive annual cash after debt and tariff, but fail the proposed distribution condition. A model that reports the pre-tariff ratios alongside a tariff-sharing proposal leaves that proposal's principal cash consequence untested.

For an illustrative committee requiring 1.25x coverage at CFADS 76 before tariff sharing, allowable annual debt service is 60.80. Discounting thirteen payments of 60.80 at the hypothetical 4.5 percent lender rate gives maximum opening borrowing of 588.72. Both Extend 600 and Cash out 680 exceed this limit. The ten-year structure supports 481.09. These are debt-sizing calculations under declared assumptions; they are not lender offers, statutory limits or actual project debt capacity.

The 588.72 constraint excludes public tariff sharing. After a tariff reduction, allowable debt service is lower again. The debt amount, gain, public share and tariff reduction move together, requiring a simultaneous recalculation. Subtracting the original tariff reduction from the cash while resizing debt using the old gain does not solve that interdependence. An actual restructuring should show the revised loan, closing contribution, reserve, share and stresses together before seeking approval.

At a 40 percent CFADS reduction, cash is 57. Extend's annual payment 61.97 produces a 4.97 shortfall before tariff sharing. Its assumed 30.98 reserve can supply a permitted draw, but the annual model has no evidence of a committed replenishment source. A reserve balance addresses liquidity for a limited period. A repeated shortfall consumes that balance. The sponsor's ability and obligation to supply capital require separate verification.

Table 7. Combined operating and tariff-sharing stress

Hypothetical USD millions; ratios in x. Cash values are pre-reserve and pre-support. The proposed distribution condition is 1.25x; cash below zero is a liquidity shortfall.

Author calculations, hypothetical assumptions. The dashed line is the proposed 1.25x distribution threshold; it is not a verified loan covenant.

12 Carry the debt tail into the public exposure comparison

At year ten, the counterfactual and Reprice are repaid. Extend still owes 170.34 and Cash out 193.05. At year seven, their respective balances are 319.61 and 362.22, compared with 217.91 under the counterfactual. These debt schedules identify how maturity extension changes the balance exposed to later events. The full model supplies intermediate balances instead of relying on the opening loan amount alone.

For illustration, a gross recognised-debt termination rule is assumed to reimburse remaining senior principal. Under that fictional rule, the balances above would determine the principal component of exposure. They are not estimates of actual termination payments. Such payments could include deductions, break amounts, cash, insurance, asset value or amounts excluded from compensation. Default cause and the executed agreement would govern their treatment. No undisclosed Gulf contract is assigned this assumed rule.

The proposed committee should compare the public cash receipt with the changed exposure schedule, using agreed scenarios and measurement dates. A public share paid at completion does not automatically compensate an increased late-year payment obligation. An amendment excluding cash-out borrowing from recognised debt would create a different lender risk. The financing and consent documents would need to resolve that risk rather than leave an unapproved difference between lender security expectations and public compensation.

Two years remain between the thirteen-year debt maturity and the fictional concession expiry. Planned lifecycle spending is already within CFADS, but the model does not quantify a separate handback remediation obligation. An engineer's condition assessment and expenditure schedule would be needed before representing those two years as sufficient protection. The next analysis should reconcile funded maintenance, residual life, reserve releases and final permitted distributions to the actual handback standard.

Author calculations; debt balances follow scheduled year-end repayment. No actual termination compensation is estimated.

Table 8. Principal used in the hypothetical recognised-debt comparison

USD millions. Year zero precedes the first payment; subsequent balances are after year-end service. A gross recognised-principal rule is hypothetical.

13 Negotiate a public benefit that the closing can fund

The hypothetical Extend case creates a positive gain of USD 53.39 million, measured at the assumed ten percent discount rate over fifteen years before implementation of covenant restrictions. Its public share is USD 26.70 million under the assumed equal split. Available closing cash before that payment is only USD 1.02 million. Most of the measured benefit arrives through changed future cash flows. A consent paper therefore needs a delivery proposal with an identified funding source. Describing the public share as payable at completion leaves a USD 25.68 million sponsor contribution to be committed and paid.

