Land Finance and Construction Finance: Why Lenders Split Them

A development loan looks like a single facility from the outside. From the inside it is almost always two. The ground is financed on one set of terms and the building on another, because the ground and the building expose the lender to different risks, at different moments, with different remedies if something goes wrong. A developer who understands why lenders draw that line, and what each tranche demands, presents a file that moves. A developer who asks for one loan against one future asset is asking the lender to price two risks as if they were one, and lenders decline to do that.

The split is not bureaucracy and it is not caution for its own sake. It is the credit logic of development finance expressed in structure. Everything else in a development facility, the conditions, the controls, the order in which money is released, follows from this first decision to treat land and works as separate problems.

Two risks that only look like one

The risk in the land is a risk of permission and value. A plot produces no income while it waits. Its worth depends on what may lawfully be built on it, which in turn depends on zoning, planning consents and permits that sit in the hands of authorities, not of the borrower. If the project never starts, the lender is left holding security over an asset whose value was always conditional on an outcome that did not arrive. Land risk is, at bottom, the risk that the story attached to the ground does not come true.

The risk in the works is a risk of execution. Once construction begins, the questions change entirely. Will the contractor perform. Will the budget hold. Will the timetable survive contact with the ground conditions, the supply chain and the weather. A half-built structure is worth less than the money spent on it, sometimes dramatically less, because no buyer wants someone else’s unfinished problem. Construction risk is the risk that money goes into the site faster than value comes out of it.

These two risks do not move together. Land risk falls as consents are secured and is largely resolved by the time the first foundation is poured. Construction risk barely exists at acquisition and is at its most intense in the middle of the works, when the maximum amount of money has been spent and the asset is at its least sellable. A single facility covering both would have to be priced and secured for the worst point of each curve simultaneously.

That is why the lender separates them. Each tranche can then carry its own conditions, its own security tests and its own release mechanics, matched to the risk it actually covers. The borrower is not paying construction-grade margins on money that only buys ground, and the lender is not extending construction money before the questions that belong to the land have been answered. The split is what allows either side to enter the transaction at all.

What the land tranche demands

The land tranche finances the acquisition of the site, and the lender’s analysis concentrates on what stands between the ground as it is and the ground as the project needs it to be. Planning status is the first question. A site with full consent, a site with outline permission and a site with nothing but a zoning designation are three different assets, whatever the purchase contract says. The lender will want the current status documented precisely, the remaining steps identified, and an honest view of what could still refuse, delay or condition the consent. In the Netherlands as in Spain, the calendar of those procedures is set by the authority, not by the transaction, and the lender will expect the timetable to acknowledge that rather than compress it.

Because the land produces nothing while it waits, the lender also asks how the tranche is carried. There is no rent to pay interest, so servicing depends on the borrower’s own resources or on interest being capitalised into the facility, and either answer has to be shown to work for the whole period the consents may realistically take. Equity is concentrated here for the same reason. The developer’s own money goes into the ground first, so that the lender’s exposure at the riskiest, least productive stage of the project is as small as the transaction allows.

What goes wrong at this stage is usually optimism. A developer buys with conditions unresolved and treats the permit as a formality, because in his experience of the municipality it always has been. The lender does not underwrite experience of the municipality. It underwrites the documented position, and a file that waves at the planning question instead of answering it invites either a refusal or terms that price the uncertainty in full.

The remedy is order. The path from the plot to a consented site is set out step by step, with the evidence for each step attached, and the maturity of the land tranche is aligned with that path rather than with hope. When the consents arrive, the land risk has been retired, and the project arrives at the works tranche with its first question already answered. That sequencing is also what later allows the works tranche to open on schedule, because a lender rarely reopens questions that were closed with evidence.

