Comparable Sales Analysis for Note Investors: A Practical Guide

Why Should Note Investors Learn to Pull Their Own Comps?

When you buy a mortgage note, you rarely have the luxury of ordering a full appraisal on every property in a loan pool. Appraisals cost $300 to $500, take weeks, and require interior access you almost never have during the pre-bid phase. Automated valuation models give you a starting point, but they cannot see the boarded-up windows, the fire damage, or the fact that the neighborhood two blocks east trades at half the price of the neighborhood two blocks west.

That leaves comparable sales analysis -- pulling your own comps -- as the single most cost-effective way to develop a defensible estimate of fair market value (FMV). It costs nothing but your time, and it produces a better result than any algorithm because a human is evaluating the data, applying judgment, and accounting for nuances that no formula can capture.

The problem is that most note investors never learned how to run a proper comp analysis. They pull up Zillow, glance at the three properties the algorithm suggests, and move on. That approach works when the equity cushion is enormous. It fails when precision matters -- when the loan-to-value (LTV) ratio is tight, when the market is thinly traded, or when the property has characteristics that make algorithmic estimates unreliable.

This article walks through how to identify, evaluate, and adjust comparable sales like a professional -- without paying for one.

What Makes a Property Truly Comparable?

Not every recent sale near the subject property qualifies as a comparable. A comp must be similar enough that the price it sold for tells you something meaningful about what the subject property would sell for today. Five factors determine whether a sale is a genuine comparable or noise.

Proximity

The closer the comp is to the subject property, the more reliable it is. A sale on the same street or in the same subdivision is the gold standard. A sale one mile away in a different neighborhood is weaker. A sale three miles away across a highway, a school district boundary, or a socioeconomic dividing line may be useless.

In practice, start with a half-mile radius and expand outward only if you cannot find enough sales. In dense urban markets, you can often stay within a quarter mile. In rural areas, you may need to expand to three to five miles or even further -- which introduces its own set of problems discussed later in this article.

Recency

Real estate markets move. A sale from 18 months ago may reflect a fundamentally different market than today. Prioritize sales from the last three to six months. In rapidly appreciating or declining markets, even six-month-old data can mislead you.

If you must use older sales, adjust for market movement. County-level price indices from the Federal Housing Finance Agency (FHFA) or local MLS data can tell you whether the broader market has risen or fallen since the comp sold. A comp that sold for $120,000 twelve months ago in a market that has appreciated 8% since then has an adjusted value closer to $130,000.

Property Size

Square footage is the single most important physical characteristic for comp selection. A 1,100-square-foot ranch is not comparable to a 2,200-square-foot colonial, even if they sit on the same block. As a general rule, keep comps within 20% of the subject property's living area. A 1,500-square-foot subject should be compared to properties between roughly 1,200 and 1,800 square feet.

Lot size matters too, particularly in markets where land value represents a significant portion of the total property value. A quarter-acre lot versus a two-acre lot in a semi-rural area is a material difference that affects price.

Condition and Quality

This is where comp analysis gets difficult from a distance. You can see square footage and sale price in public records, but you cannot see whether the comp had a renovated kitchen, new HVAC, or a crumbling foundation. Two homes with identical square footage and bedroom count can trade at vastly different prices based on condition.

Look for clues in the listing data. If a comp sold above the neighborhood average, it may have been updated. If it sold below, it may have been distressed. Listing photos on Zillow, Redfin, and Realtor.com -- even from closed sales -- often remain available and can give you a visual read on the comp's condition relative to the subject property.

Bedroom and Bathroom Count

Match the bedroom and bathroom configuration as closely as possible. A three-bedroom, two-bathroom home and a two-bedroom, one-bathroom home serve different buyer pools at different price points. A missing bathroom is not a minor difference -- in many markets, the jump from one bathroom to two adds $10,000 to $20,000 or more to the sale price.

How Do You Adjust Comps for Differences?