One proposed arrangement is a funded sponsor contribution alongside the refinancing. The sponsor would compare that immediate payment with the retained future benefits and its own liquidity constraints. The approval record should specify the contribution instrument, account, funding deadline and consequences of non-payment. Any shareholder loan used to supply the money would require a new model run covering its interest and repayment. The existing example assumes the contribution as an equity cash outflow; adding a repayment obligation changes the economics and could change the shareable gain.

A tariff reduction offers a different delivery schedule. In the hypothetical Extend calculation, an annual reduction of USD 3.51 million over fifteen year-end payments has the same present value as the public share. That equality depends on the assumed discount rate, payment dates and full delivery of the reductions. It does not establish equal liquidity, credit exposure or budget treatment. The proposed amendment would identify the affected payment component, its commencement and final dates, and the treatment of missed, disputed or revised payments. Any indexation or accelerated settlement mechanism would require its own calculation.

Deferral also consumes operating headroom. Extend's base coverage becomes 1.48x after the tariff reduction, using the selected refinancing debt schedule. Reprice falls to 1.23x, below the author's assumed new 1.25x distribution threshold. The Reprice tariff option is rejected in its submitted form. Its unrestricted cash-flow NPV cannot be described as an actually distributable gain. A blended payment would need a fresh calculation of the upfront amount, remaining tariff reduction and resulting coverage. The analysis supplies no approved blend and no assumed lender waiver. The original financing has a separately assumed 1.10x threshold; its base cash flows pass that legacy condition.

14 Record a distinct consent outcome for each structure

For Reprice, the proposed committee disposition is conditional further negotiation. The structure requires USD 18.08 million of sponsor cash to meet the modelled completion uses, including the public payment. Its base coverage passes the assumed distribution test by a narrow margin. A ten percent CFADS loss produces a lock-up in the stated stress calculation. Proposed consent conditions would therefore include evidence of the sponsor contribution, confirmation of the selected public-payment route and a funded liquidity plan. Substituting the full tariff alternative would reopen the coverage decision because that version fails the base distribution threshold.

For Extend, the proposed disposition is negotiation of the public-delivery and downside package before final approval. The longer amortisation produces greater base headroom, although the twenty percent CFADS reduction still falls below the assumed distribution threshold. The closing contribution and the later debt tail require separate decisions. Conditions would identify the accepted contribution or tariff schedule, the permitted maturity, the debt recognised in the hypothetical termination comparison and the treatment of required lifecycle spending. A reviewer should be able to find the precise document and model version satisfying each condition.

For Cash out, the proposed disposition is to withhold approval of the submitted distribution pending redesign. Its modelled completion cash permits a USD 40.00 million sponsor distribution after the public payment. Its coverage falls below the assumed distribution threshold under the ten percent CFADS stress. A committee choosing to require continued distribution eligibility under the twenty percent stress would have to reduce debt or change the financing package. The model's USD 588.72 million debt limit uses the thirteen-year structure and applies before any tariff-sharing deduction. Coverage after tariff sharing is an additional constraint requiring recalculation.

These proposed outcomes distinguish conditions for negotiation from completed approvals. None represents an actual lender or public decision. If the committee instead accepts distributions being suspended during a specified stress, it should state that policy choice and test how retained cash supports recovery. The approval record should also state whether the sponsor must replenish reserves or supply new equity. A higher initial distribution cannot serve as evidence that future support has been committed. Any redesigned borrowing changes the gain, reserve requirement, public share and funds flow together.

15 Translate annual coverage into a payment-delay liquidity test

The annual model uses cash available for debt service of USD 95.00 million after operating costs, tax and planned lifecycle expenditure. It does not specify invoice dates, receivable collection or intra-year debt-payment dates. Consequently, its annual ratios cannot establish that cash is available on every payment date. The next lender test would construct a dated receipts-and-payments schedule from the proposed commercial arrangements. It would preserve annual totals in the undelayed case and move only the selected receipts in the delay case.

