RealTorch / Field Guide / New-construction CMA
The Real Torch Field Guide to Homebuilding
What makes a good comparable market analysis for new homebuilding?
A good CMA compares alternatives a buyer would actually consider.
A good new-construction comparable market analysis starts with the decision, identifies homes the same buyer would realistically consider, and scores them for recency, location, lot, size, plan, finish, incentives, and status. It separates closed evidence from active competition and pending signal, explains every adjustment, and returns a supported range with confidence and exceptions.
Use this guide to define the subject, choose and weight comparable homes, handle new-construction incentives and options, and return a range that a builder can inspect rather than a number it has to trust blindly.
Start with the decision the CMA needs to support.
A comparable market analysis, or CMA, is not one fixed report. The right comp set changes with the question.
- What should we list this completed spec home for?
- What finished-home value should support a land residual?
- Which product belongs in this community?
- How should a community be positioned against nearby builders?
- What premium does this lot or feature appear to support?
State the subject, decision, geography, and as-of date before selecting records. A CMA for today's list price should emphasize current competition and recent buyer behavior. A CMA for a home that will deliver in eighteen months needs a wider discussion of pipeline, time, and uncertainty.
A comparable sale is not automatically a good comp.
A comp is a substitute, not merely a neighbor. Proximity creates a candidate pool. A record deserves weight only if the same buyer might plausibly have considered it.
How a sale or listing earns its place in a new-home CMA
- Candidate recordsNearby sales, pending homes, active listings
- Buyer-substitution gateWould the same buyer realistically consider it?
- If noExclude it and record the reason
- Score similarityRecency, home, plan, lot, location, finish, terms
- Assign its evidence roleClosed, pending, active, withdrawn
- Explain adjustmentsMarket-supported differences, not rules of thumb
- Supported range + confidencePosition, exceptions, and what could change it
Recency
Recent evidence usually deserves more weight, but a thin market may require older sales. If market conditions changed between contract and the effective date, test and explain a time adjustment. “Within six months” is a filter, not a conclusion.
Home and product
Compare new versus resale, age, finished square footage, bedrooms, baths, stories, parking, plan type, finish level, custom versus spec, and delivery timing. Square footage alone cannot tell you whether the same buyer saw the homes as substitutes.
Lot and site
Compare lot size, usable area, topography, orientation, view, privacy, corner or cul-de-sac position, adjacent uses, and location inside the community. A larger lot beside a lift station can deserve less weight than a smaller, quieter site.
Location through work, school, and play
Ask whether each home reaches the same jobs, solves the same school decision, and provides the same amenities and daily life. Straight-line distance or ZIP code can miss how buyers actually substitute between locations.
Offer terms, options, and incentives
Normalize base price, lot premium, structural options, design-center upgrades, closing-cost help, rate buydowns, seller concessions, spec discounts, and included features. In new construction, headline price and net effective price can be different evidence.
Use closed, pending, and active homes differently.
These records answer related questions. They should not be blended into one average.
| Status | What it tells you | Important limitation |
|---|---|---|
| Closed | What a buyer actually paid | It reflects an earlier contract and market |
| Pending | What recently found a buyer | Contract price and concessions may be unknown |
| Active | What buyers can choose today | Asking price is not accepted value |
| Withdrawn or expired | What the market did not accept | The cause may be price, condition, timing, access, or seller behavior |
Closed sales anchor demonstrated value. Pending homes provide fresh direction. Active homes define the competitive shelf. Withdrawn and expired homes can reveal rejection, but only after the reason is investigated.
New construction requires adjustments a resale CMA often misses.
Community stage
An early phase sells against uncertainty and unfinished surroundings. A late phase may have proven demand but fewer preferred lots.
Builder and execution
Brand, model presentation, construction quality, delivery reliability, and customer experience can affect buyer substitution.
Build time
A completed spec and a to-be-built home do not solve the same timing problem.
Included features
Base price becomes misleading when one builder includes what another sells as an option.
Lot premiums
Separate the home from the site. The same plan can close at different totals because the lots carry different buyer value.
Incentives
A rate buydown, closing-cost credit, or upgrade allowance can change effective economics without changing the recorded headline price.
Custom, spec, and to-be-built
Control, immediacy, design burden, and certainty are part of the product.
Planned inventory
A CMA is an as-of-date view. A land or product decision must also consider homes that will compete when the subject reaches market.
Score first. Adjust second. Reconcile last.
Similarity scoring decides which records deserve attention. Adjustments then make the remaining differences visible. Reconciliation explains why some adjusted indications deserve more weight than others.
Define the candidate universe.
Set a broad enough geography and time window to avoid selecting only records that support the initial hypothesis.
Apply the buyer-substitution gate.
Exclude candidates the same buyer would not realistically consider, and record why.
Score similarity.
Use visible dimensions such as recency, location, home, plan, lot, finish, options, terms, and incentives. The decision determines the weights.
Make supported adjustments.
Use market evidence rather than a generic price-per-square-foot rule. Explain the data and technique behind material adjustments.
Reconcile the evidence.
Return a supported range, central position, confidence, influential comps, exceptions, and the evidence that would change the answer.
CMA and absorption answer different questions.
A CMA tests price support and market position. Absorption tests pace. A product can have strong price evidence and still sell too slowly for the builder's capital plan. Current inventory, recent sales pace, community stage, and planned pipeline belong in the risk discussion without being treated as closed-sale value.
