Contents
  1. Data source and coverage
  2. Baseline window — why Jan 6, not Jan 1
  3. Rental volume calculation
  4. Ready sales volume — weekly averages and the DLD batch pattern
  5. Off-plan exclusion
  6. Rental and sales prices — the selection effect
  7. Sales pipeline timing
  8. Community segmentation and daily averages
  9. Ramadan overlap

Data source and coverage

All rental and sales transaction data is sourced from ReidIN, a licensed DLD (Dubai Land Department) data portal. ReidIN provides access to the full official DLD register of rental contracts and property sales transactions.

Data typeSourceCoverage
Rental contractsReidIN — DLD rental registerJan 1 – Mar 14, 2026 (~146,000 rows)
Ready salesReidIN — DLD sales registerJan 1 – Mar 15, 2026 (~12,000 rows)
ReidIN export behaviour: exports are incremental, not cumulative. Each download covers only the selected date range. New data files are merged with previous files — not used to replace them — to maintain full history.

Only residential property transactions are included. Off-plan sales are excluded from conflict-period analysis (see Section 5).

Rental contracts are filtered to Rent Type = 'New' or Rent Type = 'Renewed' only. One row in each export contains filter metadata as the Rent Type value — this is excluded by filtering to exact values rather than dropping nulls.

Baseline window — why Jan 6, not Jan 1

The pre-conflict baseline is January 6 – February 27, 2026 (53 days).

The first week of January is anomalously low due to the New Year holiday lag — residents who signed agreements in late December often register them in the first days of January, creating an artificial trough. Including this period in the baseline would artificially depress the pre-conflict average and overstate the conflict impact.

January 6 was selected as the first full trading week. The baseline ends February 27 (the day before the conflict started on February 28).

Summary: Baseline = Jan 6 – Feb 27 = 53 days. Post-conflict = Feb 28 onwards.

Rental volume calculation

Rental contract volumes are expressed as daily averages within each window:

Pre-conflict daily avg = total contracts in baseline ÷ 53
Post-conflict daily avg = total contracts post-Feb 28 ÷ number of days elapsed
% change = (post avg − pre avg) / pre avg × 100

Using daily averages (rather than raw totals) is essential because the baseline period (53 days) and the growing post-conflict window have different lengths. Comparing raw totals would produce artificially large declines as the ratio of baseline-to-post days widens.

Historical bug (fixed Mar 16, 2026): An earlier version of the dashboard compared raw totals rather than daily averages. This produced community-level declines of ~80% when the correct figure was ~30–40%. All figures now use daily averages.

Ready sales volume — weekly averages and the DLD batch pattern

DLD registers ready sales transactions in weekday batches. The daily pattern is stark: weekend days (Friday and Saturday in UAE context, plus Sunday) show near-zero registration volumes, while weekday volumes spike accordingly. The daily breakdown from the post-conflict period illustrates this clearly:

Day of weekTypical volume
Saturday / Sunday1–22 transactions
Monday–Friday83–183 transactions

Because the post-conflict period began on Saturday February 28, the first two days contributed only 26 transactions combined — heavily distorting a simple daily average calculation.

The correct approach is weekly averages on complete weeks only:

Pre-conflict: 6 complete weeks (Jan 12 – Feb 22) = avg 1,187 transactions/week
Post-conflict: 2 complete weeks (Mar 2 – Mar 15) = avg 655 transactions/week
Weekly average change: −45.1%

The simpler daily average approach produces −52.1%, which overstates the decline due to the weekend-heavy composition of the first post-conflict days. The weekly average is the more defensible figure.

Caveat: Only 2 post-conflict weeks are available in this dataset. Treat the −45.1% figure directionally, not as a statistically robust estimate. The April figures will provide the first clean multi-week read.

Off-plan exclusion

Off-plan transactions are excluded from all conflict-period analysis on this dashboard.

Off-plan transactions have a DLD registration lag of 30–90 days from the point of agreement. A transaction agreed in October 2025 may appear in the DLD register in January 2026. A transaction agreed post-conflict in March 2026 will not appear in DLD data until April–June 2026 at the earliest.

This means: any off-plan data visible in the conflict window was agreed weeks or months before the conflict started. Including it would create a false baseline and make the conflict impact appear smaller or larger than it actually is.

Policy: Off-plan is excluded from all pre/post conflict comparisons. Ready transactions only. A note is displayed on the dashboard wherever off-plan exclusion affects interpretation.

Rental and sales prices — the selection effect

Both rental prices and sales prices per sqft show stable or slightly rising medians in the post-conflict period. This is counterintuitive given the volume collapse. The explanation is the selection effect.

What is the selection effect?

