Data room / Run 1, stage 2: the nine cross-examinations

Securities and consumer finance

Cross-examination, written by the securities and consumer finance reviewer, 5,893 words, about 26 min to read

run-1/reviews/crossexam-securities-consumer-finance.md Open the file as written

Equity Flow — Cross-Examination: Securities & Consumer-Finance Reviewer (Stage 2)

Reviewer role: Senior attorney, securities law + consumer-finance regulation (same reviewer as Stage 1, reviews/securities-consumer-finance.md) Date: August 1, 2026 Materials: My Stage 1 report; the full spec (disclosures/equity-flow-invention-disclosure.txt, §§1–39); the eight other panel reports (retail-customer, investor, tax-attorney, estate-bankruptcy, real-estate-title, product-operator, insurance-actuarial, red-team). Confidence tags: SETTLED LAW / LIKELY INTERPRETATION / STATE-DEPENDENT / AMBIGUOUS / SPECULATIVE.


PRELIMINARY NOTE — POSITIONS MAINTAINED AND ONE UPDATE

I re-adopt my Stage 1 positions in full: Howey capture of investor positions (§1.1), general-solicitation destruction of 506(b) (§1.3), Gary Plastic on the packaged secondary market (§1.4), unregistered broker/adviser exposure for Noteworth (§§1.4–1.5), Equity Shield as unauthorized financial-guaranty insurance or a guarantee-security/swap (§1.6), the §3(c)(5)(C) ambiguity as the load-bearing ICA wall (§2.2), credit recharacterization as the central consumer-side risk (§3.1), UDAAP exposure independent of recharacterization (§3.3), and the marketing-vs.-spec catalog (§6).

The title reviewer's citations to Olson v. Unison Agreement Corp., No. 23-2835 (9th Cir. Aug. 7, 2025) (HEI meets Washington's Reverse Mortgage Act definition) and the CFPB amicus in Roberts v. Unlock (D.N.J. Jan. 15, 2025) (HEI is "credit"/a "residential mortgage loan"; the Reg Z investment-plan exception requires the investor to share loss of value) strengthen my Stage 1 §3.1 and §4 (reverse-mortgage functional statutes). The Roberts loss-sharing point is especially damaging here: Equity Flow's 1.0x escrow-priority capital return fails the investment-plan test more cleanly than the lump-sum incumbents. I adopt both authorities. [STATE-DEPENDENT (Olson, WA); LIKELY INTERPRETATION (Roberts framework as applied)]

One correction to my own Stage 1: red-team M5 is right that I omitted McGill v. Citibank, 2 Cal. 5th 945 (2017) (public-injunctive-relief waivers unenforceable in California; Blair v. Rent-A-Center, 928 F.3d 819 (9th Cir. 2019) rejects FAA preemption). My §3.5 arbitration analysis is amended to include it. [SETTLED LAW in CA; LIKELY INTERPRETATION federally]


QUESTION 1 — FINDINGS I DISPUTE

I dispute nothing in the tax-attorney report (its Rev. Rul. 83-51 / SAM analysis is the strongest single document in the file and is fully consistent with my §3.1), and nothing material in the title report. The disputes below are real, and several matter to the panel's aggregate verdict.

DISPUTE 1 — Estate-bankruptcy §4.1: the "SETTLED LAW" tag on the discharge-wipeout outcome is overstated. The framework is settled; the outcome is conditional.

