France's Energy Rating Ban: What 14,000 F/G Sales Reveal One Year In
G-rated apartments were banned from new rental agreements in France on 1 January 2025. English-language coverage predicted price collapses and fire sales. We tracked every F and G apartment transaction across France through the end of 2025 to see what actually happened.
TL;DR
In 2025, 14,051 F and G apartments sold in France (+2% vs 2024). The average G median price moved +1.7% year-on-year. No panic. No fire sale. The market absorbed the G ban with no observable disruption. But the F ban of January 2028 affects 9,608 annual transactions, 2.2 times the G-rated volume, and shows no sign of being priced in yet.
The law: three phases, three deadlines
The Loi Climat et Résilience (2021) created a rolling schedule of rental prohibitions based on DPE energy rating. The ban applies to new tenancy agreements only: existing leases for G-rated properties were not voided in January 2025, and they can still be renewed under the old tenant. The full prohibition kicks in when the current tenant vacates.
Banned from new leases: 1 Jan 2025
Cannot sign a new lease. Existing tenancies are exempt until renewal. Buyers purchasing to live (not rent) are unaffected.
Banned from new leases: 1 Jan 2028
Planned under current legislation. Affects 2.2× more annual transactions than G. No evidence this is priced into the purchase market yet.
Banned from new leases: 1 Jan 2034
8 years out. E-rated properties represent the largest share of the affected stock and currently show near-zero purchase price discount vs D.
One year of data: the market did not blink
We tracked every F-rated and G-rated apartment sale recorded in our DVF×DPE dataset from Q1 2023 through Q4 2025. The vertical line marks Q1 2025, the first quarter under the G ban.
Source: DVF DGFiP × DPE ADEME (dpe_dvf_matches), apartments, prix_m2 between 800 and 25,000 €. Q1 2023 to Q4 2025. N = 27,833 total F+G transactions over the two-year period shown.
G-rated sales in Q4 2024 (the last pre-ban quarter): 1,204. In Q1 2025 (the first quarter under the ban): 1,000. A dip, but not a cliff. By Q3 2025, volumes recovered to 1,193 — matching pre-ban levels.
The annual comparison is equally clear. 4,443 G-rated apartments sold in France in 2025, compared to 4,544 in 2024, a 2% decline that is well within normal quarterly variance. F-rated sales, in contrast, increased 4% year-on-year to 9,608 transactions.
The price story is similarly undramatic. The average quarterly G median was 3,927 €/sqm in 2024 and 3,995 €/sqm in 2025 — effectively flat, with quarter-to-quarter variance larger than the year-on-year change.
City exposure: where F/G concentrates
The proportion of F/G apartments in sales varies sharply by city. Paris stands out, with one in four apartment sales carrying an F or G rating. The table below compares the F/G share before and after the G ban, and the F/G price median for each period.
| City | % F/G (2024) | % F/G (2025) | Change |
|---|---|---|---|
| Paris | 25.7% | 25.2% | -0.5 pp |
| Lille | 9.9% | 10.1% | +0.2 pp |
| Rennes | 10.3% | 9.2% | -1.1 pp |
| Nantes | 7.7% | 7.9% | +0.2 pp |
| Bordeaux | 7.9% | 6.8% | -1.1 pp |
| Toulouse | 7.8% | 6% | -1.8 pp |
| Nice | 6.9% | 6% | -0.9 pp |
Apartments, January 2024 to December 2025, prix_m2 between 800 and 25,000 €. Source: dpe_dvf_matches (DVF DGFiP × DPE ADEME). Marseille and Lyon not shown (Marseille: 4.8% F/G share, below query threshold for 2024-2025 window; Lyon similar low share). pp = percentage points.
Paris: 25% F/G — stable through the ban
The share barely moved (25.7% to 25.2%). Paris F/G medians rose 2% year-on-year, from 9,607 to 9,800 €/sqm. The Haussmann buildings that dominate the F/G pool are irreplaceable by location, so owners did not lower prices to compensate for the rental restriction.
Lille and Rennes: highest shares after Paris
Lille (10.1%) and Rennes (9.2%) have the highest F/G concentrations outside Paris, driven by large stocks of pre-war stone buildings. Lille saw the only meaningful F/G price decline in our dataset (-7%), while Rennes saw prices rise +7.5%. Neither moved in a direction that suggests panic.
Why purchase prices did not react: the location paradox
In most French cities, G-rated apartments sell for the same price per sqm as A-rated ones, or for more. This is not a malfunction of the market. It reflects the fact that F/G properties are concentrated in pre-1975 stone construction in city centres, while A/B properties are almost exclusively post-2012 new-builds on city fringes.
When a buyer compares city-level medians, they are comparing addresses, not insulation. A Haussmann apartment on the Ile Saint-Louis (G, 12,000 €/sqm) does not compete with a new-build studio in Evry (A, 3,200 €/sqm). The rental restriction removes one exit route (letting it out) but does not make the address less desirable to owner-occupiers.
The F ban of 2028: the larger, unpriced risk
If the G ban caused no visible market disruption, should investors ignore the F ban of January 2028? Probably not, for three reasons.
2.2× the volume
In 2025, 9,608 F-rated apartments sold in France, compared to 4,443 G-rated. When the F ban takes effect, the affected pool of rental-restricted properties doubles overnight. A larger pool means a larger group of investors who need to either renovate or sell, even if the average owner-occupier buyer remains indifferent.
Shorter renovation window
The G ban gave the market a 3-year signal (legislation passed 2021, ban effective 2025). The F ban gives the same 3-year window from now. Renovation lead times for an F-to-C upgrade typically run 6-18 months, meaning investors who want to exit before the ban need to start planning in 2026 to avoid owning a rental-restricted property at 1 January 2028.
