Opendoor vs Offerpad (2026): Fees, Margins and the FTC Order
Some links pay us a referral fee; each one says so. Disclosure
Quick Answer
As of October 13, 2026, both big iBuyers state the same headline fee, 5%, and both earn a thin margin on each house. Opendoor Technologies (OPEN) calls it “our 5% service fee” in its 2025 Form 10-K; Offerpad Solutions (OPAD) shows “5% Service Fee of purchase price” plus about 1% closing costs on its own website today. In the quarter ended June 30, 2026, Opendoor sold 2,339 homes for $883 million of revenue, a 9.7% gross margin and a 5.8% contribution margin, and still lost $162 million ($100 million of that is stock compensation tied to share-price targets). Offerpad sold 206 homes for $77.7 million, a 9.2% gross margin, a 6.8% contribution margin (5.1% after interest) and a $9.3 million loss. Opendoor's cost of revenue was 90.3% of revenue and Offerpad's Cash Offer cost was 92.8% (our arithmetic), which is the ceiling on what sellers were paid per $1 of resale price, before renovation costs that sit in the same line. The FTC's final order against Opendoor (docket C-4777, October 21, 2022) required $62,000,000; no filing we read discloses a comparable order against Offerpad. Opendoor is 7.7 times Offerpad's size by 2025 revenue (our arithmetic). Neither filing uses going-concern language, but Offerpad has $33.1 million of cash, uncommitted credit lines and did a 1-for-10 reverse split on June 8, 2026.
Key Takeaways
- Fees as the companies state them: Opendoor, “our 5% service fee” (10-K filed February 19, 2026). Offerpad, “5% Service Fee of purchase price” and about 1% closing costs on its site, read October 13, 2026; Offerpad's 10-K and 10-Q give no fee percentage. The FTC complaint says Opendoor's service charge once “ranged from 6 to 14 percent” of the offer price.
- On a hypothetical $400,000 sale (our arithmetic): 5% is $20,000 at Opendoor; 5% plus about 1% closing is $24,000 at Offerpad; the 9% of traditional costs that Opendoor's old offers assumed (6% commission, 2% concessions, 1% overlap) is $36,000. The fee is not the whole price: the offer itself is set below the company's valuation, and that discount (the “spread”) is not disclosed per deal.
- Per $1 of revenue in Q2 2026 (our arithmetic): Opendoor's cost of revenue was 90.3 cents, Offerpad's Cash Offer 92.8 cents. The price paid to the seller, net of the fee, is inside that line together with renovation. Both then paid selling costs: 3.3% of revenue at Opendoor and 2.3% at Offerpad.
- Unit economics, Q2 2026: Opendoor gross margin 9.7%, contribution margin 5.8%, adjusted EBITDA of negative $4 million. Offerpad gross margin 9.2%, contribution margin 6.8%, contribution margin after interest 5.1%, adjusted EBITDA of negative $6.2 million. Contribution margin is a non-GAAP measure each company defines for itself.
- Volume collapsed at both: Opendoor revenue fell 44% to $883 million and Offerpad's 51.6% to $77.7 million against Q2 2025. Opendoor then bought 4,378 homes against 2,339 sold, taking inventory from 3,420 to 5,459 homes in the quarter, and carried $1.8 billion of asset-backed debt against $1,845 million of inventory.
- Opendoor, the FTC and the courts: FTC docket C-4777 (matter 1923191), $62,000,000, no admission of wrongdoing, final October 21, 2022; the FTC said in April 2024 it was sending refund checks to 54,689 consumers. A securities class action over its pricing algorithm ended with final approval and dismissal on January 6, 2026, settled within insurance limits.
- Capital structure: Opendoor has about 971 million shares, issued $650 million of 0.00% convertible notes due 2030 on August 19, 2026 (initial conversion price about $4.71) and planned a $158.0 million buyback. Offerpad has 4,832,749 shares after a 1-for-10 reverse split, equity of $37.3 million against a $525.9 million accumulated deficit, and credit lines it describes as “not fully committed”.
- Zillow, as context: on November 2, 2021 its board decided to “wind down Zillow Offers operations” after a $304.4 million inventory write-down in the third quarter of 2021.