Estate §4.1 concludes — tagged SETTLED LAW — that the investor's interest is a dischargeable §101(5) claim, only a lien survives discharge, §34 disclaims lien status, therefore "the debtor keeps the home and the future appreciation." The statutory framework is indeed settled (11 U.S.C. §§101(5), 101(12); Johnson v. Home State Bank, 501 U.S. 78 (1991); Pa. Dep't of Pub. Welfare v. Davenport, 495 U.S. 552 (1990)). But the wipeout outcome depends on an antecedent state-law characterization the estate report itself concedes is open in its own §4.2: whether the instrument is (a) a contingent contract right to payment (→ claim → discharged absent a lien) or (b) a completed pre-petition conveyance of an equitable property interest in future appreciation — which, under Butner v. United States, 440 U.S. 48 (1979), and 11 U.S.C. §541(a)/(d), never enters the debtor's estate to the extent conveyed and cannot be "discharged" because there is nothing in personam to discharge and the res is not the debtor's. The spec's §24 tax framing ("prepaid forward sale of a partial interest") and §34 recorded memorandum are the investor's evidence for branch (b). A court could go either way; branch (a) is more probable on these facts (the 1.0x repayment obligation is the tell), but "more probable" is LIKELY INTERPRETATION, not SETTLED LAW. Estate §4.2 says exactly this ("whether anything WAS conveyed... is exactly the classification question the spec leaves open"); §4.1's headline tag contradicts §4.2's hedge. Correct tagging: SETTLED LAW (framework) / AMBIGUOUS-to-LIKELY (application). This matters because downstream reviewers treated §4.1 as unconditional — see Dispute 2.

DISPUTE 2 — Red-team F2/M4: the "industrialized bankruptcy-discharge playbook" is materially overstated.

Red-team F2 asserts the extraction strategy (enroll, collect for years, file Chapter 7 in a strong homestead state, keep house and appreciation) is "SETTLED LAW away from industrialization" and that "the legal version does not even need" intent. Four objections, all from settled bankruptcy/consumer-finance doctrine:

  1. The characterization fork (Dispute 1) makes the outcome contested, not automatic. The first test case is a coin-flip with published-opinion stakes for both sides.
  2. Intent is not free. A homeowner who enrolls planning to file has made false representations at origination — spec §39(i) (no pending bankruptcy; no intent to sell) and the application file — triggering 11 U.S.C. §523(a)(2)(A)/(B) non-dischargeability for money obtained by actual fraud or materially false written statements on which the investor reasonably relied, plus §727(a)(2)–(a)(5) denial-of-discharge exposure for scheme conduct, plus the spec's own Category C personal-recourse remedy (§37). The "advertised playbook" version red-team fears — consumer bankruptcy attorneys marketing the strategy — manufactures precisely the intent evidence that defeats it. [SETTLED LAW as to §523/§727; LIKELY INTERPRETATION as applied]
  3. The trustee sells houses with equity. The target book is, by underwriting design, high-equity (§39(g): CLTV ≤ 80%). Outside the unlimited-homestead states (TX, FL — which the estate and title reviewers have already excluded from the launch footprint on independent grounds), non-exempt equity means the Chapter 7 trustee liquidates the home; the debtor does not "keep the house and 100% of future appreciation" — the trustee keeps the appreciation for creditors. The playbook's geographic domain is roughly coextensive with states the panel has already written off.
  4. Correlation is asserted, not shown. "Legal-strategy contagion" requires a published win (contested, per 1–3) plus a debtor population willing to take a Chapter 7 on their credit record to escape a contract that costs them nothing monthly.

The underlying drafting defect (estate §4.1's core point — the "not a lien" posture is self-destructive) is real and I endorse the fix. The tail is real. The industrialized-certainty framing is not. Correct severity: material unpriced tail, MEDIUM likelihood even after an adverse first case — not the correlated book-wide loss red-team prices. [LIKELY INTERPRETATION]

DISPUTE 3 — Product-operator §2.3 (FCRA): right statute, wrong pressure point.

Ops tags as LIKELY that the platform is "plausibly a 'user' of consumer reports — triggering permissible-purpose requirements and adverse-action notice duties." Two corrections from my domain:

  • Permissible purpose is not the live problem for the application flow. For a consumer-initiated application, 15 U.S.C. §1681b(a)(3)(A) (credit) or (a)(3)(F)(i) (legitimate business need in connection with a consumer-initiated business transaction) covers the pull regardless of the credit/not-credit fight. The permissible-purpose exposure is confined to pre-application uses — prescreening the "free tool" audience against tradeline data without complying with the firm-offer rules of §1681b(c). That narrower point is real and worth its own finding; the general framing overstates it.
  • The adverse-action duty is firmer than "plausibly." §1681m(a) applies to any "adverse action," and §1681a(k)(1)(B)(iv) defines adverse action to include any determination adverse to the consumer's interests made in connection with a consumer-initiated transaction — expressly not limited to credit. A lock-in-score denial or a silent payout haircut (retail C7) based in part on a consumer report requires an adverse-action notice even if Equity Flow wins the not-credit argument everywhere. [SETTLED LAW as to the statutory text; LIKELY INTERPRETATION as applied]

Net: ops's conclusion (the pipeline is unusable without an adverse-action framework) survives; its mechanism should be corrected before anyone relies on it, because "permissible purpose" is the branch the platform would actually win.