That schedule should track unrestricted operating cash, the debt-service reserve and any committed working-capital facility separately. An invoice whose receipt has moved remains a receivable until collected. The reviewer would identify the first date unrestricted cash is insufficient, the permitted reserve draw, the facility availability and the replenishment requirement. If a facility is proposed, its fees, interest, maturity and draw conditions become explicit inputs. The current model contains no such facility and provides no basis for assigning it a zero cost or certain availability.

The modelled reserve equals half of annual debt service. That convention says nothing about the actual timing of collections or permitted reserve uses. Under the annual forty percent CFADS stress, Extend has a debt-service shortfall of USD 4.97 million before any tariff reduction. An assumed opening reserve may supply liquidity for a permitted draw, but a repeated shortfall needs a replenishment or restructuring response. The approval should identify who bears that response and whether shareholder payments remain blocked. The gain calculation should be revisited if the chosen response changes financing costs or distribution dates.

With the fixed tariff reductions retained, the twenty percent CFADS stress produces coverage of 0.98x for Reprice, 1.17x for Extend and 1.02x for Cash out. All fail the proposed distribution threshold. Reprice also has an annual cash shortfall after debt service and the tariff reduction. The approval file should distinguish those outcomes and avoid treating the positive unrestricted NPV as permission to pay shareholders. Any cash trapping and later release would need to enter a covenant-aware distribution schedule before an executable shareholder valuation is claimed.

16 Keep lifecycle and handback funding inside the consent perimeter

Planned lifecycle spending is already deducted in the hypothetical CFADS assumption. A reviewer should request the underlying expenditure schedule before adapting the model to a transaction. Each item needs an asset, planned date, amount and treatment in CFADS or a separately funded account. This mapping matters when a reserve is released: the released balance could otherwise be presented as excess cash while its original maintenance purpose remains unfunded. A proposed consent condition would require reconciliation between the engineer's expenditure plan and the financial model before any distribution.

An additional overhaul or earlier replacement would enter a revised case as an incremental dated payment. The team should show the resulting coverage, cash balance and support requirement. It should also identify expenditure that is technically deferrable, the consequences of deferral and the person authorised to accept those consequences. The existing constant-CFADS illustration establishes none of those engineering conclusions. A maintenance reduction cannot be used to improve the financing case without a separately evidenced change in the underlying work requirement.

Extend and Cash out amortise through year thirteen within the assumed fifteen-year project horizon. Their remaining debt at year ten is USD 170.34 million and USD 193.05 million respectively. The model's hypothetical debt-recognition rule makes that remaining balance visible as a public-exposure comparison. Actual termination compensation would need its own verified contractual calculation. Proposed consent conditions should identify approved debt recognition, final maturity, inspections and any handback funding restrictions. The two years after repayment provide a timing interval in the illustration; their adequacy for handback cannot be established without asset-condition and expenditure evidence.

17 Execute the selected funds flow and amend dependent schedules

For an Extend closing with an immediate public payment, modelled sources comprise new debt, release of the old reserve and the sponsor contribution. Uses comprise old-debt repayment, fees, hedge settlement, funding of the new reserve and the public payment. No sponsor distribution remains. The completion statement should carry those exact categories and identify the paying account, receiving account, amount, value date and release condition. A funding confirmation should relate to cleared funds or an executable draw, with its status stated. A signed commitment alone does not record a completed transfer.

The hypothetical arithmetic assumes that the old reserve is fully available at closing. In a transaction, the completion team would first verify its balance, release conditions and any claims against it. A lower available balance increases the funding gap unless another use also changes. The same discipline applies to the hedge settlement and lender payoff: their final amounts and value dates must replace the provisional model inputs. The proposed closing control would prohibit release of a sponsor distribution while a required public payment or reserve deposit remains unfunded.

A last-minute change requires an explicit revision. An increased fee reduces the calculated gain; a changed public share then affects the sponsor cash requirement. Moving completion alters discount timing and could change accrued debt settlement. The revised statement should therefore reconcile both cash and present value, with the authorised parties accepting the changed result. Rounding differences should be resolved in the payment instructions while calculations retain full precision. A residual balancing entry labelled other uses would leave the transaction's allocation incomplete.

