Stagnation in Azerbaijan's property market: developers keep raising prices while sales stand still

Studies of Azerbaijan's property market usually show what developers and owners are asking for a property - whether an apartment, a house or commercial space - rather than what buyers actually pay for it.

The second of those figures barely exists in the public domain in Azerbaijan at all.

Yet if you set the price charts against the only available indicator of volume - the number of registered property rights - the picture of recent years looks nothing like the way it is usually described. Apartments gained roughly ten percent in price over the year, while the number of apartment registrations did not move by a single percent.

For land, a record plot price comes alongside a fall in registrations of an eighth. Mortgage lending is not growing, and the number of technical passports issued is falling.

Physical supply is contracting at the same time: investment in housing construction fell by almost a third in early 2026.

The only segment showing a marked rise in volume is private houses - but that figure, as will become clear, needs to be read with care.

In this article we set out to explain why prices are not falling despite frozen volumes, why an apartment yields roughly three times less in cash than an ordinary bank deposit and is still considered a sound investment, and finally why nowhere in the chain of listing - valuation - collateral is there a single point at which a buyer confirms the price.

And above all - why the figure whose absence we are discussing does in fact exist.

Since February 2023 all settlements for property purchases in Azerbaijan have been made by non-cash transfer. That means the entire banking system sees real transaction prices. The problem is not that the data is missing. The problem is that it is not published.

This is not a story about a bubble: a bubble requires leverage, which this market does not have.

It is a story about a market that is neither rising nor falling but standing still - and that looks as though it is growing only because it is measured by the single figure the supply side reports about itself.

* * *

In August 2026 ABB (the International Bank of Azerbaijan) released the fifth issue of its study of the Baku property market. Thirty-six charts, a breakdown across ten districts, separate sections on land, private houses, commercial space and rentals.

It should be said at the outset that this is rare and valuable work. Azerbaijan has almost no institutions that publish sectoral analysis systematically and regularly in the public domain - not only on property, but across most sectors of the economy. The bank has now done so for five issues running, free of charge, with transparent sourcing and stated methodological limits.

A significant share of the figures cited below is available only thanks to this report. The analysis that follows is possible because ABB has taken on work that nobody else takes on, and arguing with the substance is far more useful to the market than having no substance to argue with.

Even so, one thing is missing from the report. The number of transactions.

Not a single figure on how many apartments in Baku were bought and sold. The "sövdələşmə payı" indicator in Table 36 is not the share of transactions but the size of the negotiating discount.

In other words, this is a report on a market that contains no market volume.

This is not a quibble about presentation. Without volume a price chart says almost nothing. A price at which no transaction takes place is not a price but an asking figure. And once the missing number is put back in, the picture changes fundamentally.

What the statistics missing from the report show

According to the State Service for Property Issues, 115,613 property rights were registered across the country in January-April 2026 - 3.8% fewer than in the same period a year earlier. Calculations based on the Service’s four-month statistics indicate that between 2023 and 2026 the figure for the comparable period fell by roughly a third, from 179,416.

The breakdown explains more than the total does:

Now compare this with the ABB report.

For apartments, the bank records price growth of 8.7% on the primary market and 10% on the secondary market over 2024-2025. The number of apartment registrations for the first four months of 2026: minus 0.2%. Precisely zero movement.

For land, the divergence is sharper still. The report calls the land market the main driver: the average plot price in Baku peaked at 41,700 manat in the first quarter of 2026, and land’s share of the value of a private house rose from 57% in 2021 to 71% by the second quarter of 2026. Meanwhile land registrations are down 13.5%. A record price on a volume contracting by an eighth.

For private houses - the only segment with rising activity: registrations up 22% alongside a rise in the average price from 313,000 to 353,000 manat. This figure, though, needs treating more cautiously than any other - why, is explained below.

Mortgage lending is not growing. A 1.8% fall in contracts against 10% price growth means the credit channel is not expanding demand, merely holding it up.

Technical passports deserve separate mention: they are issued before a property enters official circulation, which makes them a leading indicator of market intentions. Over the same four months: minus 4.7%.

A qualification is needed here, without which every figure above reads wrongly. Registration of a property right is not the same thing as a purchase and sale. This dataset includes inheritance, gifts, division of property in divorce and, most importantly, first-time registration of properties that previously had no documents at all. Strictly speaking, we are measuring not transactions but any change in the register.

This noise, however, is distributed very unevenly across segments - and that is the whole point.

Houses and land plots in Azerbaijan circulated for years without proper documents, and their gradual legalisation may account for a substantial share of registrations. This is precisely why the 22% rise in private houses cannot be called growth in sales without qualification: part of that figure is not new buyers but long-standing owners registering what they have owned all along. The same caution applies to reading the fall in land.

