Construction loan vs forward funding

Both routes get a building financed before it produces anything. The difference is who is holding the project on the day the contractor says it will be six months late.

Every row follows from one asymmetry: long-term capital does not price construction risk well, and developers do.

A facility that does not yet exist can fail to exist. It can cost more than budgeted, arrive later than promised, or arrive built to a specification the occupier will not accept. None of those risks resemble the risk of owning a completed, contracted building, and the capital that wants the second is generally poor at pricing the first.

A construction loan resolves that by leaving the risk where the competence is. The lender advances against progress, holds security, and relies on the developer's guarantees, a fixed-price contract and a contingency to absorb what goes wrong. In exchange the developer keeps the whole of the upside: if the completed asset is worth more than it cost, the difference is the development margin and it belongs to the developer.

Forward funding resolves it differently, by bringing the long-term owner in early and paying it to accept a version of the risk. The funder releases capital against progress and takes the asset on completion at a price set in advance. It gets the asset at a development-inclusive price rather than a completed one; the developer gets certainty of exit and a much smaller capital requirement, and gives up most of the margin.

The control row is the practical consequence sponsors most often underestimate. A funder that will own the building specifies the building. Approval rights over design, contractor selection, material changes and the occupier's fit-out are the norm, not an imposition, and they slow decisions during the phase where speed is most valuable.

The two failure rows are the real test and they are worth stating together. A construction loan fails where the developer cannot absorb an overrun, because the guarantees are then worth less than the exposure they cover. Forward funding fails where nobody will commit — and a funder's refusal to commit is usually a statement about the offtake or the site rather than about the building.

The word forward describes a commitment made before completion; it does not, by itself, say when the purchase price is paid or who funds the build. That missing fact is what determines the developer's capital requirement.

In a forward-funding structure, the investor's capital is released during construction under an agreed draw process. The agreement normally states what evidence supports each draw, which costs qualify, who certifies progress, what the investor may retain, and what must happen before the next release. The investor is therefore exposed before the completed asset exists, while the developer usually remains responsible for delivering it to the agreed specification. Completion risk may be allocated, capped or supported, but it does not disappear.

In a forward purchase, the buyer commits before completion but generally pays the acquisition price when the completion conditions are satisfied. The commitment may remove exit-market risk, but it does not ordinarily fund the construction period. The developer still needs its own equity, a construction loan, or another source to carry the build to the point at which the buyer must close.

Forward commitment is often used as an umbrella label for an agreement made in advance, and sometimes as a local synonym for one of the two structures above. It is not safe to infer the funding mechanics from that label. The documents have to answer the operative questions: when cash moves, what conditions each payment, whether the commitment can be terminated, who owns the work in progress, and who supplies additional capital if the budget is exceeded.

The funding schedule is therefore the practical dividing line. A sponsor comparing proposals should map the sources for land, deposits, construction draws, change orders, contingency and the gap between practical completion and final acceptance. A deal can be described as fully committed while still leaving one of those periods unfunded.

A sources-and-uses total can balance while the project still runs out of cash between milestones. The useful comparison is therefore a calendar: when each use must be paid, which party is obliged to fund it, what evidence releases the money, and who carries it if the release condition is not yet satisfied. That exercise often exposes the real difference between a construction facility, forward funding and a forward purchase faster than the transaction name does.

Long-lead equipment is the most common timing trap. A utility, transformer or switchgear deposit may be due before the investor's definition of eligible construction cost begins, and a forward purchaser that pays only at completion does not solve the gap at all. Even under forward funding, the investor may decline to recognise a reservation until it becomes a firm order, require title or vesting over work in progress, or retain part of each draw until delivery and acceptance. The sponsor then needs equity, a separate procurement facility or an express early-cost provision.

The same issue appears at the other end of the build. Practical completion may release most funding while tenant acceptance, final commissioning, lien releases and retention remain outstanding. If rent has not commenced and the final investor payment is still conditional, interest and operating carry need a named source. A structure is not fully funded merely because an investor is committed; it is fully funded only when the committed draw conditions and the project's payment dates fit each other.

A construction loan wins where the developer has both the capacity to carry the risk and a reason to want the reward.