DISPUTE 4 — Retail-customer V1/verdict: "payday-loan to subprime priced" treats ex-post realized cost as if it were the legal measure of unfairness. It is not.

Retail's cash-flow modeling is competent and I adopt its disclosure findings (T1–T9) without reservation. But the rhetorical move from "34.7% APR-equivalent in a 3-year/5%-HPA exit" to "payday-loan priced" imports a legal conclusion my domain does not support:

  • Contingent-consideration doctrine. In usury and cost-of-credit analysis, compensation contingent on a bona fide hazard is traditionally evaluated ex ante, not on the realized path; where repayment of the return component is genuinely at risk, most states' usury law does not treat the realized upside as interest (e.g., the Rubenstein v. Small, 273 A.D. 102 (N.Y. App. Div. 1947) line: repayment subject to hazard defeats loan/usury characterization). [STATE-DEPENDENT; the doctrine is old but durable]
  • Retail's own table proves the contingency: the identical contract prices at 0.1%–4.4% in flat markets. A single instrument is not "payday-loan priced" and "nearly free" at once; it is a contingent instrument whose expected cost is the fair metric — and whose realized cost should be disclosed as a scenario range (which is retail's actual, correct prescription).
  • What survives, and it is worse for the sponsor than the rhetoric: the components that are not contingent — the 1.0x repayment floor (assured up to proceeds), the origination fee on projected disbursements, and the gross-basis repayment mechanic — are exactly the facts that weaken the contingency defense if the product is recharacterized (my Stage 1 §3.1; tax D-1). Post-recharacterization, hot-market early exits genuinely can pierce state caps (tax D-4(d)). The usury exposure is real; the "payday" framing is the wrong doctrinal path to it.

[LIKELY INTERPRETATION]

DISPUTE 5 — Retail-customer P4 (and my own Stage 1 §3.5): the arbitration-cost premise is partly wrong; the real defect in §38 is one neither of us named in Stage 1.

Retail P4 (and my §3.5) assumed the homeowner must "fund AAA arbitration" at prohibitive cost. Under the AAA Consumer Arbitration Rules and Consumer Due Process Protocol, AAA applies its consumer rules to arbitration clauses in consumer-facing agreements regardless of the rules the drafter designated, caps the consumer's filing fee (currently ~$225 territory) with the business paying the balance, and reviews clauses for Protocol compliance before administering. So the raw cost-asymmetry point is overstated — a correction I apply to my own Stage 1 as well.

The sharper defect: spec §38 designates the AAA Real Estate Industry / Commercial Rules with fees "split equally" for a mass-market consumer product. That drafting either (a) gets overridden by AAA's consumer-rules policy (making §38's cost allocation illusory), or (b) leads AAA to decline to administer for Protocol noncompliance — and a clause whose designated forum refuses to administer it is unenforceable or collapses into court in a number of jurisdictions, with the drafter bearing the consequences of its own clause. Either branch defeats the clause's purpose. [LIKELY INTERPRETATION as to AAA practice; STATE-DEPENDENT consequences]

DISPUTE 6 — Investor O2 and product-operator §5.1: tagging Equity Shield's classification "AMBIGUOUS" understates the exposure in a way that changes launch posture.

Investor O2 ("Classification of Equity Shield (insurance? swap? security?): AMBIGUOUS") and ops §5.1 ("AMBIGUOUS for Equity Shield") are correct that the bucket is uncertain. But an unqualified AMBIGUOUS tag reads as "launch and see." The correct structure — my Stage 1 §1.6, now joined by actuarial 2.1 (LIKELY unlicensed insurance; NY Ins. Law §1101(a)/§1102; Article 69 monoline requirement SETTLED LAW) — is: it is AMBIGUOUS which regulated regime captures it, and LIKELY (approaching certain) that at least one does — insurance (N.Y. Ins. Law Art. 69 analogs), guarantee-of-a-security (Securities Act §2(a)(1)), or swap/security-based swap (CEA §1a(47); Exchange Act §3(a)(68)). There is no identified characterization under which an unlicensed platform can lawfully sell a fee-funded 1.0x floor. The distinction matters: "ambiguous" invites a risk memo; "no lawful path as structured" requires withdrawal, which is where the actuarial reviewer and I both land.