With apartments, in our view, the situation is fundamentally different. Housing in multi-unit buildings is documented when the building is commissioned; an apartment is not usually registered for the first time - it is bought and sold. Inheritance and gifts exist in this segment too, but their share is relatively stable year on year and does not change the dynamic.

In other words, the apartment figure - minus 0.2% - is the cleanest in the entire dataset, and it is the one directly comparable with ABB’s price charts, which are built primarily on apartments.

Construction is not slowing - it is falling

The standstill has a second dimension, absent from the ABB report: the dynamics of the construction sector itself. And it has to be read in the right order.

First, the record. Over the whole of 2025, 35,390 housing units with a total area of 4.63 million m² were commissioned in Azerbaijan. That is 75.6% more than in 2024, when 20,149 units covering 2.36 million m² were commissioned. Private and country houses accounted for 25,350 units - 2.2 times more than a year earlier - and apartments in buildings for 10,040 units, up 16.4%.

The figure is impressive but misleading. Commissioning reflects projects begun two or three years earlier. This is the market’s past, not its present condition.

The present looks fundamentally different. According to the State Statistical Committee, 207.7 million manat was invested in housing construction in Azerbaijan in January-February 2026 - 31.4% less than in the same period a year earlier. Investment in the construction sector as a whole fell by roughly a third over the first two months of 2026.

The figures that follow are no better. In January-May 2026 value added in the construction sector contracted by 23% in real terms. Construction’s share of gross domestic product shrank to 4.6%, against 6.5% for January-October 2025.

This is happening against a background of no movement in the wider economy. In January-May 2026 the country’s GDP was unchanged in real terms: 51,783.8 million manat at zero growth. The non-oil-and-gas sector added 0.4%, the oil and gas sector lost 1%. Investment in fixed capital fell by 2.1%, and investment in the non-oil sector by 14.7%. And one of the principal drags proved to be construction.

The official explanation for the decline came from the Minister of Economy. In an interview with Public Television (İTV) and AZERTAC published on 11 June 2026, Mikayil Jabbarov named construction as the only sector with a sharp decline over the first four months of the year, and stressed that this decline was linked more to state than to private construction. He gave the reason directly: growth came in below expectations mainly because the timing of a number of state projects had been shifted, and the corresponding works would be carried out at the next stage.

The explanation is credible, but we would add a couple of points. Rescheduling explains the state share of the decline, and the difference between money allocated and money disbursed really is fundamental here: the funds are in the budget, what has shifted is the schedule for spending them. But a 23% fall across the sector as a whole means private construction is not growing either.

In the same interview, incidentally, the minister cited another figure: investment in fixed capital in the non-state part of the non-oil-and-gas economy rose by more than 15%, and the non-state sector’s share of investment climbed to 51.1%.

That figure cannot be read as an indicator of private construction, however - it covers the entire non-oil-and-gas economy: industry, communications, agriculture, transport. If anything it measures the depth of the contraction on the state side: with total investment in the non-oil sector down 14.7%, a rise in the non-state share to half means exactly that.

State construction, meanwhile, is no abstraction. In the same interview the minister lists among state investments roads and housing for citizens returning to the liberated territories.

The 2026 budget, incidentally, allocated 3.5 billion manat for the restoration and reconstruction of the liberated territories - against 4 billion manat allocated a year earlier.

All of this, however, concerns the country as a whole.

Central Baku has constraints of its own on top of that. Property experts attribute the shortage of new supply here to two causes: tighter construction rules and the absence of free land plots, particularly in the central part of the city (even though a look at Baku from the air, let alone at the rest of the country, reveals expanses that are empty or occupied by old buildings).

The conclusion of this section matters for everything that follows. Supply on the market is constrained and in 2026 is physically contracting, for reasons both objective and subjective. That in itself explains why prices are holding high: scarcity supports price. But there are no transactions even amid scarcity. Even contracting supply, in other words, is failing to find a buyer - and the explanation for that lies in price, in household purchasing power and in the availability of mortgage lending, which, as the figures above show, is used to a very limited extent.

One qualification about the commissioning figures. 10,040 apartments for the whole country in a year looks modest against the real volume of construction. Part of the new housing stock is probably commissioned late or does not enter these statistics at all, so the indicator should be read as a measure of direction rather than as an exhaustive market volume.

Four markers inside the report itself

The most interesting part: ABB’s own data confirms this picture. The conclusions drawn from it are simply written more gently than the figures warrant.

First. The bank itself publishes the gap between listing and transaction. Table 36: at budgets up to 200,000 manat the negotiating discount runs at 3-5%, at 200,000-350,000 it reaches 7%, and at 350,000-500,000 it reaches 10%. In the expensive segment, in other words, one tenth of the stated price does not exist. Every chart in the report is built on asking prices. Which means the more expensive the segment, the more the chart overstates reality - and it overstates systematically, not at random.

