Articles · Paycheck Protection Program (PPP)
How Much Money Did Womply Really Make?
The most-cited number about the most-investigated fintech of the pandemic is "$2 billion." It appears in a congressional staff report, several hundred news articles, and at least one arbitration brief filed by a lender trying not to pay its invoices. Nobody who cites it can tell you what it counts. We rebuilt the number from the bottom — from Womply's own published partner roster, the SBA's loan-level data, an arbitration award nobody read, and the contracts the litigation pried into the open — and the strange result is that the number itself survives the audit. Everything said about it does not.
I. A number with no denominator
In December 2022, the House Select Subcommittee on the Coronavirus Crisis published a staff report titled "We Are Not the Fraud Police," and in it three claims about Womply's money:
- "Lenders paid Womply over $2 billion in processing fees"
- Womply had "a windfall 2021 net revenue of over $2 billion"
- Womply "secured a gross profit of $1.8 billion and gross profit margin of nearly 90 percent"
These sound like one fact stated three ways. They are three different accounting concepts — billings, recognized revenue, and accrual gross profit — and the report does not distinguish them or note that the first one was, at the moment of publication, demonstrably false in at least one particular: the largest single number in the "paid" column was sitting in a JAMS arbitration precisely because it had not been paid.
The press, given three numbers, chose the biggest and rounded the concept off to "Womply made $2 billion." So define "made," and count.
There are four questions inside "how much money did Womply really make," and they have different answers. Public records answer the first three:
| Question | The answer is called |
|---|---|
| What did Womply bill its lenders? | Revenue (billed) |
| What did lenders actually pay? | Revenue (collected) |
| What was left after costs and taxes? | Earnings |
| What did Womply's owners actually end up with? | Realized proceeds (not public) |
II. The fee pool: $2.5 billion, and whose name is on the checks
The money starts at the SBA. Congress paid lenders processing fees to make PPP loans — in 2021, for loans up to $50,000, the lesser of 50% of the loan or $2,500. On the 2021 books of the six Fast Lane lenders below, 99.6% of loans were for $50,000 or less, four in five paid their lender exactly $2,500, and the average fee was $2,328 (SBA loan-level data, priced loan by loan).
Womply published its own lender roster. A page titled "Womply's PPP Network" ran at womply.com/ppp/faq-ppp-overview/womplys-ppp-network/ with a table of every partner, its type, its channel, and whether it was active. The Wayback Machine captured it repeatedly (April 17, May 20, June 7, and, post-program with the complete historical roster, August 5, 2021). Cross-referenced against the SBA's loan-level data, the list reads:
Fast Lane lenders (the flagship 2021 product):
| Lender | PPP loans (SBA file) | SBA fees (lifetime) | Corroboration |
|---|---|---|---|
| Harvest Small Business Finance | 408,173 | $1,002.6M | also Subcommittee: 800k+ referred apps |
| Benworth Capital | 311,022 | $697.4M | JAMS award: 304,897 Womply loans (exact) |
| Fountainhead | 272,227 | $640.6M | loan-file litigation: "more than 263,000" loans through the platform |
| Capital Plus Financial | 395,359 (86,521 Womply) | $937.3M ($186.9M on the Womply loans) | N.D. Tex. complaint; appears only in the post-program capture; the Blueacorn Order Form that covered its Womply loans is dated May 10, 2021 |
| DreamSpring | 28,435 | $54.2M | also Subcommittee |
| Sunshine State Economic Development Corp. | 6,924 | $12.2M | |
| TMC Financing | not in SBA file under this name | de minimis |
Non-Fast Lane (referral channel — all but Funding Circle already inactive by the April 2021 capture, i.e., predominantly 2020/round-1 relationships): Cross River Bank, Celtic Bank, Kabbage, Fundbox, OnDeck, Funding Circle, plus Lendio and SmartBiz, which the table classifies as "Platform," not lenders.