After closing, the operating records should carry the selected debt schedule, reserve requirements, public-delivery schedule and distribution conditions. A tariff amendment requires corresponding changes to invoicing and any dependent forecast. The repayment schedule should agree with lender records; reserve balances should agree with bank evidence. A proposed short post-completion reconciliation would compare expected and actual transfers and identify outstanding documents or amounts. The signed completion model would then preserve the calculation actually implemented, alongside the original counterfactual and approval history.

18 Make the evidence chain testable by an independent reviewer

Each transaction input should connect to a specific source passage or record. The opening debt balance would connect to a lender statement and payoff confirmation; reserve cash to account evidence and release terms; new debt service to the signed facility economics; hedge settlement to the dated settlement statement. Public sharing and delivery would connect to the agreed contractual mechanism. The hypothetical inputs in this paper instead connect to the declared assumptions. They carry no documentary confirmation of a real asset, market price or payment right.

The reviewer should rebuild the debt schedules independently and reconcile total principal to opening borrowing, final balances to repayment and reserve releases to the selected maturity. The shareholder comparison should include closing contributions as negative cash flows and closing distributions as positive flows. It should retain the old reserve release in the counterfactual and the new reserve release in the refinancing case. Removing either future release should change the result; this deliberate test can expose a missing or duplicated reserve entry. Fees already paid at closing must not be deducted again from the computed gain.

A second set of tests should address decisions. Replacing lump-sum sharing with the calculated tariff schedule should preserve the assumed public present value while changing operating coverage. Raising debt should alter proceeds, debt service, reserves and future balances together. Applying the twenty percent stress should suspend distributions where the stated threshold fails, without automatically labelling positive cash after debt service as a payment default. A dated delay test should move cash availability even when annual revenue is unchanged. These checks connect model behaviour to specific approval questions.

The final review opinion should separate reproduced arithmetic, verified source inputs and unresolved assumptions. An unexplained discrepancy would remain open with its financial consequence stated. The committee could then approve a bounded negotiation, request a redesigned structure or stop the proposed distribution. Completion would require the conditions attached to that chosen outcome to be satisfied. The practical contribution is a traceable connection between refinancing value, available money, retained obligations and the decision that authorises their allocation.

19 Worked payment-delay diagnostic: annual coverage and an intra-year funding gap

A separate author-created liquidity case makes the timing question executable. It uses the Extend debt-service total of 61.97, but assumes four equal quarter-end instalments solely for this diagnostic. It does not change the main annual shareholder valuation or establish actual facility payment terms. Hypothetical monthly invoices are 27.92 and cash operating costs are 20.00, giving annual receipts of 335 and operating costs of 240. These amounts reproduce annual CFADS of 95. Tax and lifecycle cash are assumed included in the operating amount for this simplified test.

Opening unrestricted operating cash is assumed to be 20.00, separate from the debt-service reserve of 30.98. The diagnostic is after closing and before tariff sharing. The first three invoice receipts arrive together in month four. Later invoices are paid in their invoice month. Annual receipts remain unchanged. There is no assumed committed working-capital facility. When operating cash is insufficient to pay the monthly operating amount, the schedule records required sponsor support. That entry identifies an unresolved funding requirement; sponsor agreement and ability to pay remain unverified.

At month one, opening cash pays the operating costs and falls to zero. Months two and three each require support of 20.00 for operations. The quarter-end debt instalment of 15.49 in month three then draws on the reserve, assuming that the reserve documents permit such a draw for debt payment. Peak operating support is 40.00 and the reserve falls to 15.49. The undiscounted annual coverage ratio remains 1.533x throughout this timing comparison, although cash is unavailable on the first payment dates.

Month four receives 111.67 from four invoices and pays 20.00 in operating costs. The diagnostic waterfall replenishes the 15.49 reserve draw and refunds the 40.00 support. Refund assumes no lender, covenant or subordination restriction; actual eligibility remains unverified. Cash after those movements is 36.18. Later months retain normal collection and quarter-end payment. At year end, cash is 53.03, the reserve is restored and support refunded. Opening cash of 20 plus annual CFADS of 95 less debt service of 61.97 reconciles to that ending balance.

