938 Triangle resales: why a good-looking spread can still lose money

938 Triangle resales: why a good-looking spread can still lose money

Alexis DeAngelisBusiness Member
Property Manager · Raleigh, Raleigh-Durham area · Member since 2025 · 52 posts · 41 votes

A lot of flip analysis stops at buy price vs. sale price. My team maintains a dataset of recorded sales across the NC Triangle, so we checked how that view holds up once rehab is counted.

Step 1: Score the house against its ZIP

Divide the asking $/sqft by the median sold $/sqft in that ZIP.

For example, the median in 27701 is $265/sqft. 80% of that is $212/sqft, or about $275K on a 1,296 sq ft house.

Here’s how past purchases resold, grouped by how far under the ZIP median they were bought. These numbers are before rehab, carry, and selling costs:

  • Under 60%: about 72% median gain, 4.3% lost money (94 resales)

  • 60–80%: about 33% median gain, 5.5% lost money (289 resales)

  • 80–100%: about 5.4% median gain, 11.8% lost money

Purchases under 80% and at or below $290K had about a 61% median gain, and 2.4% lost money (208 resales). The same tier order held going forward in all 12 markets we track.

Step 2: Subtract the rehab

Here are 938 repeat sales by price band, after estimated rehab and carry:

  • Cheapest fifth (median buy $145K): +29.35% net, 54.3% cleared zero, 174-day hold

  • Entry-level ($225K): +7.53% net, 61.7% cleared zero, 168 days

  • Middle ($310K): −9.36% net, 31.3% cleared zero, 153 days

  • Upper-middle ($434.5K): −13.17% net, 23.9% cleared zero, 196 days

  • Top tenth ($633.5K): −17.2% net, 11.7% cleared zero, 251 days

Overall, the typical resale netted −6.14%, and only 38.5% cleared zero. The two lists group deals differently, so you can’t subtract one from the other. Still, a low gross loss rate says little about what you keep.

How I’d use this

  • Treat the ratio as a screen, not a verdict. Deeper discounts resold better, but under 30% of the ZIP median is usually a wrong size, a lot, or a partial interest.

  • Choose the house, not the ZIP. Ranking ZIPs by past gross spread didn’t hold up in later sales (rank agreement 0.38).

  • Underwrite to the low end of ARV. Across 137 flips, actual ARV landed between 20% below and 43% above the modeled figure four times out of five.

  • Walk the property. Below about $158/sqft here, condition sets the price, and public data can’t see condition.

Limits: Rehab is estimated per deal, not taken from receipts. Selling costs aren’t called out in the net figures, so budget for them yourself.

If you flip in the Triangle, how close is this to your actual net, especially above $300K?

2Reply
155 views

Most Popular Reply

Diana KhanPro Member
Attorney · 10451 Mill Run Cir #755 Owings Mills, MD 21117 · Member since 2024 · 716 posts · 259 votes
1d
Quote from @Alexis DeAngelis:

A lot of flip analysis stops at buy price vs. sale price. My team maintains a dataset of recorded sales across the NC Triangle, so we checked how that view holds up once rehab is counted.

Step 1: Score the house against its ZIP

Divide the asking $/sqft by the median sold $/sqft in that ZIP.

For example, the median in 27701 is $265/sqft. 80% of that is $212/sqft, or about $275K on a 1,296 sq ft house.

Here’s how past purchases resold, grouped by how far under the ZIP median they were bought. These numbers are before rehab, carry, and selling costs:

  • Under 60%: about 72% median gain, 4.3% lost money (94 resales)

  • 60–80%: about 33% median gain, 5.5% lost money (289 resales)

  • 80–100%: about 5.4% median gain, 11.8% lost money

Purchases under 80% and at or below $290K had about a 61% median gain, and 2.4% lost money (208 resales). The same tier order held going forward in all 12 markets we track.

Step 2: Subtract the rehab

Here are 938 repeat sales by price band, after estimated rehab and carry:

  • Cheapest fifth (median buy $145K): +29.35% net, 54.3% cleared zero, 174-day hold

  • Entry-level ($225K): +7.53% net, 61.7% cleared zero, 168 days

  • Middle ($310K): −9.36% net, 31.3% cleared zero, 153 days

  • Upper-middle ($434.5K): −13.17% net, 23.9% cleared zero, 196 days

  • Top tenth ($633.5K): −17.2% net, 11.7% cleared zero, 251 days

Overall, the typical resale netted −6.14%, and only 38.5% cleared zero. The two lists group deals differently, so you can’t subtract one from the other. Still, a low gross loss rate says little about what you keep.

How I’d use this

  • Treat the ratio as a screen, not a verdict. Deeper discounts resold better, but under 30% of the ZIP median is usually a wrong size, a lot, or a partial interest.

  • Choose the house, not the ZIP. Ranking ZIPs by past gross spread didn’t hold up in later sales (rank agreement 0.38).

  • Underwrite to the low end of ARV. Across 137 flips, actual ARV landed between 20% below and 43% above the modeled figure four times out of five.

  • Walk the property. Below about $158/sqft here, condition sets the price, and public data can’t see condition.

Limits: Rehab is estimated per deal, not taken from receipts. Selling costs aren’t called out in the net figures, so budget for them yourself.

If you flip in the Triangle, how close is this to your actual net, especially above $300K?