First, Lendistry is not on it, although Womply's applications reached Lendistry and Womply was paid on those loans. The route ran through Harvest, and the SBA file dates the handoff. By May 4, 2021, the program's general money was gone; about $8 billion remained, reserved for community development financial institutions and minority depository institutions (ABA Banking Journal, May 4, 2021). Harvest approved 13,189 loans of its own on May 18 and 30 in the rest of the program. On May 22 Lendistry, a certified CDFI, approved 89,907 loans, and by June 29 it had approved 189,259 that the SBA file lists with Harvest as servicer: 93% of Lendistry's 2021 book. On June 8, Lendistry's chief executive emailed Harvest's Adam Seery and Womply's chief executive about fraud inquiries tied to "applicants coming through Harvest"; Seery replied that he would "connect with Womply and our accounting team to review and investigate" (House report, p. 49 and nn. 433–434). The Subcommittee's December 1, 2022 press release named Lendistry among "Womply lending partners." A federal agent's affidavit describes one of these loans from the other end: "Harvest Small Business Finance LLC purchased this loan from Lendistry before the wire was sent" (United States v. Pierre-Canel, D. Mass. No. 1:25-mj-01005, Doc. 4-1, para. 42). Lendistry's 2021 book is therefore in the fee model below, as it is in the archive's Womply loan count (more than 1,293,428 loans and $19.74 billion). Second, the names sometimes attached to Womply in secondary sources (Customers Bank, Itria/Biz2Credit, Readycap, MBE Capital, Prestamos) appear nowhere on the company's own list, in any capture. Third, the three lenders that ended up in court with Womply, Benworth, Capital Plus and Fountainhead, are all Fast Lane lenders, and the Fast Lane column is where all the money is.
On the loans that came through Womply, the six Fast Lane lenders and Lendistry collected about $3.0 billion in SBA processing fees in 2021: about $680 million at Benworth on 304,897 loans, $186.9 million at Capital Plus on 86,521, and, priced loan by loan from the SBA file, $963.6 million at Harvest, $625.1 million at Fountainhead, $512.4 million at Lendistry, $50.8 million at DreamSpring and $12.0 million at Sunshine State on their 2021 books. That is the pool the splits operate on; the non-Fast-Lane referral channel, mostly 2020-vintage, generated about $3 million. Every dollar Womply made from these seven lenders came out of this pool, as a contracted share of fees the lenders received from the government.
Several of these lenders were, before Womply, very small businesses themselves. Harvest was a niche SBA shop. Fountainhead did SBA 504 loans. Benworth was a Miami mortgage lender of, in the arbitrator's phrase, "modest financial success." Capital Plus had less than $40 million in annual revenue. These four institutions collected $3.3 billion in federal processing fees in roughly fifteen months (counting Capital Plus's whole book, not just its Womply slice). The arbitrator who examined the Benworth relationship put cause and effect plainly: without Womply's technology, Benworth "would have had to conduct manual reviews of each PPP loan applicant's information, making it impossible for it to have successfully processed billions of dollars in PPP loan applications."
III. The split: what the contracts actually said
We know Womply's economics at one lender with unusual precision, because that lender stopped paying and spent three years in arbitration losing the argument about them. Benworth's contracts — a Referral Agreement plus a Developer Order Form, both quoted in Womply's complaint in Puerto Rico — paid Womply:
- A referral fee: 1% of loan principal, expressly subject to the SBA's agent-fee cap;
- An API fee: $250 per funded loan, for the integration package (Plaid, LexisNexis, Ocrolus, Persona, DocuSign, Twilio, the Teslar portal — the entire identity-verification and loan-operations stack);
- A technology fee: a tiered percentage of the lender's SBA processing fee — 80% in the 60,001-to-300,000-loan tier — net of referral-fee credits, and waived entirely on loans where the lender's fee was $250 or less.
All-in, at the 80% tier, this works out to about 91 cents of every fee dollar the lender received on a Womply loan — a figure you don't have to model, because Benworth's own invoices prove it: Womply billed Benworth $616.7 million against the $680 million in SBA fees Benworth collected on 304,897 Womply loans. Per loan: $2,022.