DISPUTE 7 — Red-team F10 (AML): "either answer is bad" overstates; the exposure is contingent and narrower.

Two refinements: (a) The BSA question is not free-floating AMBIGUOUS — it is contingent on recharacterization. FinCEN's "loan or finance company" category (31 C.F.R. §1010.100; program rules at 31 C.F.R. pt. 1029) covers non-bank residential mortgage lenders and originators. If Equity Flow is recharacterized as mortgage lending, AML program + SAR duties attach; if the equity characterization holds, the platform is not obviously any BSA financial institution. Same whipsaw as everything else, and it belongs in the recharacterization-consequences stack (my §3.2), not as an independent ambiguity. (b) FinCEN's Residential Real Estate Rule (31 C.F.R. pt. 1031, eff. Dec. 1, 2025) targets non-financed transfers of residential real property to legal entities and trusts, reported by persons in the closing chain. Equity Flow position assignments (Flow 6) transfer a contract/recorded interest, not residential real property, and settlements occur inside ordinary financed closings — so the likelier answer is the rule simply does not reach the assignment corridor, rather than "unregulated corridor advertising itself to launderers." The investor-side KYC gap red-team identifies is a genuine diligence/reputation finding; the regulatory framing needs the correction. [LIKELY INTERPRETATION]

DISPUTE 8 — Product-operator §5.2: "each monthly ACH is arguably a new advance" is the weaker mechanism; the stronger one is void-contract exposure.

Ops's C&D-fuse finding (a state action forces the fund to choose between unlicensed lending and three-missed-payment forfeiture) is correct and important. The mechanism needs tightening: whether continuing performance of pre-reclassification contracts constitutes new unlicensed lending is genuinely STATE-DEPENDENT (many licensing statutes penalize the making of loans; regulators typically address existing books through the order's terms). The sharper and less escapable exposure is that several states void or render uncollectible loans made without a required license (and some void the interest or the entire obligation) — meaning recharacterization can retroactively convert the existing in-state book into unenforceable claims regardless of what the fund does prospectively. That is worse than ops's framing, not better, but it is a different fuse: the harm crystallizes at the moment of recharacterization, not month-by-month afterward. [STATE-DEPENDENT]


QUESTION 2 — PROPOSED FIXES THAT CREATE NEW PROBLEMS IN MY DOMAIN

FIX-CONFLICT 1 — "Record a true security instrument / performance deed of trust" (estate fix #1; title fix #1). This is the right fix and it detonates the consumer-finance posture. Everyone should price that honestly.

The title and estate reviewers are correct that lien status is the only architecture that survives foreclosure, probate, homestead, §6323, and bankruptcy. But the concession cascade in my domain is total:

