ssj2abid@ssj2abid
People are sleeping on the $OPEN Doma acquisition + mortgage product.
This is the Opendoor 2.0 full stack. Let's do some basic math.
Here's how I understand the economics to work.
🏠 OPENDOOR 1.0: THE OLD MODEL
One revenue stream. Buy homes, sell homes, profit on the spread. Everything else goes to third parties.
🏠 OPENDOOR 1.0: SINGLE HOME TRANSACTION
Buyer X finds an Opendoor home for $400,000.
→ Opendoor bought it for $370,000
→ Spent $5,000 on repairs
→ Sells for $400,000
→ Pays ~$5,000 in selling costs
→ $400,000 - $370,000 - $5,000 - $5,000 = $20,000 profit
But during that same transaction:
→ Buyer X pays ~$4,800 in mortgage fees → goes to Wells Fargo or whoever
→ Buyer X pays ~$1,800 for title and escrow → goes to First American or whoever
→ ~$6,600 leaving the building on every home sale
→ Opendoor captures none of it
Opendoor 1.0 per home: $20,000
🏠 OPENDOOR 2.0: THE NEW MODEL
Kaz has been quietly building new revenue layers on top of every transaction:
Layer 1 → Doma (title + escrow)
Every home sale requires title and escrow. Legally required. Not optional.
→ Title = verifying the seller legally owns the home
→ Escrow = neutral party that holds all money and distributes it at closing
Opendoor acquired Doma and rebuilt this with AI. What took humans weeks now takes minutes.
Layer 2 → Opendoor Mortgage
Traditionally the buyer borrows from a third-party lender who charges rate markups, lender fees, and loan officer commissions.
Opendoor's version: AI handles the process. No loan officer. No lender fees. Buyer gets a lower rate.
Opendoor originates the loan and can sell it to Fannie Mae or other investors at a slight premium. That premium is the gain-on-sale spread → that's the profit.
🏠 OPENDOOR 2.0: SINGLE HOME TRANSACTION
Same home. Same buyer. Different economics.
→ Buyer X finds an Opendoor home for $400,000
→ Opendoor bought it for $370,000
→ Spent $5,000 on repairs
→ Sells for $400,000
→ Pays ~$5,000 in selling costs
→ $400,000 - $370,000 - $5,000 - $5,000 = $20,000 home sale profit (same as 1.0)
But now the rest stays in-house:
Mortgage:
→ Buyer puts down $50,000, borrows $350,000 from Opendoor
→ Lower rate. No lender fees. AI handles the paperwork
→ Opendoor can sell that $350,000 loan to Fannie Mae at a slight premium
→ Say Fannie Mae pays ~101% of face value
→ $350,000 × 1.01 = $353,500
→ $353,500 - $350,000 = $3,500 gain-on-sale profit
Doma:
→ Closing handled by Doma. Title search by AI. Escrow in-house
→ Buyer pays ~$1,800 for title + escrow (same fee as before)
→ But now it goes to Opendoor's Doma subsidiary
→ Doma processes at lower cost with AI
→ I estimate ~$1,000 in processing costs (conservative)
→ $1,800 - $1,000 = ~$800 profit per transaction
Opendoor 2.0 per home:
→ Home sale: $20,000
→ + Mortgage gain-on-sale: $3,500
→ + Doma title/escrow: $800
→ = $24,300
Opendoor 1.0: $20,000 per home
Opendoor 2.0: $24,300 per home
21% more profit. Same home. Same buyer. From software.
🏠 OPENDOOR 2.0: AT SCALE
At current pace (~600 homes/week = 7,800/quarter):
Doma (every transaction, legally required):
→ 7,800 × $800 = +$6.2M profit per quarter
Mortgage (estimating 30% adoption):
→ 7,800 × 30% = 2,340 mortgages
→ 2,340 × $3,500 = +$8.2M profit per quarter
Full stack:
→ $6.2M + $8.2M = +$14.4M per quarter
→ 7% contribution margin becomes ~7.5%
→ At only 30% mortgage adoption. As it grows to 40%, 50%, 60%+, margin compounds
→ Expands considerably as acquisitions/market share increases
This is why I say people are sleeping on Doma + mortgage.
Do you see the power of this model now?
Every layer Kaz adds captures revenue that used to leave the building.
This is Opendoor 2.0. And the market hasn't caught up yet.
The stock is ~$4.80.
nfa, long $OPEN 🏠