Capital Plus paid on a different form. Womply's Order Form with Blueacorn, dated May 10, 2021, covered the applicants Blueacorn placed with Capital Plus and gave Womply "the first $250 from any Lender Processing Fee, plus 1/3 of the remaining Lender Processing Fee after the $250 is subtracted," with the 1% referral fee counted inside that amount (First Amended Complaint, N.D. Tex. 3:21-cv-02636, para. 65 and Order Form §§ 3.3–3.4). Applied to the $186,882,948 in fees Capital Plus received on 86,521 Womply loans, the formula gives $76,714,482.67, the exact sum Womply sued for: 41 cents of each fee dollar.
The split had a logic: the two sides ran opposite cost curves. The vendor stack itemized above ran, at Womply's expense, on every application screened, funded or not. Womply, by its own pleading, "invested heavily in marketing" to keep those applications coming, while its contract revenue arrived only per funded loan. The lender's curve ran the other way: fees collected per funded note, a 100% guarantee behind it.
Was 91% legal? This was litigated twice. A federal judge in Texas (Judge Boyle, N.D. Tex., 2022) found that the referral fees and technology fees "are for distinct services" (memorandum opinion and order) — the SBA's 1% agent cap covers helping a borrower prepare an application, not technology services sold to a lender. The JAMS arbitrator, after a seven-day evidentiary hearing featuring the former head of the SBA's Office of Capital Access, reached the same conclusion and awarded Womply $117.9 million from the lender that had argued otherwise. The SBA itself, asked in August 2021 to retroactively cap such fees, declined in writing.
Two calibration points:
First, the only adjudicated discount went one way. The Benworth contract's tier table dropped the technology fee to 70% above 300,000 loans, and the arbitrator held — against Womply — that the 70% applied retroactively to all of Benworth's loans, trimming Womply's entitlement from $616.7M billed to $551.3M. Benworth was not the only lender to stop paying. Womply's counsel swore that Fountainhead "failed to pay Womply over $90 million in fees" and that Womply had taken it to JAMS arbitration in June 2021 (Cheney affidavit, paras. 6–8); Fountainhead's own complaint says it had paid "greater than 80%" of what Womply claimed (Fountainhead complaint, p. 3 n.1). No ruling on the amounts in that dispute, or in the Capital Plus case, is in the record.
Second, the Benworth form sat near the top of the range. Capital Plus's form paid Womply 41 cents on the fee dollar; Benworth's paid 81 to 91. The Subcommittee, which obtained Womply's agreements with Harvest, Fountainhead and Benworth, wrote that Womply "often took at least half—and in some cases up to 90 percent—of all the taxpayer-funded fees allocated to lenders" (House report, p. 47). The report's body does not itemize the fees lender by lender, and the Harvest and Fountainhead agreements are not public.
IV. The build: lender by lender
Now multiply. Three kinds of number stay apart in this table: what a court record says Womply claimed or was owed; what each lender collected in SBA fees on its Womply loans; and, where no contract is public, what the Subcommittee's range of shares gives when applied to those fees.
| Lender | Womply loans | Lender fees on those loans ($M) | Fee rule | Womply's 2021 fees ($M) | Basis |
|---|---|---|---|---|---|
| Benworth | 304,897 | ≈680 | 1% referral + $250 a loan + 80% of the lender fee (70% past 300,000 loans), less the referral fee | 551.3–616.7 | Arbitrator's entitlement / Womply's claim |
| Capital Plus (through Blueacorn) | 86,521 | 186.9 | First $250 of each lender fee + 1/3 of the rest, referral fee inside | 76.7 | Womply's claim; the formula reproduces it to the cent |
| Fountainhead | more than 263,000 | 625.1 (modeled) | Not public | 450–563 | Floor: more than $90M unpaid after more than 80% paid; ceiling: 90% of the fees |
| Harvest | 402,965 | 963.6 (modeled) | Not public | 482–867 | 50–90% of the fees, the Subcommittee's range |
| Lendistry (through Harvest) | 203,207 | 512.4 (modeled) | Not public | 256–461 | Same |
| DreamSpring | 25,986 | 50.8 (modeled) | Not public | 25–46 | Same |
| Sunshine State EDC | 6,852 | 12.0 (modeled) | Not public | 6–11 | Same |
| Total | more than 1.29 million | ≈3,031 | 1,847–2,641 |
The modeled fees are the SBA's loan-level file priced loan by loan under the published schedules (86 FR 3692; 86 FR 3712). Harvest, Lendistry, DreamSpring and Sunshine State count every loan they approved in 2021, as the archive's Womply loan count does; Fountainhead's fees are those of its 2021 book, because no record gives a dollar figure for its Womply loans. The model checks out against the one lender whose numbers were litigated: run Benworth's 2021 loans through its Order Form and scale them to 304,897 loans, and it gives $618.4 million at Womply's reading of the tiers and $553.8 million at the arbitrator's, against the record's $616.7 million and $551.3 million. The figures in this table, with the record behind each one, can be downloaded from Womply's PPP fees, lender by lender.