CSV · 143 rows
Opendoor vs Offerpad: fees, margins, debt, share counts and the FTC order from SEC filings and FTC documents
143 rows: Opendoor and Offerpad quarterly and annual results, fee statements, debt facilities, share counts, the FTC order and Zillow Offers wind-down, each with form and accession number; derived rows are marked as our arithmetic.
What a seller pays: the fee each company states
Two numbers matter to a home seller comparing cash offers: the fee, and everything else that is taken off the price. The companies disclose the first and only hint at the second.
| Item | Opendoor (OPEN) | Offerpad (OPAD) |
|---|---|---|
| Service fee as the company states it | “our 5% service fee” (10-K for 2025, filed February 19, 2026) | “5% Service Fee of purchase price” (offerpad.com, read October 13, 2026) |
| Closing costs | Not stated as a percentage in the 10-K or 10-Q | About 1%, “standard title & escrow” (offerpad.com) |
| Where the fee shows in the filings | The purchase price “is net of our service fee and represents the cash proceeds paid to the home seller” | No fee percentage in the 10-K or Q2 10-Q; the site lists “$0 Repair Costs out of pocket”, $0 concessions and $0 commissions |
| Other deductions | Offers are also reduced by a spread: the “total discount to our home valuation at time of offer” less the service fee | The offer price itself is not itemised on the site; the 10-Q says its pricing adjustments “have had a negative impact on our operating results” |
| On a hypothetical $400,000 sale (our arithmetic) | 5% = $20,000 | 5% + about 1% = $24,000 |
| Direct selling costs paid by the company on resale, Q2 2026 | $29 million, 3.3% of revenue (our arithmetic) | $1.8 million, 2.3% of revenue (our arithmetic) |
Two cautions. First, the Offerpad figure comes from its website, not from a filing; its FAQ also tells the reader to ask for “a full fee breakdown in writing”, and we could not find a dated fee in the 10-K, so we show the page as read on October 13, 2026. Second, Opendoor's old numbers were different: the FTC's complaint says its “service charge” once “ranged from 6 to 14 percent of Opendoor’s offer price”. On our hypothetical $400,000 sale, 14% would be $56,000 against $20,000 at 5% (our arithmetic). The current 5% is a company statement in a 10-K, not a promise on any single offer.
The benchmark Opendoor itself once used for a traditional sale is in the same complaint: its offers showed six percent in agent commissions, two percent in concessions and one percent in “home overlap costs”, nine percent in all, or $36,000 on $400,000 (our arithmetic). Those were Opendoor's assumptions, quoted by the FTC, not measured costs.
What the margins say about the price a seller gets
The fee is only part of the gap between a seller's market value and the cash received. The filings let us measure the whole gap from the company side, on average, not for any one sale.
Opendoor's 10-K says the purchase price “is net of our service fee and represents the cash proceeds paid to the home seller”, and that price is booked in cost of revenue together with renovation and valuation adjustments. So cost of revenue as a share of revenue is a ceiling on the average cash paid to sellers per $1 of resale price.
| Measure (our arithmetic from the filings) | Opendoor | Offerpad |
|---|---|---|
| Cost of revenue as share of revenue, Q2 2026 | 90.3% ($797M of $883M) | 92.8% for Cash Offer ($66.6M of $71.7M); 90.8% for all revenue |
| Cost of revenue as share of revenue, same measure for the year 2025 | 92.0% ($4,371M revenue, $350M gross profit) | 94.1% for Cash Offer ($503.5M of $534.8M) |
| Gross profit per home sold, Q2 2026 | $36.8 thousand ($86M / 2,339 homes) | $34.5 thousand (reported) |
| Contribution profit per home sold, Q2 2026 | $21.8 thousand ($51M / 2,339) | $25.7 thousand (reported) |
| Average resale price, Q2 2026 | $377.5 thousand revenue per home sold ($883M / 2,339) | $356 thousand (reported) |
| Direct selling costs per home sold, Q2 2026 | $12.4 thousand ($29M / 2,339) | Not shown per home; $1.8M total |
What this does and does not show. It shows that, on average, 90 to 93 cents of each dollar of resale price went to what the companies paid for the house and fixed it, leaving them 7 to 10 cents of gross profit before selling, holding and overhead. It does not show what a particular seller lost against a listing: the resale price can differ from the market value at the day of the offer, renovation sits in the same line as the price paid, and Opendoor revenue includes a small amount of title, escrow and brokerage revenue. The FTC reached its own conclusion on this question in 2022 using Opendoor's internal data, quoted below.
