How We Analyze Portugal’s Real Estate Market

Good market intelligence begins with honest definitions

Real estate statistics can look reassuringly precise. A city is assigned an average price per square metre. A report says homes took five months to sell. A chart shows a discount of 5%. The numbers may be accurate within the source methodology, but they are not automatically answers to the questions a buyer or seller is actually asking.

What kind of property was included? Was the price an asking price or a completed transaction price? Does “five months” measure the time between listing and contract, or the amount of inventory available? Is the figure an official count, an estimate or an average calculated from a participating group of companies? When two periods differ, did prices change—or did a different mix of properties sell?

AGTP Market Intelligence exists to make those distinctions visible. We use selected market statistics to help Americans understand Portugal's housing market, but we do not treat a market average as a valuation of an individual home or present complicated data as more certain than it is.

The short version: AGTP selects and interprets a limited number of real estate indicators relevant to Americans buying, selling or comparing locations in Portugal. We identify the source, period, geography and property segment; distinguish source data from our calculations; explain important limitations; and connect the numbers to practical decisions.

Our principal market-data source

The principal source for AGTP's current real estate market reports is Confidencial Imobiliário, an independent Portuguese real estate databank operated by Imoestatística – Sistemas de Informação de Imobiliário, Lda.

Its SIR — Sistema de Informação Residencial, or Residential Information System, aggregates residential sales information reported by major real estate market operators. Confidencial Imobiliário uses this information to produce statistics about transaction prices, market activity, listing-to-transaction time and the difference between asking and completed sale prices.

The source workbooks used in our current analysis identify their contributors as participating real estate developers and intermediaries. Confidencial Imobiliário also explains that its quarterly releases can incorporate information from brokerage networks, developers, investors and credit institutions, depending on whether the release is preliminary or definitive.

AGTP is not the producer of this underlying market data. Our role is to select limited figures, perform clearly identified calculations where appropriate, explain the methodology in plain English and add American-specific decision context.

Required source notice for CI-derived reproductions:

Source: Confidencial Imobiliário, SIR — Residential Information System. AGTP analysis. © IMOESTATISTICA – TODOS OS DIREITOS RESERVADOS.

Is this official Portuguese government data?

Not by itself. Confidencial Imobiliário is a private market-data provider, not a Portuguese government agency.

However, CI's estimate of the number of homes sold begins with the most recent residential transaction data published by Portugal's Instituto Nacional de Estatística, or INE. CI then projects that activity forward using changes measured through a panel composed of major companies participating in SIR. Regional totals are allocated to smaller geographic and property segments using the structure of sales recorded in the SIR system.

For that reason, AGTP describes this indicator as:

CI's estimated number of residential property transactions

We do not describe it as a raw count of records in the workbook, a count of transactions handled only by participating agencies or an official registry total for the period.

What geographic areas does AGTP cover?

Coverage depends on the source export and the purpose of the report. Current files include mainland Portuguese regions and municipality-level results, but they do not include parish-level observations or the autonomous regions of Madeira and the Azores.

AGTP reports will identify the geography used in each table, chart and conclusion. We do not assume that a municipality-wide average describes every parish, neighborhood or micro-location within it.

This matters particularly in markets such as Lisbon, Cascais, Sintra and Setúbal, where location, property type, condition, view, outdoor space, parking and building quality can create substantial differences within the same municipality.

Geographic-name quality control

Some exported workbooks contain machine-translated municipality names. These translations can produce serious errors—for example, translating Portuguese place names as ordinary English words.

AGTP matches geographic records using the official municipality identifier included in the source data, then restores the canonical Portuguese name. We do not join or compare historical records solely by their translated display name.

What period does each report cover?

Every market report should identify:

• The reporting period.
• Whether the period is monthly, quarterly or annual.
• The date on which AGTP retrieved the source file.
• Whether the data is preliminary or definitive.
• The selected property type, typology and condition filters.

A file downloaded in August may describe a quarter ending in June. The download date is not the reporting period.

Historical editions remain tied to the period originally analyzed. When a newer report is published, AGTP does not silently replace an older report's conclusions with current figures.

Preliminary and definitive quarterly data

Confidencial Imobiliário uses a two-stage quarterly dissemination process.

Preliminary data is published first and is based primarily on information reported by real estate brokerage networks, which are the main contributors supporting the early release.

