Rivesta barometer of property auctions and court sales in Portugal: November 2026

On 5 November 2026, Rivesta tracked 1,701 properties for sale at auction, through court sales and by banks in Portugal. This is the first edition of the barometer: a snapshot of those listings, cross-checked with INE housing statistics, with six readings on what is for sale, where it weighs most, how it is sold and at what starting price.

Three in four of these properties (74.8%) are court sales published on the official portals, 22.5% are with auction houses and 2.6% are bank-owned. The barometer therefore mostly describes court sales.

1. Most listings are land and whole buildings; only a quarter is housing

Land and rural plots make up 40.0% of active listings. Housing, flats and houses, is 24.9%, that is 424 properties, and whole buildings 21.1%. Retail, offices and warehouses add up to 5.4%, garages and storage to 2.1%, and 6.5% of listings do not state the property type.

For anyone looking for a home, the useful universe is much smaller than the headline number suggests: about one property in four.

Source: Rivesta, tracked listings. Data as of 5 November 2026. Portugal.

2. Against each regional market, these channels carry more housing in the Centre and less in Greater Lisbon

To compare regions of different sizes, Rivesta divides active housing in each region by the dwelling sales recorded by INE in the 12 months to Q2 2026. Nationally, there are 2.7 properties per 1,000 sales.

The Centre (4.2, with 108 properties) is clearly above that average and Greater Lisbon (1.5, with 43) clearly below, even allowing for the margin of error of samples this size. The Alentejo (4.0) also comes out above, but rests on 34 properties and the gap may still be chance. Oeste and Tagus Valley (2.9) and the North (2.8) are close to the average. The Setúbal Peninsula, the Algarve, the Azores and Madeira each have fewer than 30 active housing listings and are shown together, at 2.0.

This is a relative scale, since it compares one day's stock with a year of sales. It is neither a market share nor a measure of default: it also depends on which sales reach the channels Rivesta tracks.

Sources: Rivesta, tracked listings on 5 November 2026; INE, dwelling sales in the 12 months to Q2 2026. Portugal, NUTS II regions.

3. Seven in ten court sales on the official portals are already under private negotiation

Of the 1,273 active court sales on the official portals, 72.4% were under private negotiation rather than at auction. Across all tracked listings, the share is 60.1%.

Portuguese law provides for private negotiation in several cases, including when an electronic auction or a sealed bid sale fails. For the buyer, the process changes: instead of bidding up to a deadline, they negotiate an offer with whoever runs the sale, within the set period.

The reading has a snapshot effect: these listings stay on sale longer than an auction, so they build up in a single day's stock.

Source: Rivesta, tracked listings. Data as of 5 November 2026. Portugal.

4. At least one property in six already went to auction without a sale and came back

Among active listings, 269 properties, 15.8%, came back to the market after an auction that ended without a sale: no bids, below the minimum, cancelled or withdrawn. The share is similar at auction (16.3%), under private negotiation (17.9%) and in housing (16.7%).

This is a lower bound. Rivesta only counts a second round when it knows the outcome of the previous auction, and a listing that disappears with no known outcome is left out.

Source: Rivesta, tracked listings. Data as of 5 November 2026. Portugal.

5. Housing starting prices vary widely against the INE reference

For 256 housing properties with a floor area and a municipal reference, Rivesta compared the minimum starting price with the value given by the INE municipal median times the floor area. The first reading is the spread: half of the cases sit between 50% and 111% of the reference, 44.5% are below 70% and one in three (32.0%) is at or above it.

The median is 75%, and the extreme cases do not drive it: without the 26 properties below 20% or above 200% of the reference (undivided shares of a property, for example), it stays at 75%. Flats (75%, across 108 properties) and houses (76%, across 148) barely differ. By region, only the North (83%, across 82 properties) and the Centre (77%, across 59) have a large enough sample to be read separately.

This is not a discount on the final price, which is set in the sale and can end above the minimum, nor does it mean these properties are cheap. The INE reference describes sales across the whole municipality and does not see each property's condition, occupancy or charges. It places the starting price in context; it does not value a specific home.

Sources: Rivesta, tracked listings on 5 November 2026; INE, median sale value per m², 12 months to Q2 2026. Portugal, by municipality.

6. At auction, the sale is set at the median 32 days after the listing appears

For sales scheduled from June to October 2026, across 2,205 auctions, the sale date fell, at the median, 32 days after Rivesta first saw the listing. Half fell between 28 and 48 days, and only 5.7% had 21 days or less. On the official portals the median is 34 days and with auction houses 31.

Under private negotiation the period up to the deadline for offers is longer and more variable: 46 days at the median, across 1,258 sales, with half between 30 and 68.

In practice, a bidder has about a month to visit, read the documents and arrange financing.

Source: Rivesta, tracked listings, sales scheduled from June to October 2026. Portugal.

How it was done

The listings are those Rivesta tracks, as a property investment analysis and intelligence platform, on the official court sale portals, at auction houses and on bank sale channels. This is not the whole market: a sale that none of these sources publishes is left out, and coverage is not the same in every category. Sale outcomes are not included yet, because only some sources publish them reliably. The full methodology, samples and sources are in this article's method note.