The FTC order against Opendoor, and what Offerpad discloses
| Fact | What the primary source says |
|---|---|
| Case | In the Matter of Opendoor Labs Inc., docket C-4777; FTC matter/file number 1923191 |
| Complaint and proposed order | August 1, 2022 (FTC case page) |
| Final order | October 21, 2022; the Commission vote was 4-0 (FTC press release) |
| Money | “Respondent must pay to the Commission $62,000,000” (decision and order); the FTC said it expected it “to be used for consumer redress” |
| Refunds | April 3, 2024: the FTC was sending checks to 54,689 consumers (nearly $62 million in refunds) |
| What the FTC alleged | Offers “below market value on average” and costs “significantly higher than what consumers typically pay”; the order bars unsubstantiated claims about what sellers will receive or pay |
| Opendoor's own account (10-K) | The inquiry began in August 2019; the order “became final on October 21, 2022”; Opendoor “did not admit any wrongdoing” and must hold “competent and reliable supporting data” before making cost or savings claims; it paid $62 million |
| Offerpad | Q2 2026 10-Q, Part II, Item 1: not “currently a party to any actions, claims, suits or other legal proceedings” that would be material; we found no FTC order or consent decree in the filings we read |
Opendoor's 10-K lists the order as a continuing risk: “Any alleged or actual noncompliance with the consent order could have a material adverse effect on our business.” The case page on ftc.gov still shows the case status as pending, which reflects the administrative order's reporting duties, not an open allegation. The FTC also quotes Opendoor's website as saying it made money from a fee, not from “buy-low, sell-high”, the claim at the center of the case.
Opendoor's separate securities litigation over its pricing algorithm (In re Opendoor Technologies Inc. Securities Litigation) reached an agreement in principle on March 26, 2025 “for an amount within the limits of insurance coverage”, and on January 6, 2026 the court gave final approval and dismissed all claims with prejudice. Derivative suits over the same facts were also dismissed after final approval on November 25, 2025.
Context from a third issuer, only where its own filing supports it: Zillow Group's Form 8-K of November 2, 2021 says its board “made the determination to wind down Zillow Offers operations”, recorded a $304.4 million write-down of inventory associated with Zillow Offers for the third quarter of 2021 and expected about a 25% workforce reduction. It is the one large exit from this business we can source to a filing, and the reason cited was pricing unpredictability.
Unit economics, side by side
| Q2 2026 unless stated | Opendoor | Offerpad |
|---|---|---|
| Revenue | $883M (Q2 2025: $1,567M) | $77.7M (Q2 2025: $160.3M) |
| Homes sold | 2,339 (4,299) | 206 (452) |
| Homes purchased | 4,378 (1,757) | Not verified |
| Gross margin | 9.7% (8.2%) | 9.2% (8.9%) |
| Contribution margin (non-GAAP) | 5.8% (4.4%) | 6.8% (6.3%) |
| Contribution margin after interest (non-GAAP) | Not reported; about 3.1% if the quarter's $24M of property financing interest is deducted (our arithmetic, not like-for-like) | 5.1% (3.5%) |
| Adjusted EBITDA (non-GAAP) | negative $4M (positive $23M) | negative $6.2M (negative $4.8M) |
| Net loss | $162M ($29M); $100M is market-condition RSU expense | $9.3M ($10.9M) |
| Time held | 9% of homes on the market over 120 days (Opendoor's market comparison: 27%) | Average holding period of homes sold: 141 days |
| Year 2025: revenue, net loss, homes sold | $4,371M, loss $1,300M, 11,791 homes | $567.8M, loss $46.4M, 1,591 homes |
| Year 2025 contribution margin | 3.4% (2024: 4.7%) | 4.9% (5.8%); after interest 1.7% (3.4%) |
Opendoor's margin rose in Q2 2026 on fresher inventory and adjacent services, the 10-Q says, but its net loss widened mainly through $95 million of executive market-condition restricted stock units in general and administrative cost. Excluding the $100 million RSU line, the quarter's loss would be $62 million (our arithmetic). Offerpad's gross margin of 9.2% was, in its own words, “the highest level since late 2023”.