Definitive data for quarter t is published when preliminary data for quarter t+1 is released. The definitive version draws from the broader universe of operators integrated into SIR, including other brokerage businesses, real estate developers, investors and credit institutions.

CI notes that preliminary results are less representative of new housing. Its preliminary estimate of transactions therefore uses the structure of definitive data from the preceding quarter.

AGTP will retain preliminary and definitive releases as separate internal data vintages. When a public report uses preliminary results, we will say so. If the status has not been confirmed, we will label it data status unconfirmed rather than guessing.

How AGTP handles monthly, quarterly and annual comparisons

Two numbers should be called a growth rate only when their periods and definitions are genuinely comparable.

For example, comparing a full-year 2025 average with a Q2 2026 average may reveal that the latest quarter sits above or below the earlier annual baseline. It does not, by itself, establish annual price growth.

Where we make such a comparison, AGTP uses this disclosure:

Directional comparison—not a growth rate. The latest quarterly average is being compared with a full-year annual average. The difference may reflect timing and the mix of properties transacted as well as underlying market movement.

As we collect equivalent periods over time, we will distinguish among:

• Month-over-month change.
• Quarter-over-quarter change.
• Same-quarter year-over-year change.
• Full-year change.
• Rolling annual measures.

We will not substitute one for another merely because only one comparison is currently available.

Understanding the principal indicators

Estimated residential transactions

This indicator estimates the total number of residential property transactions in the market.

CI bases the calculation on the most recently published INE transaction statistics and projects them forward using the change in activity observed through major companies participating in SIR. CI then uses the SIR sales structure to segment INE's broader regional statistics into smaller geographies and property categories.

What it can tell us

• The estimated scale of residential transaction activity.
• How activity is distributed across selected markets and property segments.
• How a market's activity changes when comparable periods are available.

What it cannot tell us by itself

• The number of raw transaction records supporting every price statistic.
• Whether an individual market is liquid at a particular price point.
• The probability that a specific property will sell.
• Complete official registry activity unless CI explicitly identifies the value that way.

Average transaction price

CI defines the transaction price as the amount paid by the buyer for the residential property, excluding taxes. When a transaction results from a real estate intermediary's activity, the price corresponds to the amount on which the mediation commission is calculated.

AGTP generally presents this as average transaction price or average sale price.

An average is not the same as a median. A relatively small number of high-value or unusually large transactions can move an average. A change in the kinds of properties sold can also change the average even if the value of an otherwise identical property has moved less.

We therefore avoid calling an average the “typical home price” unless an appropriate median is available.

Transaction price per square metre

CI calculates transaction price per square metre as:

Transaction price ÷ gross private area

This is a completed-transaction measure, not an advertised asking-price measure.

Price per square metre is useful when comparing broad price levels, but it does not normalize every difference among properties. Condition, floor, elevator, light, outdoor space, parking, view, renovation quality, energy performance and exact location can all materially affect value.

Why Portuguese area definitions may surprise Americans

The denominator is gross private area, known in Portuguese as área bruta privativa or ABP. Under the Portuguese IMI framework, this is the total surface measured by the exterior perimeter and the axes of walls or other separating elements. It can include enclosed private balconies, private basements and private attics with use equivalent to the dwelling.

This is not necessarily equivalent to the interior living-area measurement an American buyer may expect. Portuguese €/m² and U.S. price-per-square-foot comparisons should therefore be treated as related concepts, not mechanically interchangeable measures.

Gross private area and gross dependent area

Gross private area includes the primary private area measured using the exterior perimeter and wall axes, together with certain enclosed private spaces used in the same way as the dwelling.

Gross dependent area refers to covered and enclosed areas for exclusive use whose function is accessory to the dwelling. Depending on the property, these may include garages, parking spaces, storage rooms, accessible attics or basements, animal facilities and certain balconies not included in gross private area.

CI's transaction €/m² indicator uses gross private area, not gross dependent area.

Listing-to-transaction time

CI defines absorption time as the average number of months between a property being listed and the transaction.

For American readers, AGTP generally describes this more plainly as:

Average time from listing to transaction

This is a marketing-time measure. It is not “months of housing inventory,” and it should not be interpreted as an absorption rate in the sense commonly used in some U.S. market reports.

The source reports the measure in months. When AGTP offers an approximate day equivalent, we label the conversion as approximate rather than implying the source measured individual days.

