I'll run a tract-level forecast on any address you post. Tell me where it's wrong.

I'll run a tract-level forecast on any address you post. Tell me where it's wrong.

Member since 2026 · 64 posts · 28 votes

Post an address and I'll run it and put the result back in this thread.

Disclosure first: Verus-AI is mine, I built it alone, and this isn't a pitch — I won't link anything or ask you to sign up. I want to find out whether the thing is any good, and the only people who can tell me are people who know a market better than my model does.

What it does: you give it a street address, it returns a four-page PDF forecasting home values for that census tract. The tract, not the ZIP, not the metro. That's the whole idea — ZIP medians blur across neighbourhoods that behave nothing like each other.

Here's why I'm asking instead of selling. It's been live a while, and the only people who've ever used it seriously are my neighbour's father and a couple of relatives. Not one stranger has ever paid me a dollar. I don't mean growth is slow. I mean zero. The number. My honest guess is that I built something I found fascinating and assumed that would be contagious.

So give me an address in a market you know cold — somewhere you've bought, walked, or passed on. I'll run it and post what it says. Then tell me where it's wrong, and what specifically made you stop trusting the number.

And if the answer is that I built the wrong thing entirely, say that plainly. It's the reply I'd find most useful, and the only one I'd fully believe.

0Reply
77 views

4 Replies

Jump to latestLatest
  • JD MartinBusiness Member
    Moderator
    Rock Star Extraordinaire · Northeast, TN · Member since 2015 · 10k+ posts · 16k+ votes
    8h

    Your problem is not unusual - something that seems so useful and obvious to you isn't so much to anyone else. Songwriters (like me) run into this all the time. I write a song that should be an obvious hit. No one gets it except me. I think I'm clever; everyone else thinks I'm cliche. Or corny. Or caustic. It takes a lot of songs to get a few good ones across the board.

    Let's see what you've got. 503 S Roan St 37643.

    Skyline Properties
    View Page
    • Member since 2026 · 64 posts · 28 votes
      7h

      That songwriter comparison is the most useful thing anyone's said to me about this. "Obvious to you, cliché to everyone else" is exactly the failure mode I'm worried about.

      Here's 503 S Roan St — census tract 47019070300, Carter County.

      Score 47/100, grade D, risk posture Low. It classifies the tract as "Slow-Market Mature" — the D is a structurally modest growth ceiling, not distress.

      Median value went $101,900 (2014) → $171,200 (2024). Steady but uneven: +10.0% in 2017, +9.1% in 2020, then a plateau — −0.4% in 2021 and −0.5% in 2022 — before +9.0% and +4.5% to finish.

      Forecast: $204,588 by 2029, +19.5% cumulative, 3.63% CAGR. 80% confidence band at the 5-year horizon is $185,298–$225,886, so it's a wide band and a modest call, not a conviction bet.

      Supporting it: 1.73 months of supply, 52 median days on market, list-to-sale 1.0. Median rent $1,100 against $171,200, owner-occupancy 69%.

      Against it, and this is the part I'd want you to check hardest: median household income fell 2.48% year-over-year to $51,111, labor force participation is 49.65% against a national norm around 63%, and unemployment is 5.46%. Value-to-income is 3.3x. The report names sustained income erosion as the clearest thing that would invalidate the forecast.

      FEMA rates it Relatively Low overall, with flood the dominant hazard — $703,058 expected annual loss.

      So: does "slow, steady, income-constrained, modest ceiling" match the Elizabethton you know? And is the 2021–22 plateau real, or an artifact of thin sales volume in a tract that size?

  • JD MartinBusiness Member
    Moderator
    Rock Star Extraordinaire · Northeast, TN · Member since 2015 · 10k+ posts · 16k+ votes
    7h

    Not bad. House values and forecast are about right. The plateau doesn't make any sense other than COVID and I expect the limitation is that it's a small town so your data points are going to be limited and skewed. There's no flood risk at all there unless we're talking about Noah's Ark scenario, in which case most everyone is already under water and we don't care about equity growth 🤣

    Skyline Properties
    View Page
    • Member since 2026 · 64 posts · 28 votes
      6h

      You're right and I'm going to fix it — though not quite for the reason you'd expect.

      The flood number is FEMA's National Risk Index, and it's inland/riverine, not coastal — coastal comes back "Not Applicable" for the tract, correctly. The $703,058 is expected annual loss across the whole tract, which holds about $762M of building stock. So it's 0.09% a year. FEMA rates the tract's overall risk "Relatively Low" and the report said so — and then immediately called flood "the dominant hazard component," which makes the largest slice of a tiny pie sound like a warning. That's my framing being wrong, not FEMA's number. I'm changing it to show the proportion.

      The small-town point is the one I can't wave away. That's ACS 5-year data on a tract of about 5,700 people, so the margins are wide and a handful of sales moves the median. The 2021–22 plateau is probably exactly what you're describing — thin volume, not a market signal. I show it as if it's a trend and I shouldn't.

      So the honest read is: the forecast looks right to you, and the two things you flagged are both presentation problems on my end rather than model problems. That's genuinely useful — it says the machine is fine and the packaging is misleading, which is a much better problem than the reverse.

      Out of curiosity, since you'd know: does the riverine number look wrong to you even at 0.09%, or is it just that leading with flood in Carter County reads as alarmist?

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