What to take into your next appraisal
- Distinguish transactions from listings and ready property from off-plan property.
- Record filters, exclusions and the denominator used for unit prices.
- Use synthetic examples to explain a method without presenting invented market evidence.
Record what was actually extracted
DLD’s Real Estate Data service exposes filters including date, transaction type, registration type, freehold status, area, usage and property type. Its transaction fields include amount, transaction size, property size, room count, master project and project. Capture the extraction date, selected filters and raw field definitions.
The service also points to Dubai Pulse for previous-year data. Retain the exact extract and applicable reuse conditions. A displayed transaction count or a broker’s selected chart is not equivalent to a reproducible pricing dataset.
Define comparability before seeing the answer
Choose the subject’s relevant submarket, property type, stage and approximate unit sizes. Exclude records that cannot be interpreted, duplicate events and transactions whose economic basis differs from the intended comparison. Keep an exclusion log with a reason for each removal.
Do not erase inconvenient observations merely because they lower the average. An unusually low value may reveal a different transaction type, a data issue or a genuine comparable. Investigate it and preserve the original record.
Worked example: cleaning a synthetic sample
The rows below are invented to demonstrate screening, not taken from DLD. The subject is an off-plan apartment. Each rate is amount divided by the stated comparable area. The ready apartment and land transaction are excluded by the prespecified scope.
The five-row median is AED 15,000 per sq m, while the three retained off-plan apartments have a median of AED 15,500. Neither number establishes a real Dubai price. The example shows why publishing the selection rule matters as much as publishing the resulting statistic.
| Synthetic record | Amount | Area | Rate and decision |
|---|---|---|---|
| Off-plan apartment A | AED 1,200,000 | 80 sq m | AED 15,000; retain |
| Off-plan apartment B | AED 1,550,000 | 100 sq m | AED 15,500; retain |
| Off-plan apartment C | AED 1,920,000 | 120 sq m | AED 16,000; retain |
| Ready apartment D | AED 1,300,000 | 100 sq m | AED 13,000; exclude |
| Land E | AED 6,000,000 | 1,000 sq m | AED 6,000; exclude |
Keep adjustments separate from observations
A comparable’s recorded rate is evidence; the adjustment for view, floor, specification, payment plan or delivery timing is an analyst judgement unless independently supported. Show the observed rate, adjustment and adopted rate in separate columns with reasons.
If three retained records are the only usable evidence, publish that limitation. Test the appraisal over a reasonable range supported by the actual evidence rather than treating a small-sample median as a precise market forecast.
Link evidence to the unit schedule
Apply the adopted prices to the subject’s measured unit areas, not a different denominator borrowed from the dataset. Test absorption and collection terms separately. A higher contracted price under a longer payment plan may produce a lower present value of cash.
A reproducible evidence pack contains the raw extract, screening rules, retained sample, adjustment register, chosen price range and the model version using it. Refresh the evidence when the design or launch date changes.
Sources and further reading
- Real Estate Data Dubai Land Department
- Valuation of development property RICS
Published by Feasly. How we prepare our guides. Suggest a correction.