The overhead problem is visible in the gap between contribution and EBITDA: Opendoor's contribution profit of $51 million less its adjusted EBITDA of negative $4 million leaves about $55 million of operating cost the contribution measure leaves out; for Offerpad the gap is $5.3 million against negative $6.2 million, about $11.5 million (our arithmetic). A 5% to 7% contribution margin only becomes profit if volume is high enough to spread that cost, which is why both companies' Q2 revenue falls matter more than the margin rising.
Debt, collateral and share count
| June 30, 2026 unless stated | Opendoor | Offerpad |
|---|---|---|
| Cash and cash equivalents | $896M (restricted cash $66M) | $33.1M |
| Inventory | $1,845M, 5,459 homes | $94.2M |
| Secured home-financing debt | $1.8B asset-backed, non-recourse; carrying value 95.5% of inventory (our arithmetic) | $85.2M in the financing entities; 90.4% of inventory (our arithmetic) |
| Facility terms | Senior advance rates “75% to 90%” of cost; mezzanine up to 95% to 100%; mezzanine facilities at 12.50% | Senior $540M capacity, $32.5M committed, $79.8M drawn; mezzanine $80M, none committed; its facilities “are not fully committed” |
| Committed vs undrawn | $5.7B undrawn capacity, of which $25M committed | After July 29, 2026 the $200M Citibank line has no committed amount at all |
| Convertible debt | $135M 2026 Notes (due August 15, 2026) and $62M 2030 Notes at June 30; plus $650M of 0.00% notes due 2030 issued August 19, 2026 | A $14.7M revolving credit facility (long-term) |
| Equity | $914M (December 31, 2025: $1,005M) | $37.3M, accumulated deficit $525.9M |
| Shares outstanding | About 971,091,050 on July 28, 2026; 99,288,090 warrants at $9, $13 and $17 | 4,832,749 on July 27, 2026 after a 1-for-10 reverse split |
| Listing | Nasdaq | Moved from the NYSE to the Nasdaq Capital Market on August 31, 2026 (8-K of August 18, 2026) |
| Going-concern language | None found in the 2025 10-K or Q2 2026 10-Q; management expects liquidity for at least 12 months | None found; management expects liquidity for at least the next twelve months, with possible further financing |
For Opendoor the new notes are the news after the quarter. Its 8-K of August 12, 2026 priced $650 million of 0.00% convertible senior notes due 2030 at an initial conversion price of about $4.71 against a last sale price of $3.49 that day, with $52.5 million for capped calls (cap price $6.98) and $158.0 million to repurchase about 45.3 million shares. Full conversion would issue 137,960,290 shares, about 14% of the July 28 share count (our arithmetic). Whether the $135 million of 2026 Notes was repaid on August 15, 2026 is not in the filings we read; we have not verified it.
For Offerpad the March 3, 2026 NYSE notice said its average closing price had been under $1.00 for 30 trading days. It held its 2026 annual meeting on June 3, did a 1-for-10 reverse split effective June 9, and the NYSE confirmed compliance on July 1, per the 10-Q. A separate April 10, 2025 notice, about market capitalization and equity each under $50 million, gave an 18-month cure period, which runs to about October 10, 2026 (our arithmetic); we found no filing stating its outcome after the move to Nasdaq. This matters because a seller's cash offer rides on Offerpad's credit lines: the 10-Q says that lenders “may not be obligated to advance new loan funds if they choose not to do so”.
Do iBuyers work? What each year in the filings shows
Three years of Opendoor's annual table show a business that sells houses at a thin spread and has not yet turned that into a profit: revenue of $6,946 million, $5,153 million and $4,371 million for 2023, 2024 and 2025; net losses of $275 million, $392 million and $1,300 million; contribution margin of negative 3.7%, 4.7% and 3.4%. Offerpad's revenue was $918.8 million in 2024 and $567.8 million in 2025, with net losses of $62.2 million and $46.4 million and contribution margin after interest of 3.4% then 1.7%.
