Rental Yield Evolution — Algeria (2023 → 2025)

The evolution of gross rental yield between the 2023-2024 and 2025-2026 editions of the Algerian tax authority's price reference. 45,549 properties present in all four files — sale and rental, at both editions — with their yield at each date and the gap in points.

The finding: the median moves from 4.99 % to 4.78 %. Rents did not quite keep pace with sale prices. 19,187 properties see their yield fall, 10,695 rise, 15,667 stay put.

Why this product is not inflation-adjusted, and why that is the right call. A yield is a ratio between two amounts from the same year — annual rent over sale price. Inflation acts on both terms and cancels out. Restating these rates in constant dinars would have produced exactly the same numbers plus a column implying a treatment that never happened. The nominal yield IS the real yield. For inflation adjustment, our "Real Estate Sale Price Evolution" product is the one you want.

A coherence column, and it is not cosmetic. The reference contains mismatched pairs: in El Bayadh, bare land sells at 97 DA/m² and rents at 806 DA/m²/month in the same document — the price covers land, the rent evidently covers something else. The arithmetic then yields 9,977 %.

Two thresholds apply, only one of which is a setting:

• above 100 %, annual rent exceeds the sale price: the asset would repay itself in under a year. That is not a high yield, it is an arithmetic impossibility. 675 rows, marked incoherent;
• between 20 and 100 %, a grey zone: too high for ordinary real estate, not enough to be demonstrably wrong. 3,246 rows, marked a_verifier, concentrated on agricultural land.

These rows are flagged and kept, never deleted. I do not know which ones are wrong, and dropping them silently would be worse than letting you see them: you filter on the coherence_2025 column. 41,628 properties are marked plausible.

What was excluded before the calculation. Any property key matching several rows within one source file — which would be the right one? — and any pair whose base unit differs (square metre against hectare). Excluded and counted, never resolved at random.

Mind what the reference measures. These are administrative brackets used as a basis for tax assessment, not observed transactions. A yield gap says the administration revised its rental and sale scales at different rates; that is a real signal about the market, it is not the market itself.