Final discount rate

The final discount rate measures the difference between the completed sale price and the last asking price in effect when the transaction was finalized.

CI calculates the discount property by property and reports the average percentage difference.

This metric is useful for understanding the final stage of negotiation, but it is not an instruction to offer that percentage below the current asking price of every property. A newly listed, accurately priced home and an over-priced property that has already spent months on the market require different strategies.

Cumulative original-ask-to-sale adjustment

CI's cumulative discount and revision rate measures the difference between the initial asking price and the completed sale price. It therefore combines:

• Any asking-price revisions made while the property was on the market.
• The final discount negotiated from the last asking price.

The two discount measures describe different parts of the transaction journey:

Original asking price → price revisions → final asking price → final negotiation → transaction price

This distinction is especially useful for offer strategy. The current asking price does not always reveal how far a seller has already moved from the property's original position.

Market gap

Market gap compares the average transaction price of properties sold with the average asking price of all properties available for sale in the relevant market.

It differs from a transaction discount because the asking-price side includes the broader pool of available properties—not only the homes that completed a transaction.

A negative market gap can indicate that properties changing hands occupy a lower price level than the average available supply. It can also reflect differences in the composition of properties offered and properties sold, or a degree of price misalignment between effective demand and asking prices.

It does not mean individual buyers received that percentage off the asking price. AGTP will never convert a market gap into a recommended offer discount.

Asking-price revision rate

CI's review or revision rate measures changes owners make to their asking prices. It averages the percentage change in asking prices across two consecutive quarters and includes only properties whose asking prices changed.

This differs from the percentage of properties revised, which measures how many properties in the offering pool experienced a price revision during the quarter.

Where these measures are not included in a source export, AGTP will not attempt to reconstruct them from other indicators.

New and used properties

CI's definition of a new property includes:

• A property sold first-hand by the developer.
• Certain resales occurring no more than five years after construction, reconstruction or rehabilitation was completed.
• Properties sold while still in planning or under construction.

“New” should therefore not automatically be read as “never previously occupied” without examining the context of the specific property.

AGTP uses used property for the corresponding existing-property segment in the source data. We may use “resale” in explanatory prose when it improves clarity, while preserving CI's underlying classification.

Apartments, houses and bedroom typology

CI draws on Portuguese statistical and building terminology.

An apartment is a dwelling within a building containing more than one dwelling, whose main entrance generally opens onto a staircase, corridor or courtyard.

A house is a dwelling in a detached, semi-detached or row building corresponding to one family accommodation unit, whose main entrance usually leads to a street, courtyard or surrounding land.

Portuguese typology is based on the number of bedrooms:

• T1: one bedroom.
• T2: two bedrooms.
• T3: three bedrooms.
• T4: four bedrooms.

Source exports may group T1 with studios or smaller properties, and T4 with larger typologies. Each AGTP report will identify the grouping used rather than implying a more precise classification.

Transaction date

CI assigns the transaction to the year and month in which the property became subject to a Contrato-Promessa de Compra e Venda, or CPCV—the preliminary purchase and sale agreement. When there is no preliminary agreement, CI uses the deed date.

This means a report should not necessarily be described as a count of deeds signed during the period. AGTP uses “transactions assigned to the period” unless the source provides a more specific basis.

Updated selling prices and historical comparability

CI explains that sale prices made available through its Micro-SIR system are current or updated sale prices. For user-defined periods of up to 24 months, prices can be homogenized using forecasting models and local price indices produced by Confidencial Imobiliário.

This is intended to improve temporal comparability, but it also means that an exported historical value may not always be a purely nominal average frozen at the original transaction date.

Before describing a multi-period difference as appreciation, AGTP seeks to establish:

• Whether the export uses updated or homogenized prices.
• The reference date applied to the values.
• Whether historical exports may be revised later.
• Whether the compared periods use the same method.

If those questions remain unresolved, we will say so.

Values by range

Where CI provides price ranges, its labels correspond to the following points in the distribution:

• Low: 5th percentile.
• Lower middle: 25th percentile, or first quartile.
• Middle: mean.
• Upper middle: 75th percentile, or third quartile.
• High: 95th percentile.

The “middle” value in this convention is the mean—not necessarily the median. AGTP will preserve that distinction.

Missing data does not mean zero

Some geographies and market segments contain an estimated transaction count but no published price, marketing-time or negotiation metric. This can occur when a source does not provide enough information for that indicator or suppresses a result.