@Alexis DeAngelis, this is a great reminder that a good purchase price doesn't always mean a profitable deal. I don't have firsthand numbers from the Triangle market to compare, but I agree that investors need to look at the full cost of a project, not just the difference between the buying and selling prices.

I'm a real estate attorney, broker, and investor based in Maryland, and I work with clients on property purchases, contracts, and real estate transactions. I've seen how title issues, contract problems, and unexpected delays can affect a transaction. Those are additional costs and risks worth considering when investors are estimating their returns.

I appreciate you sharing the research behind this, @Alexis DeAngelis! It's interesting to see how much the numbers change once expenses are considered. Happy to stay connected!

See this reply in the discussion

4 Replies

Jump to latestLatest
  • Diana KhanPro Member
    Attorney · 10451 Mill Run Cir #755 Owings Mills, MD 21117 · Member since 2024 · 716 posts · 259 votes
    1d
    Quote from @Alexis DeAngelis:

    A lot of flip analysis stops at buy price vs. sale price. My team maintains a dataset of recorded sales across the NC Triangle, so we checked how that view holds up once rehab is counted.

    Step 1: Score the house against its ZIP

    Divide the asking $/sqft by the median sold $/sqft in that ZIP.

    For example, the median in 27701 is $265/sqft. 80% of that is $212/sqft, or about $275K on a 1,296 sq ft house.

    Here’s how past purchases resold, grouped by how far under the ZIP median they were bought. These numbers are before rehab, carry, and selling costs:

    • Under 60%: about 72% median gain, 4.3% lost money (94 resales)

    • 60–80%: about 33% median gain, 5.5% lost money (289 resales)

    • 80–100%: about 5.4% median gain, 11.8% lost money

    Purchases under 80% and at or below $290K had about a 61% median gain, and 2.4% lost money (208 resales). The same tier order held going forward in all 12 markets we track.

    Step 2: Subtract the rehab

    Here are 938 repeat sales by price band, after estimated rehab and carry:

    • Cheapest fifth (median buy $145K): +29.35% net, 54.3% cleared zero, 174-day hold

    • Entry-level ($225K): +7.53% net, 61.7% cleared zero, 168 days

    • Middle ($310K): −9.36% net, 31.3% cleared zero, 153 days

    • Upper-middle ($434.5K): −13.17% net, 23.9% cleared zero, 196 days

    • Top tenth ($633.5K): −17.2% net, 11.7% cleared zero, 251 days

    Overall, the typical resale netted −6.14%, and only 38.5% cleared zero. The two lists group deals differently, so you can’t subtract one from the other. Still, a low gross loss rate says little about what you keep.

    How I’d use this

    • Treat the ratio as a screen, not a verdict. Deeper discounts resold better, but under 30% of the ZIP median is usually a wrong size, a lot, or a partial interest.

    • Choose the house, not the ZIP. Ranking ZIPs by past gross spread didn’t hold up in later sales (rank agreement 0.38).

    • Underwrite to the low end of ARV. Across 137 flips, actual ARV landed between 20% below and 43% above the modeled figure four times out of five.

    • Walk the property. Below about $158/sqft here, condition sets the price, and public data can’t see condition.

    Limits: Rehab is estimated per deal, not taken from receipts. Selling costs aren’t called out in the net figures, so budget for them yourself.

    If you flip in the Triangle, how close is this to your actual net, especially above $300K?

    @Alexis DeAngelis, this is a great reminder that a good purchase price doesn't always mean a profitable deal. I don't have firsthand numbers from the Triangle market to compare, but I agree that investors need to look at the full cost of a project, not just the difference between the buying and selling prices.

    I'm a real estate attorney, broker, and investor based in Maryland, and I work with clients on property purchases, contracts, and real estate transactions. I've seen how title issues, contract problems, and unexpected delays can affect a transaction. Those are additional costs and risks worth considering when investors are estimating their returns.

    I appreciate you sharing the research behind this, @Alexis DeAngelis! It's interesting to see how much the numbers change once expenses are considered. Happy to stay connected!

    • Member since 2026 · 2 posts · 0 votes
      18h

      Thank you for your response and I totally agree of how interesting the change in numbers can be. We follow our 12 markets, but if you have a specific one you invest in... I'd be happy to run an analysis for you on that. @Diana Khan

  • Developer · Cary, NC · Member since 2012 · 3 posts · 1 vote
    1d

    Alexis - great analysis. Curious how you are pulling this data. Is it from MLS or from your groups listings?

    • Member since 2026 · 2 posts · 0 votes
      18h

      @Matt Murphy Our sales data comes from MLS listings and county deed records. We add public data like the Census American Community Survey for income, rent and owner-occupancy by ZIP. Each market gets its own model, trained on about 3.5 years of local sales.

      To keep ourselves honest, we test every model on sales it never trained on, and we score our estimates against what those homes actually sold for. We also publish estimates before homes sell and check them after they close. In Apex–Cary, our price range caught the actual sale price 87% of the time, across 5,169 homes.

      We also say where the data is thin. On the cheapest homes, condition drives the price, and condition isn’t in any public record. So there we tell people to walk the property and pull fresh comps instead of leaning on our number.

      Happy to go deeper on any of it.

Join the conversationCreate a free account to reply, vote on answers and follow this thread.