If the reserve cannot be drawn in month three, additional support of 15.49 is required for debt service. Peak external support would therefore be 55.49 under the same receipt schedule and opening operating cash. If opening cash is also unavailable, a further 20.00 is required before considering financing charges. These are conditional funding requirements. The calculation assumes timely support whenever a gap appears; without that support the corresponding payment cannot be funded by the accounts shown.

The result is a financing instruction for diligence: identify the permitted reserve draw and a documented source for the operating gap before approving the collection-delay case. Support fees, interest, tax effects, draw conditions and default consequences are omitted. Their inclusion would change the cash schedule and could affect refinancing value. The quarterly diagnostic is not combined with the main annual NPV, and its assumed operating buffer is not counted as a refinancing gain. A lender model must replace these simplifying dates, invoices and costs with actual records.

Table 9. Quarter-end liquidity with the first three receipts delayed to month four

Hypothetical USD millions. Operating costs are 20 monthly. Month four replenishes the reserve by 15.49 and refunds support of 40.00. Support entries are unresolved funding requirements; no commitment is assumed.

Author-created timing diagnostic. Support is an unresolved funding requirement; reserve use and payment dates are assumed separately from the annual valuation.

20 Apply the decision with explicit limitations

The available-cash valuation favours Cash out at 10 percent, while the specified resilience requirement disallows each submitted structure under the 20 percent stress. A committee adopting that requirement should withhold approval pending revised borrowing or supported operating assumptions. A committee permitting temporary lock-up should instead require a dated trapped-cash, reserve-replenishment and cure model. Neither decision follows from the positive gain alone.

The example establishes reproducible financial dependencies and conditional outcomes. It uses no observed operating dataset and estimates no market distribution of refinancing benefits. Its rates, shares, thresholds, dates and project values are author-created assumptions. Corporate announcements and public guidance support the evidence review; they do not validate those assumptions for a particular asset. Subsequent transaction use requires an executed contractual perimeter and sourced financial inputs.

The practical result is a replacement for unexplained gain and coverage tables. A reader can trace principal repayment, reserve timing, closing contributions, public delivery and adverse cases through one case. The same case also exposes its unfinished implementation questions: covenant-aware cash trapping during a stress, payment-date liquidity, transaction-specific tax, lifecycle variation and actual termination economics. These limitations identify the remaining work required before money can be distributed under an approved transaction.

21 Replication appendix: debt schedules and cash-flow reconciliation

The following schedules expose every period retained in the gain calculation. For opening principal D, annual fixed rate r and n payments, level service equals D × r ÷ [1 − (1 + r) raised to −n]. Annual interest equals opening principal × r; principal equals service less interest. Each year closes at opening principal less the principal payment. Displayed totals can differ slightly because rows are rounded to two decimals.

The shareholder schedules use CFADS less debt service, plus a reserve release only at the specified maturity. They retain all fifteen years, including periods after old-debt maturity. A delta is refinancing cash less counterfactual cash. The discounted annual deltas plus closing cash before public sharing reproduce the unrestricted gain. All rates and cash-flow inputs remain hypothetical. A transaction-specific trapped-cash waterfall is outside this instructional calculation.

Table 10. Counterfactual annual amortisation

Hypothetical USD millions. Reserve 45.00 is released in year 10; it is excluded from this debt schedule and included in shareholder cash.

Table 11. Reprice annual amortisation

Hypothetical USD millions. Reserve 37.91 is released in year 10; it is excluded from this debt schedule and included in shareholder cash.

Table 12. Extend annual amortisation

Hypothetical USD millions. Reserve 30.98 is released in year 13; it is excluded from this debt schedule and included in shareholder cash.

Table 13. Cash out annual amortisation

Hypothetical USD millions. Reserve 35.11 is released in year 13; it is excluded from this debt schedule and included in shareholder cash.

Table 14. Available shareholder cash before public sharing or covenant trapping

Hypothetical USD millions. Each reserve is released at its stated maturity. Values describe unrestricted available cash, not legally approved distributions.