No two properties are identical. Even the best comps will differ from the subject property in meaningful ways. The solution is to adjust the comp's sale price to account for those differences -- the same process a licensed appraiser uses, just without the formal USPAP framework.

The adjustment logic works in one direction: you adjust the comp to make it more like the subject property.

• If the comp has something the subject does not (extra bathroom, attached garage, renovated kitchen), subtract from the comp's price.
• If the subject has something the comp does not, add to the comp's price.

Here is a practical example:

When the total adjustments exceed 25% of the comp's sale price, the comp is too different from the subject to be reliable. At that point, the adjustments are doing more work than the actual sale price, and the result is more guess than analysis. Find a better comp.

Where Do Adjustment Values Come From?

Professional appraisers use paired sales analysis -- finding two comps that are identical except for one feature and attributing the price difference to that feature. You can approximate this by examining recent sales in the same neighborhood and observing price differences between homes with and without specific features.

Over time, you develop a feel for your target markets. You learn that in a particular zip code, a garage adds roughly $7,000, a full bathroom adds $12,000, and a finished basement adds $15,000 to $20,000. Until you build that market-specific intuition, use conservative adjustment values and acknowledge the uncertainty in your estimate.

These ranges are broad because adjustments are market-specific. A garage in Phoenix is worth less than a garage in Minneapolis. A finished basement in Houston (where basements are rare) is worth more than one in Ohio (where they are standard). Always calibrate to the local market.

Where Do You Find Comparable Sales Data?

You do not need MLS access or expensive subscriptions to run a solid comp analysis. Several free and low-cost sources provide the data you need.

Zillow's "Recently Sold" filter. Navigate to the subject property's neighborhood on Zillow, toggle the map to show recently sold homes, and filter by date range and property type. Zillow displays sale prices, square footage, bed/bath counts, and listing photos -- everything you need for a desktop comp analysis.

Redfin. Redfin's sold-listings data is sourced directly from the MLS in markets where Redfin operates. It tends to be more accurate and timely than Zillow in those markets, and it includes useful details like days on market and price history.

County recorder and assessor websites. Public records are the primary source for recorded sale prices. County recorder sites show deed transfers with sale amounts. County assessor sites show assessed values, property characteristics (square footage, lot size, year built), and often include a property sketch. The data is official, but the user interfaces range from excellent to barely functional depending on the county.

PropStream and similar investor tools. For around $99 per month, platforms like PropStream aggregate public records, tax data, and listing data into a single interface designed for real estate investors. They make it faster to pull comps, but the underlying data is the same public information available for free -- you are paying for convenience and speed.

FHFA House Price Index. The Federal Housing Finance Agency publishes quarterly house price indices at the national, state, metro, and county level. These are useful for time-adjusting older comps when recent sales are scarce.

How Does Listing Price Differ from Sold Price?

This is a critical distinction that new investors sometimes miss. Listing price is what a seller hopes to get. Sold price is what the market actually paid. Only sold prices matter for comparable sales analysis.

In a seller's market, properties routinely sell above listing price. In a buyer's market, they sell below. In distressed markets or with REO properties, the gap between listing and sold can be 15% to 30% or more. Using listing prices instead of sold prices will systematically bias your FMV estimate, usually upward.

There is one exception where listing data provides useful signal: days on market (DOM). If comparable properties in the area are sitting for 90 to 120 days before selling, that tells you the market is slow and buyers have leverage. If comps are selling in under 15 days, the market is hot and your FMV estimate should reflect current or slightly appreciating conditions. DOM does not change your FMV number directly, but it informs your confidence in the estimate and your assumptions about liquidation timelines if the note eventually goes to foreclosure and REO.

How Does Comp Analysis Differ in Rural vs. Urban Markets?

Urban and suburban comp analysis is relatively straightforward. Homes are built close together, transaction volume is high, and you can usually find five or more comparable sales within a half-mile radius from the last six months. The challenge is selecting the best comps from an abundance of options.