The developer can absorb an overrun. This is the gate, and it should be tested against a realistic overrun rather than the contingency line. Guarantees, cost-overrun support and completion undertakings are only as good as the party providing them; a developer whose whole equity is committed to the project cannot meaningfully guarantee it.

The completed value is expected to exceed cost by a real margin. That margin is what the developer is buying with the risk. Where it is thin, the risk is being taken for very little, and a committed exit at a slightly lower price is the better trade.

Specification flexibility matters. Where the occupier is not yet fixed, or where the fit-out will be negotiated during the build, funder approval rights are a genuine constraint. A developer that needs to change the design mid-build wants no third-party owner to persuade.

A refinancing route is visible. The normal end state is that completion transforms the credit — the asset now exists and is contracted, which is what long-term capital wants. Establishing at the outset that the construction facility permits that refinancing without penalty is the step most often left to later; the prepayment terms are set at a moment when the refinancing is still hypothetical. The routes out are compared on sale-leaseback vs refinancing.

It is the wrong answer where the developer is small relative to the project, where the offtake is not yet contracted and the exit therefore depends on a market rather than a counterparty, or where the construction contract cannot be placed on a fixed-price basis. In that last case the risk being retained is genuinely open-ended, and retaining an open-ended risk to earn a fixed margin is a poor trade however good the site.

Forward funding wins where certainty is worth more than margin, and where a buyer can be found who agrees.

The developer's capital is the binding constraint. A developer with more sites than equity is better off recycling capital quickly at a lower margin per project than holding one project for a higher one. Forward funding is how development capacity is multiplied.

The occupier is contracted. This is what makes the structure available at all. A funder commits during construction because it can see the cash flow it will own; without a firm, long-dated obligation from a counterparty that survives stress, there is nothing to commit against. The tests that separate a bankable contract from a signed one are on compute offtake as credit, and they gate this route as surely as they gate project finance.

The development margin is modest anyway. Where the spread between cost and completed value is thin, the developer is not giving up much, and it is exchanging a small uncertain gain for a certain exit.

Market risk at completion is the worry. A developer that fears the exit market more than the build has exactly the risk forward funding removes. It converts a market-timing question into a counterparty question.

It is the wrong answer where the developer expects values to rise materially, where the specification is genuinely unsettled, or where no funder will commit at an acceptable price. That last case deserves attention rather than a search for a different funder: a refusal to commit is normally a statement about the site or the offtake, and the same gates that govern it are set out on what makes a site financeable. Whether the sponsor is selling land or a building at all is the prior distinction, drawn on powered land vs powered shell.

One mechanic to negotiate rather than accept: what happens if completion is late. Longstop dates, price adjustment, liquidated damages and the funder's right to step in and complete are the substance of the arrangement. A developer that has read only the price has read the least contingent part of the document.

Neither structure rescues a project that is not ready to be financed, and the same preconditions gate both.

A power position that is documented and dated. Construction capital funds a building that will be energised. Where the interconnection position is a letter rather than an agreement, or where the energisation date is an aspiration, the completion risk both structures allocate is not really the construction risk — it is the risk that the finished building cannot be used.

Site control that outlasts the build with margin. An option expiring before energisation is a countdown rather than control, and it is the failure that looks like control right up until diligence.

A construction contract that allocates risk to a party that can carry it. A fixed-price contract with a contractor unable to absorb the fixed price is a fixed price in name only. Both routes lean on it, and both discover the same thing at the same moment.

A clear view of who takes the compute layer. A construction facility sized for a building should not be quietly funding equipment with an economic life of a few years. Where the compute is being bought alongside the shell, the layers should be separated at the outset rather than untangled later — the argument on lease vs own compute, and the reason corporate credit vs project finance is the decision that comes first.

An honest completion definition. "Complete" is a contractual term and the parties should mean the same thing by it. Practical completion, energisation, the occupier's acceptance and the commencement of rent can all fall on different dates, and the gaps between them are where funding stops and obligations do not.

The useful sequencing test is simple. Ask what happens on the day the project is six months late and materially over budget, and name the party who writes the cheque. Under a construction loan it is the developer. Under forward funding it is whoever the agreement says, and if the agreement is unclear the answer is decided under pressure. Both structures are workable. Only one of them should be entered without knowing which.