Short-term rental yield in France: the real calculation

The real return on a short-term rental is the net operating result, after every cost and after tax, divided by the capital actually committed — purchase price, transaction costs, works and furnishing included. On that basis, a significant share of properties does not beat standard letting once the owner's time is priced in.

The gross yield everyone quotes simply divides annual rent by purchase price. It ignores operations, tax, and the capital actually tied up. In short-term letting, where operating costs represent a far larger share than in unfurnished renting, that indicator measures nothing useful.

One example measures the gap. On a €200,000 property generating €24,000 of receipts, gross yield reads 12%. Add €16,000 of transaction costs and €12,000 of furnishing to the denominator, take €14,000 of costs and tax out of the numerator, and the net yield after tax lands around 4.3%. That is not the same investment, and it is the second figure that compares to any other asset class.

ADR is revenue divided by the number of nights actually sold. It measures how the property is valued, not how it performs: you can post a high ADR while selling almost nothing. On its own it is the easiest indicator to improve and the most misleading.

This is nights sold over nights offered for sale. The subtlety sits in the denominator: a calendar closed half the year mechanically produces a handsome occupancy rate. Comparing two occupancy rates without knowing how many nights each had open is meaningless.

RevPAR is revenue per available night: ADR multiplied by occupancy. It is the only one of the three that cannot be gamed, because it carries both price and fill. An operator steering by occupancy cuts prices until the calendar fills; an operator steering by RevPAR accepts empty nights when they cost less than a discounted one.

There is nothing exotic in the list below: these are ordinary costs, present in almost every file. They are simply absent from most online estimators, whose calculation stops at gross revenue.

Furniture replacement is the single most systematically ignored item. A short-term rental takes in one year what a standard let takes in five. Mattresses, sofas, crockery, linen and small appliances do not last: they get replaced, and that spend must be provisioned every year, not discovered the day it lands.

This point deserves precision, because it governs how every public figure in the sector should be read. The occupancy and average price data published by market aggregators are neither declared nor audited: they are estimates reconstructed by observing listings' public calendars. A night blocked by the host for personal use looks, from the outside, exactly like a booked night.

• Survivorship bias: listings that fail are withdrawn and therefore leave the sample. What remains is what worked.
• Confusion between a blocked night and a sold night, which mechanically inflates observed occupancy.
• City averages applied to one specific property, when the gap between two flats on the same street often exceeds the gap between two cities.
• Gross revenue presented as income, with no deduction for platform fees, cleaning, or tourist taxes collected and passed on.

The Le Meur act rewrote the tax arithmetic of short-term rentals, and the effect is far from marginal for smaller operators. The micro-BIC allowance for unclassified short-term rentals has fallen to 30%, with a receipts ceiling of €15,000. Above that, moving to the actual-expenses regime is compulsory.

The consequence is counter-intuitive: for many operators, moving to the actual-expenses regime is a gain rather than a constraint. Depreciating the property and the furniture often wipes out taxable profit entirely for several years. The accounting cost is real, but it compares against the tax avoided, not against zero.

This is the most underestimated risk, because it never appears in a yield spreadsheet. A property bought at the price of a short-term rental asset, in a municipality that later restricts that use, does not merely lose income: it resells at the price of a standard rental asset. The loss is in capital, not in revenue.

• Registering a short-term rental through the national online service has been compulsory in every municipality since 20 May 2026, on pain of a civil fine of up to €10,000, rising to €20,000 for a false declaration.
• The 120-night annual cap on letting a main residence can be lowered to 90 nights by municipal council decision. Lyon took that step with effect from 1 January 2026.
• Change of use is governed by article L.631-7 of the construction and housing code and exposes the owner to a civil fine of up to €100,000 per dwelling.
• An energy performance certificate rated A to E is required until 31 December 2033; class D becomes the minimum on 1 January 2034.

The energy deadline deserves particular attention because it is dated, and therefore quantifiable. A property rated E today can be operated until the end of 2033 and cannot be from 1 January 2034 without works. This is not a vague risk: it is a known cost on a known date, and it belongs in the funding plan at the moment of purchase.

The three cases below are constructed to illustrate three different outcomes, from stated assumptions. They describe no real file and carry no weight as market benchmarks.

A 28 sq m studio bought for €150,000, €12,000 of transaction costs, €10,000 of furnishing: €172,000 committed. Assumptions: 330 nights open, 70% occupancy, €85 ADR, giving €19,635 of receipts. Operating costs, management, tax and renewal provision: €11,200. Net result €8,435, or 4.9% of committed capital. It works because demand is year-round and a RevPAR of €59.50 absorbs a high turnover.

A 45 sq m flat bought for €210,000, €17,000 of costs, €14,000 of furnishing: €241,000 committed. Assumptions: a four-month high season at 85% occupancy, eight months at 25%, €95 ADR. Receipts €20,425, costs and tax €13,900. Net result €6,525, or 2.7%. The same property let annually at €780 would produce roughly €8,100 net for incomparably less management. Short-term letting destroys value here, and no listing optimisation fixes underlying seasonality.

A 70 sq m flat near a university hub, €230,000 plus €18,500 of costs. As a short-term rental: €24,000 of receipts, €15,800 of costs and tax, so €8,200 net. As a three-room house share: €19,800 of receipts, €6,900 of costs and tax, so €12,900 net. Short-term letting collects more and returns less. This is the most frequent case among those people believe settled in advance.