Rural comp analysis is a different exercise entirely. Transaction volume is low, properties are heterogeneous, and the nearest comparable sale may be five to ten miles away in a different town. Here is how rural markets differ and what to do about it.

Fewer Transactions Mean Wider Search Parameters

In a rural market, you may need to expand your search radius to three to five miles and extend your time frame to 12 months or even longer. This introduces more noise into the analysis, but the alternative -- having zero comps -- is worse. When you widen the search, be especially rigorous about adjustments. A comp five miles away in a different school district or a different town requires more adjustment than a comp on the next block.

Land Value Dominates

In rural markets, land often represents a larger percentage of total property value than it does in urban areas. Two homes with identical structures can trade at dramatically different prices based on acreage, road frontage, water access, or agricultural potential. Pay close attention to lot size and land characteristics when selecting rural comps.

Property Types Are Less Standardized

Urban markets are dominated by tract-built subdivisions where homes are structurally similar. Rural markets include manufactured homes on permanent foundations, modular homes, log cabins, homes with outbuildings, properties with mixed residential and agricultural use, and other configurations that resist easy comparison. If the subject property is atypical, you may need to search for comps by structure type rather than by strict geographic proximity.

AVMs Are Least Reliable in Rural Markets

Automated valuation models need transaction data to function. In a rural zip code with three to five sales per year, the algorithm is starving for inputs. This is precisely where pulling your own comps -- even imperfect ones -- adds the most value relative to an AVM. A human can evaluate a rural comp and say, "This sale is three miles away but it is a similar property on a similar road with similar acreage." An algorithm either has enough data to produce a number or it does not.

What Are the Red Flags in Automated Valuations?

AVMs are useful for first-pass screening, but certain conditions signal that the algorithmic estimate should not be trusted.

Wide confidence intervals. Some paid AVMs (HouseCanary, CoreLogic) include a confidence score or range. If the AVM reports a value of $100,000 with a confidence range of $70,000 to $130,000, the algorithm is telling you it does not have enough data to produce a reliable estimate. Treat wide ranges as a signal to pull your own comps.

No recent sales in the area. If the AVM is basing its estimate on sales from two or more years ago, the output is stale. Check the underlying data -- most AVM platforms will show you the comparable sales they used. If those comps are old, distant, or dissimilar, the estimate is weak.

Property type mismatch. AVMs can misclassify properties. A manufactured home classified as a single-family residence will receive an inflated estimate. A duplex classified as a single-family home will be valued incorrectly. Cross-reference the AVM's assumed property type with the county assessor's records.

Recent distressed sales in the neighborhood. If a cluster of foreclosures or short sales has occurred nearby, some AVMs will incorporate those distressed prices and drag the estimate down. Others will exclude them and produce an estimate that ignores real market conditions. Either way, the AVM may not be reflecting what an arm's-length buyer would actually pay. Pull the sales data yourself and make your own judgment about which transactions are representative.

Known condition issues. If you know from a BPO, a Google Street View inspection, or a door knock that the property has significant deferred maintenance, fire damage, or other condition problems, no AVM will account for that. Discount the AVM estimate accordingly or disregard it entirely in favor of a condition-adjusted comp analysis.

How Do You Handle Properties with No Good Comps?

Sometimes you cannot find comparable sales that meet reasonable standards for proximity, recency, and similarity. This happens most often with rural properties, unique property types (churches converted to residences, commercial-to-residential conversions), and markets in severe economic distress where transaction volume has dried up.

When comps are scarce, you have several options:

Widen the search incrementally. Expand the radius by one mile and the time frame by three months at each step. Document each expansion so you can assess how much confidence to place in the comps you find.

Use the cost approach as a cross-check. Estimate the value of the land separately (using vacant land sales in the area), then estimate the replacement cost of the structure minus depreciation. This is a rough method, but it provides a floor value that can confirm or challenge your comp-based estimate.