Second. Slowing growth on top of enormous accumulated appreciation. This concerns annual growth in the average price per square metre. On the primary market the square metre rose by 22.7% in 2023, 13.9% in 2024 and 8.7% in 2025; on the secondary market by 19.0%, 14.9% and 10.0% respectively. Each year’s increment is roughly a third smaller than the last, but it remains positive: cumulatively over three years the primary market has appreciated by about half, the secondary market by almost as much. The report calls the slowdown a transition to a mature phase. Another reading is possible: purchasing power has hit its ceiling, and growth continues by inertia rather than by demand.

Third. Shrinking floor space. The average area of an apartment on offer fell from 118 m² in 2021 to 84-87 m², with an average of 2.7 rooms. The report reads this as supply forming around compact areas. The mechanics are simpler: the ticket price is held down by cutting the product. The price per metre rises, the size of the apartment falls, the final sum stays affordable. And the spacious apartments do not go anywhere - they sit in the listings, pulling up the average asking price without taking part in a single transaction.

Fourth. Inversion of the primary and secondary markets. In Nasimi district at the start of 2025, secondary housing (as a rule, apartments with finishing) cost 352 manat per metre more than primary housing (without finishing); by the second quarter of 2026 primary had overtaken it by 99 manat. In Narimanov the gap narrowed from 606 to 117 manat. In Yasamal, Sabunchu, Garadagh and Surakhani the picture is the same.

The report explains this as strengthening demand for new builds.

In our view, though, another explanation is possible, and it is simpler. The primary price is the developer’s own price list, which he sets himself, and often marks up. The secondary price is formed by thousands of independent sellers, each of them ready to negotiate because they need the money. The convergence of these two lines may indicate not the strength of new builds but that the secondary market has stopped rising while the primary market stands still by decision of whoever listed it.

It is worth adding that the comparison actually understates the gap. Secondary apartments are usually sold with finishing, primary ones without. If primary housing without finishing has overtaken secondary housing with finishing, the real divergence between the segments is greater than the 99 manat visible in the table.

The question that settles it: how much does property actually pay in cash?

Here the report provides figures that leave no room for interpretation.

Rental yield on housing in Baku is 5.4%. The payback period on an investment through rent is 18.5 years. In Numbeo’s regional comparison, Baku looks like this:

Baku is second from bottom in the region. An apartment here is more expensive relative to its rental stream than in Istanbul, Astana, Tashkent or Tbilisi.

The total return on housing cited by the report is about 13.8% at its peak. But it is made up of two unequal parts: roughly 8.4 percentage points is growth in the price index, and 5.4 points is rental yield. The first part exists on paper and is realised only on sale. The second arrives as cash - and it is a gross figure, before vacancy, repairs, agent’s commission and tax. By our calculations, about 3.5-4% remains net.

Now the comparison that really matters.

The average rate on term manat deposits for individuals in August 2026 stood at approximately 10%. The best offers on 18-24 month deposits reach 11.5-12%. Interest income of up to 2,400 manat a year is exempt from tax, and deposits of up to 100,000 manat are insured.

A deposit yields roughly three times more actual cash than an apartment. Without illiquidity, without maintenance costs, without the risk of failing to find a buyer or a tenant. Property wins only through paper revaluation - that is, on the expectation that in a year’s time somebody will be found who pays more. And it is precisely this expectation that the registration statistics fail to confirm.

Who holds the price up

The question remains: why do prices not fall when volume is frozen?

Part of the answer lies in the report itself. In its sources section ABB states that it used data from the Central Bank, the State Statistical Committee, the State Service for Property Issues, the bank’s own internal database - and market supply data provided by MBA LTD.

That last item means asking prices.

It is worth understanding the nature of the document here. The report is prepared by the risk management department and the collateral valuation unit - that is, by the bank’s risk function rather than its marketing function. For a bank this is entirely normal.

But that is exactly why its inputs are asking prices, which is what collateral valuation uses. And that is exactly why such a document should not be read as a study of the sales market - it never was one.

A closed loop takes shape. Unsold apartments in new buildings serve as security for construction loans. The valuation of that security relies on market benchmarks. Market benchmarks are formed from asking prices. Asking prices are set by the developer himself - and lowering them means marking down the valuation of his own collateral portfolio.

This is not an accusation, nor a suggestion of collusion. It is a structural feature that arises automatically wherever the pricing chain contains no independent source of transaction prices.

What is more, the market’s own standstill amounts to an objection to this model. A price nobody pays is not a price but the seller’s opinion of a price. Declining to transact is the only form of dissent available to a buyer.

But here a question arises that changes the whole conversation.

Since 1 February 2023, payments for property purchases, share contributions to housing construction cooperatives and purchases of residential and non-residential space from developers must be made by non-cash transfer only - the relevant provision was written into the Law on Non-Cash Settlements.