  • Credit characterization near-concedes. A recorded mortgage/deed of trust "securing the homeowner's obligations," plus the 1.0x repayment obligation, is the classic SAM architecture (tax D-1; Rev. Rul. 83-51). Post-Roberts amicus, the Reg Z investment-plan escape requires sharing loss of value — which the 1.0x floor forfeits. Once the instrument is a mortgage in form, the "not credit" argument survives only on economics the product doesn't have. [LIKELY INTERPRETATION]
  • §1639c(e) voids §38. If it is a dwelling-secured consumer credit transaction, the mandatory arbitration clause and class waiver are prohibited outright (15 U.S.C. §1639c(e); 12 C.F.R. §1026.36(h)). [SETTLED LAW, contingent]
  • Licensing triggers become facial. Many state mortgage-licensing and SAFE Act MLO definitions key to taking a security interest in a dwelling in connection with an advance of money — no recharacterization fight needed once a deed of trust is recorded; the license obligation is triggered by the instrument itself in those states. [STATE-DEPENDENT]
  • TILA assignee liability poisons the Tier 2 thesis. 15 U.S.C. §1641(a), (c): assignees inherit liability for violations apparent on the face of the disclosure statement, and rescission is assertable against any assignee. Every seasoned position sold to an insurer carries a three-year §1635 rescission tail if origination disclosures were absent (they will have been — the product was originated as "not a loan"). No institutional buyer's counsel accepts that paper. [SETTLED LAW as to §1641 mechanics]
  • Retroactivity — answering the question directly: recording mortgages prospectively does not, by itself, retroactively convert the prior book; TILA obligations attach at consummation based on the transaction's actual character. But (i) the new instrument is a party admission usable against the old book's characterization, and (ii) if a court recharacterizes, the consequences run from consummation anyway (the §1635 tail, HOEPA points-and-fees testing of the 2–3% fee, usury). The fix therefore forces a choice: run two incompatible characterizations across two vintages, or concede for both. [LIKELY INTERPRETATION]
  • One partially compensating path nobody briefed: as a dwelling-secured, non-recourse advance repayable at sale/transfer/death, the fixed product may be a Reg Z reverse mortgage (12 C.F.R. §1026.33 — no age element in the regulation's definition; accord Olson under WA law), which brings §1026.33 disclosures and an exemption from ability-to-repay (§1026.43(a)) — but also drags in state reverse-mortgage counseling and licensing statutes, i.e., precisely the HUD-counseling burden spec §26 brags about escaping. There is no version of Fix 1 that preserves §25's "lighter regulatory path." [AMBIGUOUS as to §1026.33 fit; STATE-DEPENDENT]

FIX-CONFLICT 2 — Hold-to-organic-exit / evergreen 20–30-year fund (actuarial fix (ii); investor's "15-year evergreen vehicle honestly marketed"). Investment Company Act and Advisers Act consequences.

  • Exclusion arithmetic gets worse, not better. A multi-decade vehicle cannot realistically hold LPs captive under §3(c)(1) (≤100 owners, no public offering) without secondary transfers; every transfer must be policed for owner-count/QP status for 25–30 years. §3(c)(7) confines the raise to qualified purchasers ($5M/$25M — 15 U.S.C. §80a-2(a)(51)), excluding the "family office long tail" the marketing targets. §3(c)(5)(C) remains AMBIGUOUS (my Stage 1 §2.2) and a longer hold means more years of exposure to an adverse staff or judicial position with a fully deployed, unsellable book.
  • If the vehicle offers periodic liquidity (the natural demand of LPs told the fund now runs 25 years), interval-fund/tender-offer mechanics push toward registered closed-end status (Rule 23c-3) — full ICA registration, board, custody, leverage limits, and daily/periodic NAV on assets with no observable price. The alternative — promising redemptions from an unregistered vehicle holding illiquid consumer-linked paper — is the classic liquidity-mismatch enforcement fact pattern.
  • Adviser exposure compounds with duration. The manager must register (§203(m)'s $150M private-fund exemption fails at the stated ambition); a 25-year fund whose returns exist only as sponsor-marked valuations of unobservable assets faces Rule 206(4)-8 exposure on every mark, and any management fee or carry computed on unrealized appreciation is a standing conflict the PPM must carry. Red-team F11/M3's wash-trade scenario becomes the base-case temptation of a fund that cannot show realized exits for decades; note also Advisers Act §206(3) requires disclosure/consent for any principal or cross transaction with an affiliated Tier 2 vehicle — even a fair-priced affiliated first sale is unlawful without it. [SETTLED LAW as to §206(3); LIKELY INTERPRETATION as applied]
  • The pension thesis hits ERISA. Passive contingent appreciation contracts qualify for neither the VCOC nor the REOC exception under the plan-asset regulation (29 C.F.R. §2510.3-101; ERISA §3(42)); pension money above the 25% threshold makes the manager an ERISA fiduciary with prohibited-transaction exposure on every affiliate fee — including the origination fee. The Tier 2 pension buyer the restructured fund still needs is the buyer the restructuring makes hardest to take. [SETTLED LAW as to the framework; LIKELY INTERPRETATION as applied]

FIX-CONFLICT 3 — "Honest APR-equivalent cost disclosure" (retail verdict, minimum bar). Answering the posed question: does disclosing an APR-equivalent concede credit characterization?