Two things stay outside the table: the 2020 referral business, which the Subcommittee put at "just under $3 million," and the finance charges on late invoices, which are interest rather than fees.
Womply's own number sits inside the range. In a letter to the Subcommittee dated May 20, 2022, Womply put its 2021 net revenue at $2.09 billion; an April 2022 letter put its "PPP Technology Service Revenue" at $1.9 billion (House report, pp. 47 and 56). Had Harvest, Lendistry, Fountainhead, DreamSpring and Sunshine State all signed Benworth's form, the table would total $2.47 billion at the arbitrator's reading and $2.60 billion at Womply's.
The seven lenders' Womply loans come to more than 1.29 million. The FTC counted more than 3.25 million applications started through Womply and more than 1.99 million never funded (FTC complaint), which leaves about 1.26 million funded; the Subcommittee says "over 1.3 million." As a share of the pool, Womply's fees ran from 61% to 87% of the $3.0 billion the seven lenders collected on its loans. The lenders kept between 13 and 39 cents of each fee dollar, plus everything they declined to pay, for holding the license and bearing the underwriting duties that, at Benworth, the arbitrator found remained the lender's.
Answer #1: lender by lender, Womply's 2021 fees come to $1.8 to $2.6 billion, and Womply itself reported $2.09 billion of net revenue for the year. The congressional ">$2 billion" holds up as a revenue figure.
V. The collections problem: the three lenders who kept the money
If you billed about $2 billion, you did not collect $2 billion, and here the record is deliciously specific about who is the difference.
Benworth Capital collected its $680 million from the SBA, paid Womply $464,991,487, and then — with $151.7 million of billings outstanding — stopped paying entirely. Asked why, it discovered, in the arbitrator's words, an argument its own counsel "appears to have created … during the course of litigation in a creative effort to raise a legal issue where no legal issue actually existed": that Womply's entire fee structure was illegal, and that Benworth was actually owed back the $420 million-plus it had already paid. Seven days of hearings later, the arbitrator denied every substantive counterclaim, found that "it was Benworth, not Womply, that was underwriting the PPP loans," called Benworth's no-benefit position "wholly disingenuous and a complete distortion of the evidence," and awarded Womply $117.9 million.
The endgame ran through three federal courts: Womply petitioned to confirm the award in San Francisco (N.D. Cal. 3:24-cv-03975), Benworth cross-petitioned to vacate (3:24-cv-04840), and Womply chased Benworth's assets into Puerto Rico (D.P.R. 3:23-cv-01034), where the collection case was consolidated with the Federal Reserve Bank of San Francisco's own lawsuit against Benworth and its principals, Bernardo and Claudia Navarro, over its pandemic-facility borrowings. Then, on January 2, 2025, eight weeks before the scheduled hearing on the cross-petitions, Womply filed a stipulated dismissal with prejudice. The Puerto Rico claims were dismissed with prejudice in parallel, and the Fed settled with Benworth and the Navarros, confidentially, in September 2025. The lender that told Congress about Womply's "high likelihood of fraud" was, when its August 2022 document production put that quote before Congress, $86 million behind on its invoices by the arbitrator's later reckoning ($151.7 million by Womply's) — and it ended the dispute paying its vendor.