AGTP displays such values as:

Insufficient source data

We do not replace missing values with zero, carry a nearby municipality's value into the gap or infer a figure from a broader region.

Where the raw number of records supporting a statistic is not provided, AGTP does not use CI's estimated total transaction count as though it were the statistical sample size behind that metric.

How we calculate and label derived figures

AGTP sometimes performs straightforward calculations using selected CI source values. Examples may include:

• Percentage difference between two comparable reported values.
• Premium of a new-property average over a used-property average.
• Difference between apartment and house averages.
• An illustrative area obtained by dividing a budget by an average €/m².

Derived values are labeled as AGTP calculations. For example:

Source data: Confidencial Imobiliário, SIR. Calculations by The American's Guide to Portugal. © IMOESTATISTICA – TODOS OS DIREITOS RESERVADOS.

An illustrative “what a budget buys” calculation is not a property forecast. Dividing €500,000 by a municipality's average €/m² does not account for taxes, acquisition costs, condition, property mix or the features of an individual home.

Our editorial framework: number, meaning, action and boundary

AGTP analysis is designed around four questions.

The number

What does the source actually report, for which market, period and property segment?

The meaning

What does the comparison help us understand about price level, transaction activity, property size, marketing time or negotiation?

The action

How might that context help an American buyer or seller ask better questions or make a more informed decision?

The boundary

What does the number not prove—and what additional property-specific analysis is still required?

This final step matters. Market statistics can inform a valuation or offer strategy, but they do not replace a comparative market analysis, document review, physical inspection, appraisal or appropriate legal, tax, mortgage and technical advice.

Why AGTP publishes selected data rather than a database

AGTP uses CI information as supporting evidence for editorial market analysis. We intentionally publish a limited selection of figures relevant to the question being examined rather than reproducing CI's database.

We do not provide:

• Raw or cleaned CI workbooks.
• A downloadable reconstruction of CI data.
• A public API or data feed.
• Complete all-municipality tables.
• Filters that allow readers to recreate the source database.
• An exhaustive public archive of every CI observation.

This keeps AGTP focused on interpretation and respects the rights and commercial role of the underlying data provider.

Corrections, revisions and transparency

Market data can be revised. Definitions can change. Preliminary releases can be replaced by definitive results. AGTP may also identify a labeling, calculation or interpretation error after publication.

When a correction materially affects a conclusion, we will update the page and explain the correction. When a source revises an earlier period, we will distinguish the revised result from the value used in the original historical edition whenever that distinction matters.

Every report should include a publication date and Last Reviewed date. Substantive methodology changes should also be recorded on this page.

A market average is context—not a property valuation

Portugal's real estate market is highly local. Two properties with the same nominal area can have very different values because of their exact location, legal condition, state of repair, layout, light, noise, views, outdoor space, parking, elevator access, energy performance, condominium condition and documentation.

Market-level information is useful because it gives buyers and sellers a frame of reference. It can reveal whether one municipality tends to offer more space for the money, whether houses have historically taken longer to transact than apartments, or whether price revisions have been an important part of the selling process.

It cannot determine the correct price or offer for a particular property on its own.

Sources and further reading

AGTP's current methodology draws principally on:

• Confidencial Imobiliário and the SIR — Sistema de Informação Residencial: https://www.confidencialimobiliario.com/en
• Confidencial Imobiliário Concepts and Glossary, as supplied with the source-data documentation.
• Confidencial Imobiliário Terms and Conditions of Use, as supplied with the source-data documentation.
• Instituto Nacional de Estatística: https://www.ine.pt/
• Portuguese Tax Authority guidance on gross private and gross dependent area: https://info.portaldasfinancas.gov.pt/pt/apoio_contribuinte/questoes_frequentes/pages/faqs-00461.aspx

Copyright notice:

© IMOESTATISTICA – TODOS OS DIREITOS RESERVADOS

Confidencial Imobiliário/Imoestatística is the source of the underlying CI/SIR content. The American's Guide to Portugal is responsible for its own selection, calculations, explanations and interpretations. CI/Imoestatística is not responsible for the adequacy or accuracy of the surrounding AGTP analysis in which selected content is incorporated.

Questions about the methodology?

If a term is unclear or you believe a figure has been interpreted incorrectly, please contact AGTP. Good market analysis should be understandable enough to question—and transparent enough to correct.