That means the banking system has for three years been seeing neither valuations nor listings but actual transaction prices.

In other words, the figure whose absence we have discussed throughout this article does exist. The register of contracts sits with the State Service for Property Issues. The payments under them sit with the banks. The consolidated picture is most likely held by the Central Bank. The problem is not that the data is hard to collect. It has already been collected. The problem is that it is not published in full.

The register of closing prices is not disclosed publicly, so valuers, analysts and readers alike are left to rely on the one thing they can see: the listing.

Hence the observation familiar to anyone who follows property listing sites: the same large units in central districts of Baku sit there for months, sometimes years, periodically reposted with a refreshed date, and their price does not change. An absence of reductions alongside an absence of sales is not a sign of strong demand. It is a sign that there is nobody to reduce for, and no reason to.

An important clarification: this applies above all to the primary market. Secondary apartments do fall in price, and quite often - a private seller has a deadline and a need. A developer in a prime district has neither, for the time being, and that is exactly why the primary market holds its inflated price tag so steadily.

What this actually is

At first glance the market may look like a kind of bubble.

In our view, though, "bubble" is far from the most accurate description of the situation. A bubble implies leverage and forced selling: borrowers cannot meet their payments, the bank demands additional security, assets are dumped on the market and the price collapses within a year or eighteen months.

In Azerbaijan that mechanism does not exist yet. Mortgage penetration is low - 19,600 contracts over four months for the entire country, and the figure is not growing. Which means there are no margin calls and no cascade of forced sales.

What is happening is more accurately described in different terms: this is not overheating but stagnation.

A narrow circle of developers, constrained supply (particularly in central Baku), enough financial staying power to hold a property for a year or more, and a relative absence of external pressure on price. A market like this does not collapse straight away. It can stand for years, while nominal prices stay put, the real price disappears into the discount at closing, and the statistics show stability.

The obvious question here: how long will that staying power last?

The answer, in our view, lies not only in buyer behaviour but in the credit policy of the banks and in the state of the economy as a whole. What allows a developer to hold his price is not so much demand as the ability to service a construction loan and avoid a downward revaluation of collateral, together with the state contracts that many builders are working on - contracts which, as noted above, are shrinking.

The turning point will come first of all where banks begin to revise the valuation of security on construction projects (and the builders’ own staying power will matter particularly here - will developers have enough money to keep holding unsold units?).

That is what needs watching, not merely the movement of listings.

What is missing

Everything set out above is built on data that can be verified. But the key figures are still absent from the public domain - even though, as we have established, they exist and have been collected.

And for as long as that remains the case, any report on Azerbaijan’s property market - the ABB report included - will inevitably describe not the market but the shop window. Valuers value by listings. Banks take collateral by valuations. Developers post the listings. The circle closes, and nowhere inside it is there a single point at which a buyer confirms the price.

Publishing quarterly transaction statistics - broken down by district, property type and price band - would require neither new surveys nor new reporting forms. The data is already being collected.

The register is maintained by the State Service for Property Issues. The statistics ought to be published by the State Statistical Committee. Non-cash payments on transactions are visible to the Central Bank. All that is needed is the corresponding decision.

And it would cost roughly as much as it costs at present not to know what is happening to an asset against which a substantial part of the banking system’s loan portfolio has been issued.

Author: First News Intelligence Unit (FNIU) - the analytical division of 1news.az, specialising in research on Azerbaijan’s financial sector and economy.

Methodological note. Sources used: "Bakı şəhəri üzrə daşınmaz əmlak bazarının təhlili", issue V, ABB, August 2026; State Service for Property Issues statistics for January-April 2026; State Statistical Committee data on investment in housing construction for January-February 2026, on housing commissioned in 2025, and on GDP and value added by sector for January-May 2026; the interview given by Minister of Economy Mikayil Jabbarov to İTV and AZERTAC on 11 June 2026; the Law of the Republic of Azerbaijan on the State Budget for 2026; the Law on Non-Cash Settlements as in force since 1 February 2023; Central Bank data on deposit rates for August 2026; the cross-city comparison - Numbeo, June 2026 (as reproduced in the ABB report).

Registration data covers all grounds for transfer of title, not purchase and sale alone. The decomposition of total return is our own calculation based on Tables 33 and 34 of the ABB report. The average land plot value is given in the ABB report without a unit of measurement; comparison with land’s share of the value of a private house indicates that the figure refers to the price of one sotka (100 m²).

Read in other languages:

Застой рынка недвижимости в Азербайджане: застройщики повышают цены на квартиры при отсутствии роста продаж - АНАЛИТИКА

Azərbaycanda daşınmaz əmlak bazarında durğunluq: satışlar artmadığı halda tikinti şirkətləri qiymətləri qaldırır - TƏHLİL