Capital Plus Financial ran the same play for ~$77 million ($9M referral, $67M technology, per Womply's complaint), lost the legal argument in front of Judge Boyle, and then settled at mediation in November 2022, with the case dismissed with prejudice on January 5, 2023. On July 26, 2021, about six weeks before Womply filed suit on September 9 (notice of removal, para. 1), Capital Plus's parent, Crossroads Systems, paid its shareholders a special dividend of about $238.9 million (First Amended Complaint, para. 52; Crossroads FY2021 annual disclosure).
Total documented withholding: more than $318 million against Womply's claims — $151.7 million at Benworth, $76.7 million at Capital Plus and more than $90 million at Fountainhead, whose CEO gave the Subcommittee its "duct tape and gum" line. The Subcommittee printed Benworth's warning about "the high likelihood of fraud involved in many of the referred loans from your company" without mentioning that the warning's author owed its recipient at least $86 million in fees. The arbitrator would later find that the contemporaneous record contained no such complaints until the invoices stopped being paid, and the report itself would be excluded from the arbitration as "rank hearsay."
Answer #2: before the settlements, Womply collected roughly $1.6 to $2.3 billion — if Harvest, Lendistry, DreamSpring and Sunshine State, which no record shows disputing, paid in full, and before whatever the Benworth and Capital Plus settlements and the Fountainhead arbitration later recovered.
VI. From collected to kept
Revenue is not earnings. Three subtractions stand between them.
Cost of revenue. The Subcommittee's own figure — "gross profit margin of nearly 90 percent" — implies that about 12 cents of every revenue dollar (at the report's exact 87.6%) went to cost of revenue, which the report does not itemize. The third-party verification stack bought identity verification, bank-account validation, document OCR, and fraud screening run across 3.25 million applications, including the roughly two million that were never funded and earned Womply nothing. The FTC would later count those two million unfunded applications as evidence of deception. They can equally be read as the cost structure of turning people away: Womply paid the verification bill on every fraudster it screened and every applicant its lenders declined.
Operating expense and legal. Pre-PPP Womply was a SaaS business with about $30 million a year in revenue (our estimate) and was unprofitable: in the last pre-pandemic year for which the Subcommittee had figures, it lost $11 million (House report, p. 57). The 2021 scale-up — support for millions of applicants — plus three arbitrations, an FTC defense, and a congressional investigation plausibly ran $150–250 million across 2021–2024 (our estimate).
Taxes. The line item every retelling of this story omits. A Delaware C-corp clearing ~$1.8 billion of gross profit pays federal 21% (26 U.S.C. § 11(b)) plus California's 8.84% (Cal. Rev. & Tax. Code § 23151) on its apportioned share — call it ~27% blended (our model). Womply's own figures do not break out a tax line; below gross profit they show $160 million of operating expenses and $379 million of "other income expenses."
| 2021, as Womply reported it to the Subcommittee | |
|---|---|
| Net revenue | $2.09B |
| Gross profit (87.6% margin) | $1.8B |
| Operating expenses | $0.16B |
| "Other income expenses" | $0.379B |
| Net income | over $1.3B |
Those are Womply's numbers, as the Subcommittee reported them (House report, pp. 56–57). The report gives each figure separately and does not reconcile them: $1.8 billion less the two expense lines is $1.26 billion.
Answer #3: by Womply's own figures, 2021 net income was over $1.3 billion.
VII. Where it went: the December 2021 sale and the escrow
In December 2021, PPP over, invoices substantially collected, three lenders in litigation or arbitration with Womply, 100% of Womply's equity was sold to a newly formed buyer, Oto Holdings, LLC, for approximately $1.09 billion, paid to its 200-plus shareholders out of Womply's own balance-sheet cash, before contributions into escrow (Atlas Technology Group v. Oto Analytics, AAA interim award, July 10, 2023).