No, not as a matter of law — characterization is substantive in both directions; a voluntary disclosure no more converts equity into credit than the "not a loan" label converts credit into equity (Joiner; economic-substance doctrine cuts symmetrically). [LIKELY INTERPRETATION] But three real costs:

  1. Party-admission evidence. A sponsor-computed "APR" will be quoted in every recharacterization brief and CFPB exhibit ("the company itself priced this as credit"). Manageable with labeling ("illustrative annualized cost of a non-credit transaction"), but the litigation-optics cost is nonzero and permanent.
  2. A new deception vector. A contingent instrument has no single APR. One number is cherry-picking (deceptive in whichever direction it errs); a 0.1%–34.7% range is accurate and useless. The defensible format is the one emerging state HEI statutes already prescribe — scenario-based dollar-cost tables with an annualized-cost column (e.g., Connecticut's §36a-498i disclosure regime) — which supplies a compliance safe harbor the home-brewed "APR-equivalent" lacks. [STATE-DEPENDENT]
  3. It must go in the investor materials too. Once the sponsor computes the homeowner's cost of funds, omitting it from the PPM while disclosing it to consumers (or vice versa) is a Rule 10b-5/§17(a) omission problem of the sponsor's own making.

Bottom line: do the disclosure — non-disclosure is the worse UDAAP exposure (my §3.3) — but in the statutory HEI-table format, never under the label "APR."

FIX-CONFLICT 4 — Hardship desk, pause governance, homeowner-elective suspension (ops 3.1; red-team M1; the Medicaid fix implied by estate 2.4(a)). Servicing-obligation and characterization costs.

  • Licensing: answering the posed question — hardship/loss-mitigation infrastructure does not itself require licensing while the not-credit characterization holds; state servicer-licensing statutes key to residential mortgage loans. But it is contingent exposure: on recharacterization, the entire apparatus (payment administration, hardship review, payoff quotes, annual statements) is unlicensed mortgage servicing in the licensing states, and the servicer-of-record the other fixes require (Fix-Conflict 8) must be licensable on day one. [STATE-DEPENDENT]
  • Substance-over-form cost: a lender-style loss-mitigation function is affirmative evidence for recharacterization (it is what creditors do; investor AS4's "lender behavior" point). Priced correctly, still worth it — the UDAAP/abusiveness exposure of the unmitigated pause (my §3.3) exceeds the evidentiary cost.
  • The elective-suspension trap: giving the homeowner the right to decline or suspend disbursements at will (the natural fix for estate 2.4(a)'s Medicaid problem) converts the "single fixed committed amount" into a draw-at-will facility — the functional shape of an open-end credit plan (12 C.F.R. §1026.2(a)(20) factors) and the destruction of the tax defense in tax P-5. This specific fix should be rejected in favor of a narrow, event-defined deferral (documented benefits-eligibility events only). [LIKELY INTERPRETATION]
  • AVM governance bonus: if recharacterized, the quarterly pause AVMs fall under the interagency AVM quality-control rule (Dodd-Frank §1125, 12 U.S.C. §3354; final rule eff. 2025) — model governance is then mandatory, not best-practice. Build it now either way. [SETTLED LAW as to the rule; contingent application]

FIX-CONFLICT 5 — Regulatory-event suspension clause in the homeowner contract (ops 5.2's minimum fix). This clause is itself a UDAAP problem.

A term letting the investor suspend "binding" payments upon regulatory events is (a) flatly inconsistent with the "binding obligation" marketing (§31) — so it must be disclosed prominently, which ops itself concedes is a conversion-killing disclosure; (b) an unfair-term candidate under §5531(c) (substantial injury — loss of budgeted income; not reasonably avoidable — the trigger is the sponsor's compliance failure; no countervailing consumer benefit); and (c) an invitation to an illusory-promise/unconscionability attack on formation. If drafted, it must be symmetric (suspension tolls the term, freezes the ratchet, and gives the homeowner a no-penalty termination right with appreciation-share forfeiture intact) — i.e., the version that protects the fund least. [LIKELY INTERPRETATION]

FIX-CONFLICT 6 — Equity Shield rebuilds (actuarial 2.2 and verdict; my own Stage 1 §8 fix (a)/(b)). Both escape routes have securities costs the proposals don't price.