Lean on the income approach. If the property is in a rental market, estimate FMV based on what it would generate as a rental. A property that rents for $800 per month in a market where investors pay a 1% price-to-rent ratio suggests an FMV of roughly $80,000. This is an imprecise method, but it provides an independent data point.

Order a BPO. When free methods fail to produce a confident estimate, a $50 to $100 BPO from a local agent who knows the market is the most efficient path to a defensible value. The agent has MLS access, local market knowledge, and the ability to physically observe the property -- advantages that no amount of remote research can replicate.

A Step-by-Step Comp Analysis Workflow

Here is the process, start to finish, for running a comparable sales analysis on a property securing a note you are evaluating.

Step 1: Identify the subject property's key characteristics. Record the square footage, lot size, bedroom and bathroom count, year built, property type, and any notable features from the county assessor's website. This is your baseline.

Step 2: Search for closed sales within a half-mile radius from the last six months. Use Zillow, Redfin, or the county recorder. Filter for the same property type and similar size. You are looking for three to five solid comps.

Step 3: Evaluate each comp against the five criteria -- proximity, recency, size, condition, and bed/bath count. Discard comps that fail on two or more criteria. If you are left with fewer than three comps, expand the search radius or time frame.

Step 4: Adjust each comp's sale price for differences from the subject property. Add value for features the subject has that the comp lacks. Subtract for features the comp has that the subject lacks. Flag any comp where total adjustments exceed 25% of the sale price.

Step 5: Reconcile the adjusted values. If your three adjusted comps come in at $118,000, $124,000, and $121,000, you have a tight cluster and high confidence that the FMV is in the $118,000 to $124,000 range. If they come in at $95,000, $130,000, and $155,000, you have a problem -- the comps are telling different stories, and you need to understand why before settling on a value.

Step 6: Cross-check against AVMs. Pull two or three free AVM estimates and compare them to your comp-based value. If the AVMs agree with your analysis, you have confirmation. If they diverge significantly, investigate the discrepancy. The AVM may be using stale data, or you may have missed a relevant comp.

Step 7: Document your analysis. Record the comps you used, the adjustments you made, and the final estimate. This documentation protects you when you revisit the deal later, when a partner or lender asks how you arrived at the value, or when you need to compare your pre-bid estimate to the actual outcome.

How Accurate Does Your Comp Analysis Need to Be?

This depends entirely on the collateral position. The tighter the equity, the more accuracy you need.

For deep-equity deals, a rough comp analysis with conservative assumptions is sufficient. For thin-equity deals, you need the most rigorous analysis you can produce -- and you should seriously consider supplementing your own work with a professional BPO or desktop appraisal.

The goal is not perfection. The goal is an estimate that is accurate enough for the equity position at hand, produced at a cost that is proportional to the deal size. A note investor who spends 90 minutes running comps on a loan with a $20,000 UPB against a $100,000 property is overinvesting in precision. A note investor who spends five minutes glancing at a Zestimate on a loan with a $95,000 UPB against a $100,000 property is taking an unnecessary risk.

Key Takeaways

Comparable sales analysis is the foundation of property valuation for note investors. It bridges the gap between free-but-unreliable AVMs and expensive-but-precise professional appraisals, giving you a defensible FMV estimate at no cost beyond your time.

The quality of your comp analysis depends on selecting genuinely comparable properties -- matched on proximity, recency, size, condition, and configuration -- and making honest adjustments for the differences that remain. Use sold prices, never listing prices. Recognize that rural and urban markets demand different approaches. Treat AVM outputs as data points to confirm or challenge, not as answers.

Build the habit of documenting your comp analyses. Over time, the adjustments you apply and the estimates you produce become a track record you can calibrate against actual outcomes -- the ultimate feedback loop for improving your valuation skills and pricing your note acquisitions more accurately.