What the numbers say, without a view on any stock: the model earns a gross margin of 7% to 10% on revenue that is mostly resale price, pays 2% to 3% in selling costs, then pays interest on 75% to 100% advance rates and a fixed overhead. In 2025 and 2026 volumes fell because less inventory was bought in 2025 (Opendoor bought 8,241 homes against 14,684 in 2024), and each company says affordability and 6% to 6.5% mortgage rates are suppressing sales. Opendoor then bought 4,378 homes in Q2 2026, 1.87 times the homes it sold (our arithmetic), which is the opposite of the 2025 cut and the reason inventory rose 59.6% in the quarter (our arithmetic). A reader can see from these tables that the next two quarters' margins on that inventory are what the filings, not this page, will answer.
This is not a recommendation to buy, sell or hold OPEN or OPAD, and we give no price target.
What a reader can do with this
A seller comparing offers can use the filings to ask sharper questions: each company states a 5% service fee, so the offer's other lines (the spread to valuation, the repair deduction, the closing costs) decide the difference, and the only way to see them is to request the offer in writing with every line itemised and to put it beside a listing estimate with 6% commission, concessions and overlap costs stated separately. A seller who also needs to know the tax bill on the sale can start with our page on the home-sale exclusion; a landlord selling an investment house faces different rules, covered in capital gains tax on real estate, and may also compare the seller-financing route.
An investor reading the same documents can follow four numbers each quarter: homes purchased against homes sold, contribution margin, adjusted EBITDA and the share of debt that is committed. For the other side of the trade, the people who flip the homes iBuyers sell, see house flipping in 2026. Opendoor is also one of the non-REIT service companies that appear inside real estate ETFs, covered in REIT ETF holdings. Both companies' Q3 2026 10-Qs will update every figure above.
Update alert · free
An email when the iBuyers numbers change
When a rate, rule or filing behind this page changes: what changed, the one number that matters, and the source to check it yourself.
Frequently Asked Questions
All figures are from Opendoor Technologies' Form 10-K for 2025 (accession 0001801169-26-000010), its Form 10-Q for the quarter ended June 30, 2026 (0001801169-26-000021) and its 8-Ks of August 12 and August 17, 2026 (0001140361-26-032600, 0001140361-26-033739); Offerpad Solutions' Form 10-K for 2025 (0001193125-26-067147), Form 10-Q for the quarter ended June 30, 2026 (0001193125-26-330713) and 8-Ks of March 6, June 9, August 18, August 31 and September 8, 2026; the FTC decision and order, case page (matter 1923191) and press releases of August 2022, October 21, 2022 and April 3, 2024 on ftc.gov; Zillow Group's Form 8-K of November 2, 2021 (0001617640-21-000085); and Offerpad's website page read on October 13, 2026. Gross, contribution and EBITDA measures are the companies' own, non-GAAP where labelled. Per-home figures, shares of revenue, the $400,000 examples and the gaps between contribution and EBITDA are our arithmetic. This is analysis of public documents, not investment, legal or tax advice.
Keep reading.
- 0126 min read
1033 Exchange for Rental Property: Fire, Condemnation, Deadlines
Section 1033 for a landlord, read from the statute, the regulations and IRS Publications 544 and 547: a rental destroyed by fire or storm gets a 2-year replacement clock and a narrow test; a condemned one gets 3 years and a like-kind test; 4 years is only for a principal residence. With basis arithmetic, the election and the 2026 disaster notices.
- 0223 min read
506b vs 506c: 30% of Real Estate Form Ds Now Choose 506(c)
How filers actually choose between Rule 506(b) and 506(c), counted from the SEC's Form D data: 35,494 real estate notices from January 2019 to September 2026, with the 506(c) share rising from 16.7% to 30.2% in 2025 (10.3% across all industries). Plus what each choice does to offering size, minimum, unsold offerings and non-accredited buyers, the verbatim rule text and the staff's March 12, 2025 letter on verification.
- 0330 min read
Assumable Mortgages for Investors: FHA, VA and Conventional Rules
Which mortgages a buyer can assume, and whether an investor who will not live there can, read from HUD, VA, USDA, Fannie Mae and Freddie Mac texts on October 13, 2026: FHA loans closed after December 15, 1989 need an occupant, VA assumptions need no veteran and no occupancy (0.5% funding fee, 45-day decision), conventional loans carry due-on-sale, and the gap is $100,000 on a $400,000 house.