When demand contracts, not all demand contracts equally. The first movers to pause are marginal, budget-sensitive participants — renters who were still deciding between communities and were not committed, buyers who were circling without urgency. These are typically the lowest-price transactions in the distribution.

The transactions that continue to complete are from more committed, above-market participants: expats whose contracts require them to move, buyers with pre-approved mortgages closing on agreed deals, investors executing pre-planned allocations.

When the bottom of the distribution drops out, the median rises mechanically — not because prices went up, but because the cheapest transactions stopped happening.

Observed: Post-conflict rental median AED 73,000 vs pre-conflict AED 70,000 (+4.3%)
Interpretation: NOT landlords raising prices. Budget-sensitive movers paused. Above-market signings continued.

Observed: Post-conflict sales median AED/sqft 1,500 vs pre-conflict 1,489 (+0.8%)
Interpretation: Same pattern. Cheapest transactions dropped out. Committed buyers continued.

When will prices actually move?

The lag between a volume shock and a price correction is typically 3–6 months in the Dubai rental market and 4–8 weeks for seller repricing in the sales market (the time it takes for landlords and sellers to acknowledge that the market has changed).

At 15 days post-conflict, no price correction is expected. If volumes remain compressed through April–May, median rental prices will begin drifting downward as the pool of committed signings is exhausted and marginal landlords begin competing on price.

Watch for: Rental new contract medians in the May–June data. If still elevated while volumes remain suppressed, the selection effect is persisting. If medians begin declining while volumes remain flat, price correction has started.

Sales pipeline timing

The DLD register reflects completed registrations, not agreements. A buyer who decides to purchase today will not appear in the DLD data for several weeks, depending on transaction type:

Transaction typeTypical DLD registration lagFirst post-conflict decisions visible
Cash sale~30–45 daysLate March 2026
Single-side mortgage (buyer or seller financed)~45–60 daysMid-April 2026
Double-side mortgage (both parties financed)~60–75 daysLate April 2026
Off-plan developer registration30–90 days (batch)Excluded from analysis

Implication: the sales volume figures in the early post-conflict period (February 28 – March 15) largely reflect transactions agreed before the conflict. The first genuinely post-conflict buyer decisions will appear in the data from late March onwards. The April dataset will provide the first clean read.

Community segmentation and daily averages

The community tracker uses daily averages per community, not raw totals. The same logic applies as for the overall volume calculation: the baseline is 53 days and the post-conflict window is growing, so raw totals are not comparable.

Community pre-conflict avg/day = community total in baseline ÷ 53
Community post-conflict avg/day = community total post-Feb 28 ÷ post-conflict days
% change = (post avg − pre avg) / pre avg × 100

Communities with fewer than 5 contracts per day in the pre-conflict period produce highly volatile % change figures. These are flagged in the dashboard — the directional trend is meaningful but the exact % is not.

Community segments (cross-project decision X14)

SegmentCommunities
Prime ApartmentsDowntown Dubai, Dubai Marina, Palm Jumeirah, Business Bay, Dubai Creek Harbour, JBR
Mid-Tier ApartmentsJVC, JVT, Al Furjan, Dubai Sports City, Dubai Silicon Oasis, Arjan
ValueInternational City, Discovery Gardens, Town Square, Damac Hills 2, Dubai South, Liwan
VillasArabian Ranches, Jumeirah Golf Estates, Tilal Al Ghaf, Damac Hills, The Springs, Dubai Hills Estate, Emaar South

Majan is excluded from the community tracker. It shows 234 transactions over 6 days in the post-conflict window — a known anomalous bulk registration event. Including it would distort the value segment average.

The full 251-community table in the dashboard shows the complete DLD register and is not filtered by segment.

Ramadan overlap

Ramadan 2026 began February 18 — ten days before the conflict started on February 28. This creates an analytical challenge: the post-conflict period and the Ramadan period overlap almost entirely.

Ramadan typically suppresses real estate transaction volumes in Dubai, as decision-making slows and move activity reduces during fasting hours. The pre-conflict baseline (Jan 6 – Feb 27) captures the first 10 days of Ramadan (Feb 18–27). The post-conflict period (Feb 28 onwards) is entirely within Ramadan.

This means the observed volume decline is likely a combination of two effects: conflict-driven caution and Ramadan seasonality. It is not possible to cleanly separate the two with this dataset.

How the dashboard handles this: The Ramadan period is shaded on the rental volume chart and a note is displayed in the context strip. Interpretations are careful not to attribute the entire volume decline to the conflict. The April data (post-Ramadan) will provide the first conflict-only read, as Ramadan ends around March 19–20, 2026.

Historically, the first two weeks of post-Ramadan activity see a rebound in transaction volumes as delayed decisions complete. If volumes do not rebound by mid-April, the conflict effect is clearly dominant.