TURKSTAT states that house-sales statistics are published monthly at national, provincial and district levels using TKGM/TAKBİS transaction records, with breakdowns such as first/second-hand and mortgaged/other sales. Use transaction volume as an activity signal, not as a price per square metre.
Comparable Property Selection Guide
Core point
A comparable should resemble the subject in micro-location, building type, size, age/condition, floor/view and transaction date.
How to verify it
Distinguish asking prices from completed sale prices and adjust for time and terms when observations come from different periods.
Decision impact
Exclude a comparable requiring large or indefensible adjustments; many weak comparables do not compensate for poor similarity.
Selecting comparables before making adjustments
A strong comparable competes in the same buyer market, not merely at a similar square metre count. Prefer the same micro-location, building type, age, floor, size, condition and material rights such as parking or view, while recording the observation or transaction date. Separate asking prices from transaction evidence; if verified sale data are unavailable, label the sample as asking-price evidence rather than presenting it as completed-sale prices.
Adjust only after selecting the sample. A date difference may be contextualised with the TCMB RPPI, but an index is not a reason to mechanically multiply one apartment’s price. Size, renovation or floor adjustments should be explainable from market evidence rather than chosen to force the result toward the target price. Keep rejected comparables and the rejection reason; transparent exclusion of a weak comparable is as important as the records ultimately used.
A good comparable matches the attributes that actually move price
Distance alone is insufficient. Compare tenure/title type, building age and condition, floor/lift, usable area, view, parking, occupancy, payment terms and transaction date. Exclude or separately flag related-party, distressed or otherwise non-arm’s-length transactions, and distinguish an unclosed asking price from a documented transaction. When true comparables are scarce, use a range and explain each adjustment rather than manufacturing false precision. A broad market index can be used as a reasonableness check for time movement, but it cannot make a materially different property into a valid comparable.
A comparable is useful only when its differences from the subject property are visible. Match location, property type, size, age or condition, legal status and transaction period, and distinguish asking prices from completed-sale evidence. If a weaker match is retained because the sample is thin, identify the mismatch and explain the adjustment instead of averaging it as if it were identical.
Market data is useful only when its source, date, geography and method are known.
Topic-specific tests Define the sample before calculating: same micro-market, property type, age, size, floor and condition where possible; separate asking from transaction data and disclose outlier handling and adjustments.