What the works tranche demands

The works tranche finances a building that does not yet exist, and so it is assessed on the plan for bringing it into existence. The lender expects a total budget that includes everything, not only the construction contract but the professional fees, the connections, the financing costs of the facility itself and a contingency for the things no budget foresees. It expects a construction timetable that a professional would defend, an identified contractor whose capacity can be verified, and clarity on how much of the cost is fixed under the contract and how much floats with the market. It will also want to know how the interest accruing during construction is funded, because a project under construction pays its own financing costs out of the facility, and a budget that omits them is understated by definition.

The defining feature of the tranche is that the money is not handed over. It is released in stages, against evidence that the works have actually progressed, so that the lender’s exposure grows only as the value on the site grows. The mechanics of that release, who certifies the progress and what happens when reality deviates from the plan, are a discipline of their own. What matters here is the principle: the works tranche is not a sum, it is a schedule.

The lender also wants to see the developer’s cushion. Construction costs move, and a budget with no contingency is a budget that fails on the first surprise. The question the lender is really asking is simple: if this costs more and takes longer than planned, is there enough capital in the structure to finish. A project that runs out of money at the halfway point does not lose half its value. It loses far more, because an unfinished building is the worst asset either party could hold.

Finally, the works tranche is underwritten against its own ending. The lender wants to know what repays it: sale of the completed units, or refinancing of the finished, income-producing asset onto ordinary investment terms. That exit is not a detail for later. It shapes the amount, the term and the conditions of the tranche from the first conversation, because a facility whose ending is undefined is a facility no credit committee can approve.

A Dutch project, financed in phases

In one transaction we structured, a developer needed to finance both the purchase of a plot and the construction of a Dutch project that was not yet generating any income. Presented as a single request, the file carried every problem described above at once: conditional land value, execution risk and no cash flow to point to. The financing was therefore separated into a land tranche and a works tranche, each with its own conditions.

The preparation did the real work. A total budget was built covering the construction cost, the equity contribution, the contingency and the financing costs, alongside a construction calendar and the specific conditions attached to each drawdown. Nothing in the package asked the lender to take anything on trust. Each stage of the project had its number, its date and its condition precedent.

The disbursements were tied to the certified progress of the works, so money entered the site as the building rose and not before. Financing the project in phases kept the cost overruns under control and, just as importantly, preserved capital to finish. At no point was the structure exposed to the situation every development lender fears, a site full of spent money and an unfinished building with no funds behind it.

The exit was prepared in parallel rather than left for the end: sale of units and refinancing of the stabilised asset. By the time the works tranche was fully drawn, its repayment was not a question but a plan already in motion. The project reached completion financed at every stage on terms that matched the risk of that stage.

The split protects the developer as much as the lender

It is tempting to read the two-tranche structure as the lender protecting itself, and it is that. But the discipline cuts both ways. A developer who draws a single large facility at the outset pays interest on the whole amount from the first day, including on money the project will not need for a year. Phased drawings mean the debt grows with the building, which is the cheapest way to carry a construction project and the strongest defence of the developer’s own margin.

The split also forces the right questions in the right order. Before the land tranche closes, the planning position must be documented. Before the works tranche opens, the budget, the contingency, the contractor and the exit must survive scrutiny. Every one of those questions would have had to be answered eventually. The structure simply refuses to let them be answered late, when the answers are expensive, instead of early, when they are merely work.

Developers sometimes resist the structure because it feels like control, and it is control. But the alternative is not freedom. The alternative is a single facility priced for the worst risk in the project across its entire life, or more commonly no facility at all. The lenders who finance development, whether traditional banks, family offices or specialist credit institutions, differ in appetite and in flexibility, but on this point they converge, because the logic is not institutional preference. It is the nature of the asset. A structure that respects that logic travels well between institutions. A structure that fights it has to be rebuilt for each one, and usually is not.

One project, two tranches, is therefore not a complication to be negotiated away. It is the shape a development takes when its risks have been honestly separated and each has been given a structure that answers it. The developers who internalise that arrive at the table with the split already built into their own numbers, and they are the ones whose projects get financed.

Montclare does not act as a lender. It structures transactions, prepares the file and coordinates financing with authorised institutions.