Table 15. Period-by-period discounted incremental shareholder cash

Hypothetical USD millions except the discount factor. Closing cash includes fees, hedge settlement and net reserve funding; no cost is deducted a second time.

Table 16. Delivery and reconciliation of the fifteen-year public tariff stream

Hypothetical USD millions except factors. The stream replaces upfront sharing. This equality assumes the same discount rate, payment dates and full delivery.

References

1. World Bank. Guidance on PPP Contractual Provisions, 2019 edition. Chapter 5, printed pp. 93–100. https://ppp.worldbank.org/sites/default/files/2021-03/Guidance%20on%20PPP%20Contractual%20Provisions_2019%20edition.pdf
2. European PPP Expertise Centre, European Investment Bank. EPEC Guide to Public-Private Partnerships, published 16 April 2026. Printed pp. 197–198. https://www.eib.org/files/publications/20250124-160426-epec-guide-to-ppp-en.pdf
3. HM Treasury. Calculation of the Authority’s Share of a Refinancing Gain (SoPC4 guidance); inspected note undated. Sections 1.2–1.4 and 1.6. Government collection published 25 July 2013, updated 7 December 2018. https://assets.publishing.service.gov.uk/media/5a7c1d8ce5274a25a91408cb/06_pfi_refinancingguidance21307.pdf ; collection: https://www.gov.uk/government/publications/pfippp-finance-guidance
4. HM Treasury. Standardisation of PF2 Contracts, December 2012, Draft. Chapter 28, sections 28.3, 28.5 and 28.6. https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/221556/infrastructure_standardisation_of_contracts_051212.pdf
5. HM Revenue & Customs. BIM64370: PFI benefit-sharing arrangements. Manual published 22 November 2013; manual update displayed 4 August 2026. https://www.gov.uk/hmrc-internal-manuals/business-income-manual/bim64370
6. ACWA Power. Successfully completes refinancing of Rabigh 3 Independent Water Project. Undated issuer announcement; precise closing date unresolved in inspected source. https://www.acwapower.com/en/media-center/latest-news/acwa-power-successfully-completes-refinancing-of-rabigh-3-independent-water-project/
7. ACWA Power. Report for the three and nine months ended 30 September 2025. Printed p. 8, section 2.3.1. https://acwapower.com/media/4aujzqo1/acwa-power-q3-report-2025-eng.pdf
8. ACWA Power. Barka completes RO 73.32 million refinancing and RO 16.6 million financing. Undated issuer announcement; publication and transaction dates unresolved. https://acwapower.com/en/media-center/latest-news/acwa-power-barka-completes-ro-7332-million-refinancing-and-ro-166-million-financing/

All public sources accessed 9 October 2026. The worked-case calculations, tables and figures are original author-created illustrations using the declared hypothetical inputs. No actual project outcome is estimated.

About the Author

Chennakeshav (CK) is a corporate finance and investment banking executive with 25+ years of global experience in deal origination, structuring and execution across M&A, growth capital and corporate strategy. He has led value-creation mandates for founders, corporates and funds — bridging the boardroom view to hands-on execution and close.

His career spans Morgan Stanley, HSBC, Lloyds Banking Group, EWEC, ADQ portfolio companies and Emirates Growth Fund, across TMT, real estate, fintech, deeptech, cleantech, infrastructure and energy. He has partnered with C-suite leaders, private equity and venture funds, sovereign wealth funds and family offices to finance complex fund raises and scale-up ventures, and has led M&A due diligence, post-merger integration and business-transformation initiatives to create value.

At Matchpoint Partners he is Managing Partner, leading the firm's corporate finance, M&A and capital-raising practice. He holds an MBA from London Business School, an engineering degree from VTU and a Master of Laws (LLM, in progress) from UCL London.

An active start-up mentor, CK mentors at Techstars, DIFC FinTech Hive, Startup Grind, Founder Institute and IN5, serves as Entrepreneur Mentor in Residence (EMiR) at London Business School, and judges the Entrepreneurship World Cup.

https://www.linkedin.com/in/ckadya/ https://www.matchpoint-partners.com/team/ck-adya.html