We Screened 1,574 Listings South of Houston. Putting Closing Costs Inside the Loan Cut the Board in Half.
Twenty-one zip codes across Harris, Galveston and Brazoria. 814 listings had enough closed sales in their own subdivision to value honestly, and 65 ranked. Then we stopped letting the borrower bring the closing costs and made the advance carry them — purchase, repairs and closing, all inside the ARV. Twenty-four files within 20% of the ask became twelve. Eleven within 12% became three. One clears at the asking price on a full rehab assumption. Six do on a light one.
The corridor, and what was actually screened
1,574 active single-family listings across twenty-one zip codes in Harris, Galveston and Brazoria counties. The growth suburbs first — Friendswood, Pearland, Manvel, Alvin, Clear Lake, Webster, Seabrook, Kemah, League City, Dickinson — then the La Porte and Shoreacres waterfront, the Brazoria towns of Lake Jackson, Clute, Richwood and Angleton, and Channelview and Sheldon out to the east.
3,551 closed sales were pulled behind them and matched inside each property's own subdivision, within 20% of its square footage and 15 years of its vintage. 814 listings cleared three or more matched sales, which is the floor for publishing a value at all. The rest get no after-repair value from us, and a thin comp set is a caveat we print rather than a reason to quietly drop a house. 65 ranked.
Zero out of pocket is a stricter test than it sounds
Most boards size a hard money loan like this: after-repair value times the advance rate, minus repairs, and whatever is left is what you can pay for the house. Then the borrower brings the closing costs. That gets called zero down, and it is not zero out of pocket.
The honest version puts all three inside the advance — the purchase price, the repairs and the closing costs. The advance covers repairs first, then closing, then whatever remains goes to purchase. Run at both a 70% and a 75% advance, because the advance rate is the dial and 75% is the ceiling. If the numbers leave a surplus, the rate comes down. If they leave a gap, the rate goes up to 75% and no further.
The closing-cost schedule is the hidden variable
Percentages break down at the bottom of the range. Three percent of a $90,000 purchase is $2,700, which will not cover origination, title, survey and prepaids on any file. The real schedule runs in bands: roughly $10,000 up to a $200,000 purchase, $11,000 to $15,000 from $200,000 to $300,000, and $18,000 to $20,000 from $300,000 to $400,000.
That has to be solved circularly, because the band depends on the purchase price the formula is trying to produce. On a $150,000 house the $10,000 band is 6.7% of the purchase — more than double what a flat 3% assumption would have charged, and every dollar of the difference came straight off the maximum offer.
What it did to the board
On the looser basis, 24 of the 65 supported an offer within 20% of the asking price and 11 within 12%. With closing costs financed instead of carried:
- 12 within 20% of the ask, down from 24
- 3 within 12%, down from 11
- One clears at the asking price on a full rehab assumption
- At a 70% advance, also one
- Every maximum on the board fell $10,000 to $20,000, and the ranking changed order
That is the whole point of running it this way. A board that quietly leaves $15,000 outside the loan is telling a borrower he can close on money he does not have.
Nine came off the board on the listing record, not the math
The remarks pass is the cheapest step in this process and it removes the top file more often than not. Nine ranked houses were struck off after reading their records:
- Financing considered read cash sale only — not a lending file whatever the spread
- A HUD listing requiring a valid NAID to bid at all
- A pre-foreclosure with showings switched off — no interior access means no repair estimate, only a guess
- A second pre-foreclosure routing offers through an online auction platform, with no possession or timing guarantee
- A patio home whose value had been built from detached sales on the same streets
- A 55-and-over age-restricted community, which cannot be rented to a general tenant
- Two already-renovated houses — new roof, fresh paint, new kitchen and bath — where the entire apparent discount was a rehab allowance applied to a finished home. Both were the largest clears on the board before the pass ran.
- One house listed under the wrong city, comped against a different suburb's sales entirely
Retail condition is the exit, never the entry. A cheap price per foot on a finished house is not a discount, and the two biggest numbers on this board were both that.
The repair band is the entire argument
Applying a full gut allowance by vintage to a house that needs paint, flooring and a kitchen fabricates cost that is not there. Applying a cosmetic allowance to a house that needs a roof and a foundation buries the borrower. So both get published:
- Full vintage band: one file clears at the asking price.
- Half that band: six clear at the asking price, and several more land inside a few thousand dollars.
Scope cannot be settled from a listing photograph. Somebody has to walk the house, and that number moves the deal in both directions.
Where the negotiation actually is
A required price cut is the job, not a failure. Four of the twelve survivors have been listed more than seventy days and one has sat 191 days. One has already come down more than $20,000 with remarks that say the seller is motivated twice over.
The market-level medians say the rest plainly. In the La Porte zip the median required cut across every valued listing was 38.8%; across Lake Jackson, Clute, Richwood, Angleton, Channelview and Sheldon it was 38%, and not one of the 119 valued listings there cleared the ask on a full rehab assumption. Those are negotiation books, not clearing books. The growth corridor to the south is where a borrower gets sent first.
Two things this board does not yet answer
Rent comps and debt service coverage are not in it. On an earlier Gulf-side run that pass reversed the conclusion — three properties cleared the purchase test while ten carried debt at the asking price — so it is being added across all twenty-one zip codes rather than assumed.
And one rule worth stating, because it is where most rental screens go wrong: a rental listed without a tenant in it does not come off the list. If the house has rent history, that history is better evidence than any subdivision median, and it survives the tenant leaving. Vacancy is not the disqualifier. No rent evidence is.
Want the addresses?
The property-level list is licensed MLS data, so it goes out one to one rather than published here. Reply to this and say which end of the corridor you work, or call 936-522-8951, and the board comes back with the maximum offer at both advances, the required cut, and the cash it would take to close at the asking price.
Market data is from public sources and is provided for information only. Nothing here is an offer of credit, a rate quote, or investment advice. Financing is subject to underwriting, lender approval, appraisal or valuation, title, insurance, property eligibility and program availability.
Financing an investment property in this market? Start with a conditional pre-approval, or run the numbers yourself.
Start a Request Get Pre-Approved Deal CalculatorEvery brief also posts on Facebook and LinkedIn — follow there for the short version.