Where Baku Actually Borrows: Mortgage Demand by District

As at 30 June 2026, the Mortgage and Credit Guarantee Fund of Azerbaijan had 59,082 mortgage loans on its books nationally, of which 45,607 - 77.2% - are in Baku. Inside Baku the distribution is not where the marketing points: Xətai district holds 10,063 loans and Səbail, the seafront, holds 1,490. That is a 6.8-fold difference between the district that borrows most and the most expensive address in the city.

We pulled and cross-checked these figures ourselves from the Fund's published infographics. The methodology and the limits are at the bottom of the page.

Mortgage loans by Baku administrative district

District	Loans	Share of Baku
Xətai	10,063	22.1%
Yasamal	8,880	19.5%
Binəqədi	5,851	12.8%
Suraxanı	5,256	11.5%
Nizami	3,456	7.6%
Nəsimi	3,428	7.5%
Nərimanov	3,264	7.2%
Sabunçu	2,034	4.5%
Səbail	1,490	3.3%
Qaradağ	947	2.1%
Xəzər	889	1.9%
Pirallahı	49	0.1%
Baku total	45,607	100%

What the Xətai-Səbail gap actually tells you

Two districts, 22.1% against 3.3%. The obvious reading - that Xətai is more popular - is the wrong one. What this measures is credit-financed demand, and credit-financed demand is a mid-market phenomenon everywhere. Səbail is not short of buyers; it is short of buyers who need a mortgage. The seafront is substantially a cash market, and the mid-market districts are where the banking system is actually underwriting households.

For an investor that reframes two things. First, resale liquidity in Xətai and Yasamal is underpinned by a financing channel that barely touches Səbail - 41.6% of all Baku mortgages sit in those two districts alone. Second, a price-per-square-metre table flattens exactly this distinction. Two districts can show similar prices and have completely different buyer mechanics behind them.

How much deposit buyers really put down

This is the question every entry-level guide answers with a guess. The Fund publishes the actual loan-to-value distribution across all 59,082 loans:

LTV band	Loans	Share	Implied deposit
0-50%	5,398	9.1%	50%+
51-60%	6,190	10.5%	40-49%
61-70%	17,670	29.9%	30-39%
71-80%	17,287	29.3%	20-29%
81-90%	12,537	21.2%	10-19%

78.8% of borrowers put down 20% or more. Only 21.2% of loans exceed 80% LTV. The modal borrower sits in the 61-80% band - a deposit of 20 to 39 per cent.

What the market actually finances: 41 to 80 m²

Floor area	Loans	Share
under 41 m²	3,438	5.8%
41-60 m²	18,771	31.8%
61-80 m²	17,632	29.8%
81-100 m²	9,290	15.7%
101-120 m²	4,361	7.4%
over 120 m²	5,590	9.5%

61.6% of all mortgaged property is between 41 and 80 m². Anything over 100 m² is 16.8% of the book. If you are buying for resale rather than for yourself, that distribution is the shape of your future buyer pool - and a 130 m² apartment is competing for the attention of one sixth of the financed market.

Affordability stress: 42.9% pay over 30% of income

Annuity as share of monthly income	Loans	Share
0-10%	3,303	6.2%
11-20%	15,120	28.3%
21-30%	12,127	22.7%
31-40%	12,127	22.7%
41-50%	6,683	12.5%
51-60%	2,966	5.5%
61-70%	1,103	2.1%
over 70%	52	0.1%

42.9% of borrowers pay more than 30% of monthly income on the annuity, and 20.2% pay more than 40%. This is the constraint on how much further local demand can push prices without wage growth, and it is worth reading next to the price trend: per the State Statistical Committee, the primary market rose 12.38% year on year in Q2 2026 against 8.58% for the secondary market.

The part that applies to a foreign buyer - and the part that does not

You cannot use these programmes. The Fund's subsidised and standard mortgage lines are for local borrowers; a foreign buyer is looking at private bank lending on different terms, or at cash. We cover what that looks like in the guide to mortgage rates and the lending market, and the ownership rules in the guide for foreign buyers.

So why does this data matter to you? Because it is the shape of the demand you are buying into. The district distribution tells you where households with financing are concentrated, the area distribution tells you which product has the deepest resale pool, and the PTI distribution tells you how much headroom local demand has left. None of that is visible in a price-per-square-metre average.

The other half: how many deals actually registered

MCGF shows the stock of credit. The State Service on Property Issues publishes the flow of transactions — and in 2025 the two moved in opposite directions.

National state registrations of property rights, 2025 against 2024:

	2024	2025	change
All registrations	378,495	364,136	−3.8%
of which first registration	64,454	70,596	+9.5%
of which repeat registration	314,041	293,540	−6.5%
Apartments, first	24,994	27,690	+10.8%
Apartments, repeat	60,944	65,198	+7.0%
Houses and dachas, first	21,429	26,348	+23.0%
Houses and dachas, repeat	41,577	45,340	+9.1%
Land, first	13,658	12,056	−11.7%
Land, repeat	202,925	174,539	−14.0%
Mortgage registrations	73,809	63,836	−13.5%

The headline figure is misleading and the detail matters. Registrations fell 3.8% overall, which reads like a cooling market. It is not a housing story. Repeat land registrations alone fell by 28,386 — nearly twice the entire net decline of 14,359 — and land is 48% of all registrations, so it dominates the total. Strip it out and housing grew: apartments up 8.1% across both registration types, houses and dachas up more.

The figure worth sitting with is the mortgage line. Registered mortgages fell 13.5% in a year when apartment transactions rose 8.1%. More housing deals, fewer of them financed. That is the same pattern the district table above shows at city scale — credit concentrated in the mid-market while the expensive districts transact in cash — appearing again at national scale.

One limit, stated plainly: mortgage registrations cover every property type, not apartments alone, so "more apartment deals done in cash" is a reasonable reading but not something this data proves on its own. What it does establish is that transaction volume and credit volume diverged.

Source: State Service on Property Issues of the Republic of Azerbaijan, emlak.gov.az/az/page/view/22, registration statistics, annual series, retrieved 11 October 2026. Granularity is annual and national, broken down by property type and by first versus repeat registration. There is no monthly series and no breakdown by Baku district — we checked, because that would have been the more useful dataset. Note also that the price figures elsewhere on this page are Q2 2026 against Q2 2025, while these volumes are full-year 2025 against 2024; the periods are adjacent, not identical, and should not be read as one series.

Methodology and limits

• Source: Mortgage and Credit Guarantee Fund of Azerbaijan, published infographics, mcgf.gov.az/az/infoqrafikler. All series dated 30.06.2026. Retrieved and parsed 11 October 2026.
• Cross-check: the district figures sum to 45,607, which matches the Fund's own regional dataset independently. The national total of 59,082 appears consistently across the LTV, floor-area and borrower-income series.
• Known limit: the PTI series totals 53,481 rather than 59,082, so roughly 5,600 loans have no PTI figure recorded. The PTI percentages above are shares of the 53,481 that do. We have not adjusted for this and do not know the reason.
• Known limit: the Fund's portfolio value series is labelled "mln manat" while its band labels are in manat. Because that unit is ambiguous in the source, we have left it out rather than guess.
• Scope: these are loans through the Fund. Private bank mortgages outside the Fund's channel are not included, so this is a large sample of the subsidised and guaranteed market, not the whole credit market.
• Price trend context: State Statistical Committee of Azerbaijan, residential property price indices, series 007_1, data through Q2 2026.

This page is informational and is not investment, tax or legal advice. The figures are accurate as at the measurement dates stated above.