I screen inventory against one line of arithmetic and I wanted to share what it returned in Dallas this week, because the result was not what I expected.

The screen: every active single-family listing inside ten miles of downtown Dallas, $80,000 to $300,000, 1,000+ square feet, built 1950 to 2010. Thirty-five zip codes, 330 listings. No filter on days on market and no filter on listing language, because "as-is" and "investor" in the remarks tell you what the agent wrote, not what the numbers do.

Sorted by price per square foot, cheapest hundred carried forward. That hundred ran $74 to $145 a foot.

The test: ARV x advance rate, minus repairs, minus closing costs, equals the most you can pay. Run at both 70% and 75%.

ARV came from closed sales in each property's own subdivision over twelve months, matched to within 20% of size and 15 years of vintage. Fewer than three matched sales and the row gets flagged thin. No zip-code averages, which is the fastest way I know to be confidently wrong about a value.

Ninety got an ARV. Thirty went through full sizing. Twenty-four survived.

Five of the twenty-four can be bought at the asking price with the loan still inside 75% of after-repair value. Two of those clear on a plain median-condition ARV with no renovated premium assumed at all.

The part I did not expect: nine of the twenty-four cover debt service at a DSCR of 1.25 or better paying full list. That is measured against closed leases on the subject property's own street, size-matched, not a portal rent estimate. Fifteen clear 1.10. Gross yields ran 9.3% to 21.3%.

So in Dallas right now the rental math is easier to satisfy than the flip math. A flipper has to win a negotiation to make these work. On nine of them a landlord does not.

For contrast, the identical screen in Harris County returned 3,666 active listings and not one cleared at asking on a median-condition ARV. Three cleared only once the exit was priced at a renovated sale.

The six rejections taught me more than the passes:

One property's comp pool topped out at $1,925,000 because the subdivision name is shared by streets with nothing in common. Four comps, and an implied ARV near three times the asking price.

Three needed an after-repair value 28% to 91% above the highest closed sale on their own street in twelve months. If the finished house has to set a neighborhood record, that is not a comp, it is a hope.

One was a main house plus a detached second dwelling under a single price. There is nothing in the sold data shaped like that.

The pattern underneath all six: a low price per square foot usually means the house is oversized for its street, not that it is cheap.

What I am still unsure about is the repair side. I band repairs by vintage when I have no photos, roughly $50 a foot on pre-1960 stock trading cheap down to $25 on newer construction, and every maximum moves once someone actually walks the property.

For those of you buying in DFW at this price point: what are you using for a repair number before you get inside, and how far off does it usually turn out to be?