Womply's own figures put 2021 net income above $1.3 billion; the equity cleared at $1.09 billion. The terms that would explain the gap, including any payouts to shareholders before closing, are not public. The Subcommittee's report speaks to the question twice: its summary says Womply's CEO and its President were "likely taking tens of millions in taxpayer-funded PPP processing fees as personal profits," and its body estimates from their 2020 ownership stakes that they "may be entitled to" as much as $324 million and $90 million, "assuming a $1.8 billion gross profit" (House report, pp. 5 and 58).
Out of the escrow came the $26 million FTC consent judgment (entered April 3, 2024) — a settlement whose implied statutory maximum was, on our model, $103 billion and which therefore resolved at 0.025 cents on the theoretical dollar, paid from money already set aside in 2021. The stipulated order (Section II) enters the full $26 million as a judgment, payable within seven days of entry, and suspends no part of it. Out of the escrow also came the $27 million Atlas Technology Group award, the unpaid M&A advisory fee on the sale itself — Womply's one clean loss in the post-PPP litigation ledger.
In the same season the FTC was settling for escrow money, the JAMS arbitrator was awarding Womply $117.9 million against the lender that had branded its fees illegal — a sum 4.5 times the FTC's entire judgment, flowing the other direction, in the same city, in the same year. The FTC's complaint concerned Womply's advertising: it alleged false, misleading or unsubstantiated claims to small businesses that they would get PPP loans by applying through Womply and that Womply would review and process their loan submissions within 24 hours (FTC complaint, Counts I–IV).
VIII. The scoreboard
| The question | The range | Central |
|---|---|---|
| What did Womply bill? | $1.8 – 2.6B | $2.09B reported |
| What did lenders actually pay, before the settlements? | $1.6 – 2.3B | — |
| What did Womply keep after costs? | — | over $1.3B reported |
This build says the Subcommittee's revenue accounting was approximately right: $1.8 to $2.6 billion of fees, lender by lender, against the ">$2 billion" it reported from Womply's own letter. What was wrong was every word wrapped around those numbers. "Lenders paid over $2 billion": no — three lenders withheld more than $318 million of it, and the largest holdout, Benworth, told Congress the fees were the scandal. "Windfall": the windfall column belongs to the lenders, whose seven Womply lenders kept 13 to 39 cents of each dollar of a $3.0 billion fee pool plus the withheld money, and whose broader cohort of PPP fintechs' lenders collected multiples of what any technology vendor saw. And "net revenue" is not money in pockets: after the verification stack, $160 million of operating expenses and $379 million of "other income expenses," Womply's own figures leave net income of over $1.3 billion, a figure consistent with the $1.41 billion of cash Womply said, in the Atlas arbitration, was on its balance sheet just before the December 2021 closing.
Take the verified numbers at any point in their range — $1.8 to $2.6 billion billed, $1.6 to $2.3 billion collected, over $1.3 billion kept in a single year — and they make Womply one of the most profitable tech startups founded in the last 25 years. A company that had lost $11 million in its last pre-pandemic year cleared, in one year, a profit few venture-backed companies founded since 2001 have ever reported.
The most-investigated fintech of the pandemic, it turns out, was investigated by everyone except an accountant.
Methodology, confidence tiers, and what would change the numbers
Lender roster discipline: the roster is Womply's own published "PPP Network" table (womply.com/ppp/faq-ppp-overview/womplys-ppp-network/, Wayback captures April 17 through August 5, 2021, read directly from the archive), which classifies every partner by type and channel: Fast Lane lenders Harvest, Benworth, Capital Plus, Fountainhead, DreamSpring, Sunshine State EDC, and TMC Financing; non-Fast-Lane lenders Cross River, Celtic, Kabbage, Fundbox, OnDeck, and Funding Circle; platforms Lendio and SmartBiz. It is corroborated by the JAMS award (Benworth), Womply's N.D. Tex. complaint (Capital Plus), and the Subcommittee report (Harvest, DreamSpring). Names attributed to Womply in secondary sources but absent from the company's own table — Customers Bank, Itria/Biz2Credit, Readycap, MBE Capital, Prestamos — are excluded from the model entirely. Lendistry, also absent from every roster capture, is a different case: Womply's applications reached it through Harvest and Womply was paid on those loans, so its 2021 book is in the fee model, as it is in the archive's Womply loan count (more than 1,293,428 loans / $19.74 billion).
