Comps (Comparables)
Comps are recently sold properties similar enough to your subject property to be used as evidence of market value. They are the empirical foundation of every ARV estimate, cap rate validation, and rental income projection. Without reliable comps, every return metric is speculation.
There are two types: sales comps (used to estimate purchase price and ARV) and rental comps (used to estimate market rent).
Sales comps vs. rental comps
| Type | What it answers | Used for |
|---|---|---|
| Sales comps | What would this property sell for? | ARV estimation, offer price, appraisal prep |
| Rental comps | What can this property rent for? | Income projection, pro forma, market rent |
Both follow the same principle: find properties sufficiently similar to your subject, sold or leased recently, and use their price or rent as a reference.
What makes a reliable sales comp
Not every "comparable" is actually comparable. Apply these filters:
| Criterion | Guideline |
|---|---|
| Sale type | Closed only — not active listings, pending, or expired |
| Date | Within 90–180 days (90 days preferred; lenders use 6 months maximum) |
| Location | Within 0.5 miles in urban/suburban areas; up to 1 mile in rural |
| Size | Within 15–20% of subject's square footage |
| Bed/bath count | Same or within one bed/bath |
| Condition | Similar finish level — renovated comps for a post-renovation ARV |
| Property type | Single-family to single-family; duplex to duplex |
Worked example: ARV estimation
A 3-bed/2-bath, 1,450 sq ft property in need of renovation. You are estimating post-renovation value.
Step 1 — Find 3–5 closed comparable renovated sales:
| Address | Sq ft | Bed/Bath | Sale price | $/sq ft | Date |
|---|---|---|---|---|---|
| 412 Maple St | 1,380 | 3/2 | $278,000 | $201 | 45 days ago |
| 88 Oak Ave | 1,510 | 3/2 | $302,000 | $200 | 62 days ago |
| 201 Birch Ln | 1,490 | 3/2 | $289,000 | $194 | 28 days ago |
Step 2 — Analyze the range:
| Metric | Value |
|---|---|
| Price range | $278K – $302K |
| Average $/sq ft | $198 |
| Median sale price | $289,000 |
Step 3 — Apply adjustments for subject differences:
- Subject has slightly dated layout: adjust $/sq ft down by 2.5% → $193/sq ft
- Subject ARV estimate: 1,450 × $193 = $279,850 → use $280,000
Using $280K (conservative, near the low comp) rather than $302K (the high) protects your downside. The low comp is your floor; the high comp is what the market may support at best execution.
Adjusting comps for differences
No comp is perfect. Apply dollar adjustments for material differences:
| Difference | Estimated adjustment |
|---|---|
| Each bedroom ± | ± $8,000–$15,000 |
| Each bathroom ± | ± $5,000–$10,000 |
| Square footage (per 100 sq ft) | ± $15,000–$25,000 |
| Garage vs. no garage | ± $10,000–$20,000 |
| Recent renovation vs. dated | ± $15,000–$40,000 |
| Backyard/lot size | Market-dependent |
Adjustments are estimates — the goal is directional accuracy, not precision. If all your comps require large adjustments to fit the subject, you need better comps.
Where to find comps
| Source | Sales comps | Rental comps | Access |
|---|---|---|---|
| MLS (via agent or platform) | ✅ Best | ✅ Lease data | Agent or paid access |
| Zillow / Redfin (Sold) | ✅ Good | ❌ Limited | Free |
| Realtor.com (Sold) | ✅ Good | ❌ | Free |
| County recorder / assessor | ✅ Legal record | ❌ | Free (often) |
| PropStream / BatchService | ✅ | ❌ | Paid |
| Rentometer / Rent.com | ❌ | ✅ Rental listings | Free / paid |
| Cozy / Buildium comps | ❌ | ✅ | Paid platforms |
For sales comps, always use closed sales — not listings. Listings reflect asking price; closings reflect what buyers actually paid. In a softening market, listings can be 5–10% above where transactions clear.
Rental comps: what to look for
For rental comps, the criteria are similar but slightly different:
| Criterion | Guideline |
|---|---|
| Lease type | Active leases or recent signings (30–90 days) |
| Property type | Single-family to single-family; apartment to apartment |
| Bed/bath | Same count |
| Size | Within 15–20% of subject |
| Condition | Similar amenity level |
| Location | Same street, neighborhood, or school district |
Look at actual asking rents on listing sites AND — if possible — what similar units recently leased for. Asking rent and lease rent can differ, especially in soft markets.
Common mistakes
1. Using active listings as comps. A property listed for $320K may sell for $295K. Never use asking price as ARV evidence.
2. Using old comps in a moving market. A 12-month-old comp in a softening market is dangerously optimistic. In a rising market, old comps understate value. Use the most recent data available.
3. Cherry-picking the high comps. Sellers and optimistic investors naturally gravitate toward the high end. Use all comparable sales that meet your criteria, then evaluate the range. Base your ARV on the conservative end unless you have specific reasons the subject will outperform.
4. Not adjusting for condition. A comp that sold renovated for $302K versus your subject property that will also be renovated is fine. A comp that sold at $302K in "as-is" condition is not comparable to your post-renovation ARV.
Frequently asked questions
How many comps do I need? Three is the minimum for a credible ARV estimate; five is better. With fewer than three, you have too little data to triangulate value. With more than five, you may be including properties that are too different to be genuinely comparable.
Can I use comps from outside my neighborhood? Only if the subject neighborhood has insufficient data. Crossing a major road, school district line, or neighborhood boundary can change values by 10–20%. Use the most local data possible, then expand the radius only when necessary — and note the limitation.
Do I need MLS access to find good comps? Not necessarily. Zillow and Redfin's "Sold" filters show most MLS data with a short delay. County recorder offices have public records. For serious investing, MLS access (via a licensed agent or paid platform) gives you more data and the ability to filter by days-on-market, price reductions, and listing history.
Comparable sales data provides evidence of market value, not certainty. Valuations reflect market conditions at a point in time. This is not investment advice.