How to find underpriced homes
An underpriced home is one listed below what comparable sales support, and the signals that actually predict it are measurable: extended days on market, one or more price cuts, a price per square foot well below the neighborhood median, listing language that implies condition or urgency, and photo sets that stop at the front door. None of these is proof on its own — the reliable method is to screen a whole market on several at once, then comp the survivors properly.
What actually correlates with a discount
| Signal | Strength | What it usually means |
|---|---|---|
| Days on market well above the local median | Strong | The market has already voted on the price |
| Two or more price cuts | Strong | A seller who has accepted they were wrong and may accept it again |
| $/ft² well under the neighborhood median | Medium | Either a discount or a condition problem — comp it to find out |
| "As-is", "cash only", "investor special", "TLC", "handyman" | Medium | Condition is known and priced, though rarely priced enough |
| "Estate sale", "relocation", "must close by" | Medium | The seller is optimizing for speed and certainty, not price |
| Six photos, all exterior | Medium | The interior is the problem, and buyers self-select out before viewing |
| Withdrawn and relisted | Medium | A reset counter hiding a long true time on market |
| Expired listing, now off-market | Situational | A seller who wanted to sell and couldn't — often the best conversation available |
| Listed under the automated estimate | Weak | Those models miss condition, which is the whole variable in question |
| "Motivated seller" | Weak | Written by an agent, on almost anything |
Any one of these on its own is noise. Three together — 90 days on market in a 20-day market, two price cuts, and exterior-only photos — is a house worth an hour.
Screen the market, then analyze the survivors
The mistake that costs most people their evenings is analyzing listings one at a time in the order a portal shows them. Reverse it: apply cheap filters to everything, and spend the expensive attention only on what survives.
1. Define the box first
Geography, property type, price band, and the minimum return you'd accept. Without it, "underpriced" has no meaning — cheap relative to what? See how to build a real estate buy box.
2. Filter mechanically on the strong signals
Days on market above the local median, at least one price cut, price per square foot below the area median. These are all machine-checkable, and together they typically take a market of 400 active listings down to 20 or 30.
3. Read the survivors for the soft signals
Listing language, photo count and coverage, relisting history, the shape of the price history.
4. Comp the shortlist properly
Now spend real time. Five or six qualifying sales, adjusted, with a stated range — how to run comps on a property. Most of your shortlist will turn out to be correctly priced for its condition, and finding that out is the point.
5. Underwrite the two or three that are left
Rent, expenses, financing, and the return at the asking price and at your offer. A discount that doesn't produce a return is a discount on something you shouldn't buy.
Do steps 2 through 5 at once
Faro scans a whole market against your buy box, underwrites every candidate on comps and market rent, and ranks them by the return they'd actually produce — so the shortlist arrives already analyzed.
Categories worth a standing search
- Inherited property. Heirs are usually splitting proceeds, often live elsewhere, and value a fast clean sale over the last 5%. Probate filings are public record in most counties.
- Tired landlords. A long-held rental with a below-market lease and deferred maintenance prices as a problem property, and the "problem" is a lease that ends.
- Failed sales. A house that fell out of contract twice carries a stigma out of proportion to whatever caused it. The inspection report that killed it is often available and often mundane.
- Bad listing execution. Terrible photos, no floor plan, an agent from two counties over, listed in December. The house is fine; the marketing suppressed demand.
- Awkward on paper, fine in person. An unusual layout, a bedroom without a closet, a functional-obsolescence quirk. These have a genuinely smaller buyer pool, which is a real discount you can capture if your exit accounts for it too.
What these share is a seller whose constraint isn't price. That's the whole game — you are not looking for someone who made a mistake, you are looking for someone optimizing for something you can supply.
Common questions
How do I know if a house is underpriced?
Only by comping it. Every signal below is a reason to look, never a conclusion — a house at $160/ft² in a $210/ft² neighborhood is either a discount or a house with something wrong with it, and only a comp set and a walkthrough tell you which. Skipping that step is how people buy the market's genuinely bad inventory at a small discount.
Are there really underpriced homes on the MLS?
Yes, though fewer than off-market marketing suggests, and they are usually underpriced for a legible reason rather than by mistake: an estate with no interest in maximizing, an out-of-area agent who mispriced, a house that failed inspection twice and now carries a stigma, a landlord tired of the property. Mispricing is much more common in the tail — unusual houses, small markets, awkward layouts — than in the middle of a liquid market.
How much below market is worth pursuing?
For a rental, whatever makes the return work — a 5% discount on a well-located property with strong rent can beat a 20% discount on something that won't lease. For a flip, you need the whole 25–30% reserve the MAO formula assumes, so a 10% discount is not a flip, it's a purchase.
Is days on market a reliable signal?
It's the most reliable single one, provided you compare it to the local norm rather than an absolute number. Sixty days is unremarkable in a market averaging 55 and a flashing light in one averaging 12. And check for a relisting — an agent who withdraws and relists resets the counter, so a "new" listing can be four months old.
Related guides and calculators
How to build a real estate buy box
A starter template, the return floors underneath it, and the criteria that quietly reject nothing.
How to find real estate comps
Where the sold data actually lives, and the six filters that separate a comp from a nearby house.
How to analyze a wholesale deal
End-to-end on one property, including the buyer's-side check most wholesalers never run.
How to calculate MAO
The maximum allowable offer formula, both versions, and how to pick the percentage honestly.
Screen a market, not a listing.
Faro scans a city against your criteria, underwrites every candidate, and ranks what's left by the return it would actually produce.
Estimates for analysis and educational use only — not financial, investment, tax, or legal advice. Verify every number independently before making a purchase decision.