Exact (adjudicated or court records): every Benworth figure (JAMS Corrected Final Award, June 26, 2024 — 304,897 loans, $4.02B principal, $616.7M claimed, $551.3M held due, $464,991,487 paid, $117.9M awarded; the $680M in lender fees is Benworth's figure as the award recites it); Capital Plus's 86,521 loans, $186,882,948 in lender fees, fee formula and $76,714,482.67 claim (N.D. Tex. 3:21-cv-02636); Fountainhead's "more than 263,000" loans, its "greater than 80%" paid and Womply's "over $90 million" unpaid (Fla. 18th Cir. No. 2021CA002143); Round 1's $3M and Womply's reported 2021 net revenue of $2.09B (Subcommittee); the ~$1.09B sale, $26M FTC judgment, and $27M Atlas award.
Modeled (stated assumptions): each lender's SBA processing fees, priced loan by loan from the loan-level file; Harvest's, Lendistry's, DreamSpring's and Sunshine State's whole 2021 books counted as Womply loans, as the archive's loan count does; Fountainhead's fees taken from its 2021 book; Womply's share of those five lenders' fees set at the Subcommittee's 50–90% range, with Fountainhead's floor set by its own complaint and by the Cheney affidavit, sworn by Womply's outside counsel in its JAMS arbitration against Fountainhead (their fee terms are not public — the build's widest swing).
Since resolved: the Benworth and Capital Plus disputes ended in confidential settlements from losing postures — Capital Plus at mediation (Nov. 14, 2022; dismissed with prejudice Jan. 5, 2023) and Benworth via stipulated dismissal of the cross-petitions to confirm/vacate (Jan. 2, 2025) alongside dismissal of the Puerto Rico collection case. Fountainhead dismissed its Florida suit with prejudice on Nov. 18, 2022; no award from its JAMS arbitration (No. 1100-111-808) is in the record. The settlement amounts are the last unknowns in the collections column; whatever they were, they add to the collected figures above, which conservatively exclude them.
Documents that would collapse the ranges to points: the Harvest, Lendistry, Fountainhead, DreamSpring and Sunshine State fee terms (the Subcommittee cites the Harvest and Fountainhead agreements but did not publish them); Womply's 2021 financial statements; the Oto Holdings purchase agreement; the settlement agreements and the Fountainhead arbitration file.
Sources: JAMS Corrected Final Award, Oto Analytics v. Benworth Capital Partners (No. 1210038203, June 26, 2024); Oto Analytics v. Capital Plus Financial, N.D. Tex. 3:21-cv-02636, First Amended Complaint and Mem. Op. & Order, May 11, 2022; Fountainhead SBF v. Oto Analytics, Fla. 18th Cir. No. 2021CA002143 (complaint, amended petition, Cheney affidavit); Select Subcommittee on the Coronavirus Crisis, "We Are Not the Fraud Police" (Dec. 1, 2022); SBA PPP loan-level FOIA data, priced under 86 FR 3692 and 86 FR 3712 (the archive's Womply loan count sets the loan basis); FTC v. OTO Analytics, N.D. Cal. 3:24-cv-01661 (complaint and consent order, March 2024); Atlas Technology Group v. Oto Analytics (JusMundi 2023; Del. Ch. 2023-0900); OTO Analytics v. Benworth, D.P.R. 3:23-cv-01034, complaint; SRS Acquiom secondary escrow status update (Mar. 27, 2024); United States v. Pierre-Canel, D. Mass. No. 1:25-mj-01005, Doc. 4-1 (affidavit); ABA Banking Journal, "SBA: PPP Funds Exhausted for All But CDFIs, MDIs" (May 4, 2021).