Running a first feasibility check on a Portuguese site

A first feasibility exists to answer one question: is this site worth spending money on? It should take hours, use publicly available information plus one or two phone calls, and be capable of being rerun as new facts arrive.

That purpose dictates its shape. Precision is not the objective; robustness is. A model that produces a return to two decimal places from twelve invented assumptions is less useful than a coarse model whose downside case is honest.

Three rules keep a first-pass appraisal useful:

• every revenue assumption must be traceable to evidence, not to opinion
• every cost assumption must include the items that are habitually forgotten
• the output must be a range, produced by explicit sensitivity tests, not a single number.

Buildable area, not plot area, drives revenue. The municipal plan sets capacity through a combination of parameters, and the binding one is whichever gives the smallest result:

• the building index applied to plot area
• the maximum implantation footprint
• the maximum height or number of floors above ground
• setbacks from boundaries, roads and watercourses
• alignment and morphology rules in consolidated urban fabric
• parking provision, which consumes area and can constrain the footprint.

Definitions differ between municipalities: whether basements, balconies, technical areas and circulation count towards the index is a question for the local regulation, not a national constant. At first pass, take the smallest plausible envelope. Optimism here compounds through every subsequent line of the model.

Gross buildable area is not what is sold. Cores, corridors, plant rooms, common areas and non-saleable parking absorb a significant share. The efficiency ratio is a design outcome that depends on typology and core strategy.

The first-pass discipline is to apply a conservative ratio, flag it as the largest single uncertainty in the model, and replace it as soon as an architect produces a sketch layout. Where the product includes parking, storage or terraces sold separately, they should be priced as separate lines rather than folded into a single rate.

Set the sales value per square metre from closed transactions for comparable product in the same micro-market. In Portugal, that means combining national statistical series on transaction prices with local agency evidence for the specific street, view, floor and finish level.

Adjust explicitly for:

• product differences — new build versus refurbished stock, typology mix, outdoor space
• absorption — how many units the micro-market realistically absorbs per month
• the sales period, which for larger schemes extends well beyond practical completion
• any price growth assumption, which at first pass should be zero.

A phased sales programme also changes the cash-flow profile, which changes finance cost. At first pass it is enough to assume a realistic absorption rate rather than instantaneous disposal at completion.

Construction cost is normally expressed as a euro-per-square-metre range for the product type and specification, sourced from a builder or quantity surveyor active in the region. The range matters: a single point estimate hides the variance that decides the project.

COSTS ROUTINELY OMITTED AT FIRST PASS

• site preparation, demolition, rock excavation and contaminated soil
• infrastructure and urbanisation works outside the building footprint
• utility connection and network reinforcement charges
• municipal urbanisation charges and any compensation or land cession obligations
• professional fees across architecture, all specialities, project management and site supervision
• surveys — topographic, geotechnical, archaeological, acoustic, environmental
• insurance, bonds and warranties
• marketing, agency commission and legal costs on sales
• a contingency of at least ten per cent, held separately and not spent on scope.

Acquisition taxes are paid before the deed and are rarely financed. Property transfer tax and stamp duty must be modelled at the rate applicable to the specific transaction and asset, and the VAT position on construction and on the eventual sales has to be established early because it is a cash-flow question as much as a cost question.

Finance cost is a function of time. Interest accrues from the first drawdown to repayment, across a period that includes licensing, construction, and the sales tail. A scheme with a strong margin and an eighteen-month planning delay can lose most of that margin to carry cost alone.

Model the programme explicitly:

• pre-application and design period
• municipal review, including realistic allowance for requests for clarification
• external authority opinions, which run on their own timescales
• procurement and mobilisation
• construction
• inspection, utilisation licence and technical file
• sales and completion of deeds.

Run the model at minimum three times.

• Base case, using evidenced values and mid-range costs.
• Price downside: sales values ten per cent lower.
• Delivery downside: costs ten per cent higher and the programme six months longer.

A viable site clears the required margin in the base case and remains solvent — not necessarily profitable — in both downsides. A site that fails either downside is not automatically rejected, but it can only proceed with a structural mitigation: a lower land price, a conditional contract, a pre-let, or a phasing strategy that limits committed spend.

The residual output is the number that should govern negotiation: the maximum land price that still delivers the target return after the downside test.

• Treating plot area as buildable area.
• Using the headline index and ignoring the footprint, height or parking constraint that actually binds.
• Assuming an efficiency ratio from a different typology.
• Pricing sales from listings instead of closed transactions.
• Excluding infrastructure, connection and municipal charges from cost.
• Omitting finance cost, or modelling it only across the construction period.
• No contingency, or a contingency quietly absorbed into scope during design.
• Presenting a single-point return with no downside case.

A first feasibility that survives sensitivity testing justifies the next tranche of spend: a planning opinion or PIP, a sketch scheme from an architect, a cost check from a quantity surveyor, and legal due diligence on title. Each of those replaces an assumption with a fact, and the model is rerun.

The appraisal is a living document through the whole cycle. It sets the bid at acquisition, supports the lender pack at funding, tracks variance during construction, and measures outcome at exit.