Worked example: choosing comps for a proposed new home.
The subject is a proposed 2,650-square-foot, four-bedroom, three-bath spec on a 0.22-acre interior lot. It would deliver in nine months. The builder needs a finished-value range before setting the maximum land basis.
| Candidate | Status | Similarity | Decision | Why |
|---|---|---|---|---|
| 18 Alder Bend | Closed | 91 / 100 | High weight | Same buyer, plan scale, school decision, and finish; contract four months ago |
| 42 Reed Lane | Pending | 88 / 100 | Directional | Nearly identical plan and lot; contract price and incentive unknown |
| 7 Briar Court | Closed | 76 / 100 | Medium weight | Strong location match; older resale with renovated finish |
| 110 Station Row | Active | 73 / 100 | Competition | New home at a lower ask, but smaller lot beside a collector road |
| 5 Lake Crest | Closed | 49 / 100 | Exclude | Same radius, but lake access and a different school decision attract another buyer |
$548,000 to $572,000 before final lot and incentive decisions.
$559,000 base, with the lot and included-feature package stated separately.
Moderate. One high-similarity pending comp could materially change the center of the range when it closes.
Net incentive on 42 Reed Lane, final site premium, and market conditions at the subject's nine-month delivery.
The point is not the fictional $559,000 position. It is that every record has a role and a reason. The closest sale was excluded because it solved a different buyer problem. The active listing informed competition but did not prove value. The range stayed open until the strongest pending signal becomes observable.
What a useful new-construction CMA should deliver.
- Subject and decision.What home, plan, product, lot, or community is being tested, and why?
- As-of date and market boundary.When is the evidence current, and where does the buyer actually shop?
- Candidate universe.Show the broad set before selection to reduce confirmation bias.
- Similarity method.Explain the dimensions and why they matter to this decision.
- Selected and excluded records.Show both, with reasons.
- Evidence roles.Keep closed, pending, active, and rejected offerings distinct.
- Incentive and option normalization.Separate headline price from the effective offer when the evidence supports it.
- Adjustments and reconciliation.Show how differences were handled and why certain comps received more weight.
- Range and confidence.Give a position, exceptions, uncertainty, and what would change the result.
- Pace and pipeline context.State the competitive risk without pretending it is closed-sale evidence.
Where new-construction CMAs go wrong.
They choose the nearest homes instead of the same buyer's alternatives.
Distance is easy to measure. Substitution is what matters.
They mix new and resale homes without explaining the difference.
Age, warranty, delivery, finish, and buyer control can change the offer.
They treat asking price as proof of value.
Active homes show the shelf, not what buyers accepted.
They ignore incentives, options, and lot premiums.
Headline price can hide the economics that moved the buyer.
They average weak comps.
Five weak records do not become strong evidence because they fit in a spreadsheet.
They hide exclusions.
The rejected candidates often reveal the analyst's judgment more clearly than the selected set.
They use stale sales without testing time.
Older evidence may be necessary, but its market context must be examined.
They report one precise number.
A supported range with confidence is more useful than false certainty.
Common questions about comparable market analysis.
How many comps should a CMA use?
Use enough strong records to represent the market evidence, not a fixed number. A small set of highly substitutable homes can be more informative than a long list of weak matches. Show the candidate pool, selected set, and exclusions.
How recent should comparable sales be?
Recent sales usually deserve more weight. In a thin market, older sales can still help if the analyst tests market change between contract and the effective date and explains any time adjustment.
Should active and pending homes count as comps?
Use them as evidence, but give them different roles. Pending homes show fresh buyer direction while contract terms may be unknown. Active homes show today's competition but do not prove accepted value.
Can new homes be compared with resales?
Yes, when the same buyer would realistically consider both. Explain and support differences in age, warranty, condition, finish, timing, incentives, and buyer control rather than treating the categories as interchangeable.
How should builder incentives be handled?
Identify reasonably available financing concessions, closing-cost help, upgrade allowances, and included features. Estimate their market effect when evidence supports it. Do not assume every dollar of seller cost equals a dollar of buyer value.
Is a CMA the same as an appraisal?
No. A CMA can support builder pricing, product, and land decisions. An appraisal is a formal valuation assignment performed under applicable professional and lender requirements. A strong CMA can borrow transparent comparison discipline without representing itself as an appraisal.
Sources, method, and limits.
This guide adapts sales-comparison discipline to new-home product and pricing decisions. Fannie Mae's current guidance is a useful public reference: analyze the most comparable closed sales, contracts, and offerings; understand transaction conditions and concessions; support adjustments with market evidence; test changes in market conditions; and explain why some indications receive more weight.
- Fannie Mae: Sales Comparison Approach
- Fannie Mae: Adjustments to Comparable Sales
- Federal Housing Finance Agency House Price Index
- U.S. Census Bureau Survey of Construction
Important limit: this is a builder decision guide, not an appraisal standard or a value opinion. Data availability, local disclosure rules, incentives, condition, and future competition can change the conclusion. Verify material records and terms before committing capital or publishing a price.
Put the method to work on your decision.
Atlas can combine the market evidence with your parcel, plans, files, costs, and operating context. You can inspect the inputs, the analysis, and the limits before you act.