Financing a development in Portugal

Almost every development is funded from more than one source, and each layer has a different price, a different security position and a different tolerance for delay.

• **Sponsor equity.** The first money in and the last money out. It absorbs cost overruns and it is the layer that determines whether a scheme survives a delay.
• **Senior development debt.** Usually a bank facility secured on the land and works, drawn in stages against certified progress, and repaid from sales or from a refinancing at completion.
• **Mezzanine or junior debt.** Fills the gap between senior debt and available equity. More expensive, structurally subordinated, and normally only worth using where the return justifies the cost of the whole stack.
• **Partner or co-investor equity.** Third-party capital brought in at project-company level, governed by the shareholder agreement rather than by a loan document.
• **Pre-sale deposits.** Purchaser deposits taken under promissory contracts. Useful evidence of demand, but they are contractual obligations, not free capital, and lenders treat them accordingly.

The design question is not "how much can we borrow", but "what mix survives a six-month licensing delay without a capital call".

Credit committees in Portugal look at the same file, roughly in the same order:

• **The planning position.** Is there a licence, a PIP, or nothing? An unlicensed scheme carries planning risk that a construction lender is generally not paid to take.
• **The title and the encumbrances.** Registry position, easements, charges, and whether the site as described matches the site as registered.
• **The cost plan.** Whether the build cost is a real measured estimate with a defensible contingency, or a rate per square metre applied to an area schedule.
• **The contractor and the contract form.** Who is building, on what terms, with what price certainty and what performance security.
• **The sales evidence.** Comparable transactions, absorption rates, and any pre-sales already contracted.
• **The sponsor.** Track record on comparable schemes, and the ability to fund an overrun.
• **The borrowing entity.** Incorporation, NIF, RCBE, accounts, and a clean compliance file.

A weak item in any one of these does not automatically stop a facility; it re-prices it, reduces it, or attaches conditions to it.

These are different products and should be modelled separately.

A **land facility** funds acquisition. It is exposed to planning risk, has no works to certify against, and generates no value while it sits. It is therefore shorter, smaller relative to value, and priced higher. Where a sponsor uses one, the exit is normally a refinancing into a construction facility once the licence is issued.

A **construction facility** funds the build. It is drawn progressively, secured on an asset whose value increases as the works advance, and underwritten against a fixed or substantially fixed contract price. Because it is drawn in stages, the sponsor carries the early costs — design, fees, statutory charges — before the first drawdown is available.

MODELLING RULE

The gap between the last equity euro and the first drawdown is the most dangerous point in a development cash flow. Model it explicitly, with the licensing calendar, not as a rounding assumption.

Development debt is released against work in place. The mechanism is consistent:

1. the contractor submits a valuation of the works completed in the period;

2. the project supervisor or monitoring surveyor verifies quantities and quality on site;

3. a certificate is issued confirming the value of completed work;

4. the lender releases the corresponding drawdown, less retention and less any conditions;

5. the borrower pays the contractor within the contractual period.

Two practical consequences follow. First, there is always a working-capital lag between the contractor's costs and the lender's payment, and someone must fund it. Second, the monitoring surveyor is effectively a gatekeeper: disputes about quality or measurement stop money, and stopped money stops the site.

Expect a security package built from some combination of:

• a mortgage over the land and works
• pledge of the shares in the project company
• assignment of the construction contract, the design appointments and their warranties
• assignment of insurance proceeds and of sale proceeds
• pledge over project bank accounts, sometimes with controlled disbursement
• a cost-overrun undertaking or completion guarantee from the sponsor
• personal or corporate guarantees, limited or unlimited depending on the sponsor's standing.

The negotiation that matters is not the existence of security but its scope: whether recourse is limited to the project, and what triggers a call on the sponsor.

• **Loan to cost (LTC)** — debt as a proportion of total project cost, tested at drawdown.
• **Loan to value (LTV)** — debt against valuation, tested periodically and at completion.
• **Pre-sales cover** — a minimum value or number of contracted units before construction drawdowns are released.
• **Cost-to-complete** — the facility must always be sufficient to finish the works; if it is not, the sponsor injects the difference before the next drawdown.
• **Longstop dates** — practical completion and repayment deadlines, with default consequences.
• **Interest cover or sales-proceeds sweeps** — how repayment is applied as units complete.

The cost-to-complete covenant is the one that most often converts a cost overrun into an immediate equity call, months before the sponsor expected to contribute again.

A development produces no income while it is being built. Interest is therefore either serviced from equity or capitalised and rolled into the loan. Rolled-up interest is convenient and expensive: it compounds against a fixed sales value, and each month of delay increases the amount that must be repaid from the same revenue.

This is why a licensing delay is a financial event, not an administrative one. A scheme that takes nine months longer than modelled pays nine months of additional finance cost, additional professional retainers and additional holding charges, while its gross development value stays where the market puts it.

• Build the model on a monthly timeline tied to the licensing and construction programme, not on annual totals.
• Separate committed costs from contingent costs, and show the peak equity requirement explicitly.
• Sensitise for a delay of six and twelve months on the licence, a build cost increase, and a slower sales absorption — separately and together.
• Model the moment of maximum debt exposure, not just the end position.
• Keep contingency as a real reserve with a release protocol, not as an unallocated line that gets spent early.

• Treating a term sheet as committed funding before conditions precedent are satisfied.
• Starting bank onboarding after the promissory contract is signed.
• Modelling drawdowns from the day the contract is signed rather than from certification of works.
• Assuming pre-sale deposits are available to fund construction.
• Setting contingency at a level that cannot absorb both a programme extension and a cost increase.
• Ignoring the VAT cash-flow timing on construction invoices.

• the title and registry pack, with encumbrances resolved or explained
• the planning position, with the licence, the PIP or a written route and timetable
• a measured cost plan with a stated contingency basis
• the procurement strategy and the intended contract form
• the sales strategy with comparable evidence
• the corporate pack for the borrowing entity: incorporation, NIF, RCBE, accounts, beneficial ownership
• a monthly cash-flow model with sensitivities.

Lenders do not reward optimism. A file that shows the risks and how they are covered is faster to approve than one that shows only the upside.