Price signal is absent so far
The F/G medians in our dataset show no systematic discount for F vs D or C properties at the city level. If the market were beginning to price in the 2028 risk, we would expect F-rated properties to start trading at a measurable discount to equivalent C-rated properties within the same postal code. We do not see this in the data yet.
Construction era: the fastest way to estimate DPE exposure
The DPE label of a specific property is shown on any French listing and available via the Normi API. But when evaluating a neighbourhood or a portfolio, the construction era is a reliable proxy. Based on 23 million DPE diagnoses from the ADEME national database:
| Construction era | DPE records | % F or G | % A or B |
|---|---|---|---|
| 2013+ | 3.5M | 0.1% | 65.6% |
| 2001-2012 | 1.9M | 1% | 9.4% |
| 1989-2000 | 1.6M | 3.3% | 4% |
| 1975-1988 | 2.4M | 7.3% | 4.2% |
| 1949-1974 | 4.0M | 13% | 4.5% |
| 1914-1948 | 3.1M | 18.7% | 3.8% |
| Pre-1914 | 0.5M | 18.4% | 4.1% |
Source: ADEME, 23 million DPE diagnoses (national base DPE-DL). All residential property types. Used in Normi's analysis of the passoires thermiques paradox.
The breakpoint is 2013: post-RT2012 construction has a 0.1% F/G rate and a 65.6% A/B rate. Everything built before the thermal regulation of 2012 carries meaningful F/G exposure, with the rate rising sharply for pre-1914 and 1914-1948 stock (both above 18%). If you are underwriting a pre-war property in France, assume F/G until a DPE confirms otherwise.
Investor playbook
Buying a G-rated apartment today
- •You cannot sign a new lease unless you first renovate to at least F. If the property already has a tenant, the restriction only applies when that lease ends.
- •In most cities, the G label does not compress the purchase price vs an equivalent C in the same location. Negotiate on renovation cost, not DPE label alone.
- •Use GET /v1/property-history to verify the property's transaction history, and GET /v1/stats/dpe-premium for the local G vs A median in the exact postal code.
Buying an F-rated apartment today
- •The F ban takes effect January 2028. You have roughly two years to renovate before new rental agreements are restricted.
- •Renovation to reach a C rating typically costs 25,000–50,000 € for a typical apartment (source: ANAH published benchmarks). For a 60 sqm apartment at 4,400 €/sqm, that is 5–19% of the purchase price — material but not prohibitive.
- •Factor renovation cost into the acquisition price and negotiate accordingly, especially in cities where F/G share is declining (Toulouse: -1.8 pp).
Targeting cities where DPE does affect price
- •Marseille is the one major city where the DPE hierarchy matches the price hierarchy: A fetches 4,298 €/sqm, G fetches 2,871 €/sqm (-33%). In Marseille, buying a G and renovating to C is a genuine value-creation play.
- •Lyon shows a milder version of the same pattern (A at 5,240 vs G at 4,914 €/sqm). These are the two cities where the DPE label operates as the market expects.
Portfolio-level risk screening
- •If you hold a portfolio of French rental properties, the F ban affects everything rated F or below. Use GET /v1/stats/dpe-premium for each commune to quantify the discount gap and prioritise which properties to renovate first.
- •Paris portfolios carry the highest absolute exposure (25% F/G share) but also the highest absolute values. A 10% renovation cost on a 10,000 €/sqm property is structurally different from the same cost on a 3,500 €/sqm property.
Pre-purchase DPE due diligence with the Normi API
Three API calls cover the key questions before making an offer on any French property.
Local DPE price gap
Returns the median price difference between A/B and F/G properties in the target commune or postal code. Use this to calibrate whether the DPE is priced locally or ignored.
View docs →Property transaction history
Full sale history for a specific address since 2014. Verify the price paid, the surface, and the number of prior transfers. Essential for checking if the seller is pricing above or below prior transaction.
View docs →Comparable sales nearby
Returns recent transactions within a configurable radius (100-1000 m). Filter by DPE class to compare F-rated sales with C-rated sales in the same block — the most granular measure of local DPE discount.
View docs →Methodology
Sources: DVF DGFiP (Licence Ouverte 2.0, data.gouv.fr, property mutations 2014-2025) cross-referenced with DPE ADEME (open data, base nationale DPE-DL). DPE by construction era uses the full 23-million-record ADEME database.
Quarterly volume analysis: apartments (computed_type_local = 'Appartement'), prix_m2 between 800 and 25,000 €, date_mutation Q1 2023 to Q4 2025. Aggregation: COUNT(*) and PERCENTILE_CONT(0.5) by quarter and DPE class. The dataset covers properties with a DPE match (approximately 38% of all DVF transactions), so absolute volumes represent the matched sample, not total market transactions.
City-level analysis: same filters, January 2024 to December 2025, for 7 major cities. Pre-ban period = 2024 (full year), post-ban period = 2025 (full year).
Renovation cost benchmarks: cited from published ANAH (Agence Nationale de l'Habitat) reference guides — not from Normi data. Normi does not have renovation cost data.
Full raw queries: content/blog/france-dpe-energy-rating-ban-investor-guide.data.json
Run DPE analysis for any French commune
Every figure in this article is reproducible with a free Normi API token. Use GET /v1/stats/dpe-premium to get the local F/G discount for any commune, or GET /v1/comparables to compare DPE-filtered sales within walking distance of a target property. Free tier: 500 credits per month. No credit card required.
Related analyses
Sources: DVF DGFiP (Licence Ouverte 2.0) · DPE ADEME (open data) · Loi Climat et Résilience n° 2021-1104 · ANAH renovation cost benchmarks.
The Normi team — 11 June 2026