  • LP-owned first-loss reserve (the actuarial/my preferred route): solves the insurance-licensing problem only if nothing about it is marketed as a guarantee — a "guarantee of a security" is itself a security (Securities Act §2(a)(1)), and "guaranteed 1.0x floor" language re-imports everything. Structurally it tranches the fund into senior/first-loss classes: a structured product requiring full offering-document treatment, class-by-class conflicts disclosure, and — the point actuarial 2.2 made that I extend — a mutual-destruction problem with the sales pitch: the feature is sold to insurers as risk transfer for RBC purposes while the securities disclosure must say no third party bears risk and the reserve may be insufficient. If the capital-treatment claim is true, it is a guarantee; if the disclosure is true, the capital-treatment claim to the insurer's regulator is false. One of the two documents is always wrong. [LIKELY INTERPRETATION]
  • Licensed-carrier / reinsurance / ILS route (my Stage 1 fix (a)): in New York-pattern states, financial guaranty is monoline-only (Art. 69) — no multiline carrier or captive can front it, and no monoline will write an unrated, dataless asset class. The realistic transfer is ILS/cat-bond format — but ILS notes are securities (another exempt offering to run correctly), and a cash-settled payout keyed to home-price levels sits near the swap/security-based-swap boundary (CEA §1a(47); Exchange Act §3(a)(68)) with eligible-contract-participant restrictions. Feasible for a mature sponsor; not a launch-window fix. [STATE-DEPENDENT / AMBIGUOUS]

FIX-CONFLICT 7 — Homeowner-side credit protection via escrowed or LC-backed committed capital (estate fix #6). Money-transmission/escrow licensing, and a collision the panel should see.

Pre-funding 84 months of disbursements into an account for the homeowner's benefit is the right consumer protection, but: (a) a platform holding and periodically releasing consumer-destined funds is doing escrow/money-transmission in a number of states absent an exempt bank/trust structure — use a bank trustee, not the platform [STATE-DEPENDENT]; (b) if the escrow gives the homeowner a sufficiently vested right to the fund, the tax reviewer's constructive-receipt analysis (tax P-4) flips against the "no income on receipt" position — the account must be genuinely conditioned; and (c) this fix trades directly against investor C3/ops 4.1 (reserve drag; the platform cannot fund the machine) — see DIRECT CONFLICTS #7.

FIX-CONFLICT 8 — Servicer-of-record, MERS-analog registry, binding payoff/subordination machinery (title fixes #3–4; ops 3.4/6.2; red-team M2). Endorsed — with two licensing riders.

This fix reduces my domain's risk overall (it is the answer to red-team M2's assignee-weaponization). Riders: (a) the permanent servicer must be built licensable as a mortgage servicer in advance, because it inherits the book on the day any state recharacterizes [STATE-DEPENDENT]; (b) any entity acquiring or servicing positions after default collects a "debt" under the FDCPA if recharacterized (15 U.S.C. §1692a(6)(F)(iii) — the in-default-at-acquisition rule), importing validation-notice and communication rules the assignment chain nowhere contemplates. [SETTLED LAW as to the rule; contingent application]

FIX-CONFLICT 9 — Restoring a priced spread/premium to fix loss-absorption (investor verdict's alternative). Confirmed lethal; adding the consumer-finance half.

The investor report already concedes this "resurrects the debt/tax problem." Completing the picture: a certain spread on advanced funds is a finance charge in substance (12 C.F.R. §1026.4 analog), collapses the contingency defense to usury (Dispute 4), and makes the Roberts investment-plan exception unavailable beyond argument. The design is genuinely trapped: every economic fix for the investor's uncompensated risk is a step toward credit characterization. [LIKELY INTERPRETATION]

FIX-CONFLICT 10 — Re-basing the origination fee to a percentage of home value (ops 1.1's fix). Fee-cap and HOEPA exposure.

Raising the homeowner-paid fee from ~$3,150 to incumbent-style 3–3.9% of a larger base fixes ops's P&L and walks into: HOEPA points-and-fees triggers on recharacterization (15 U.S.C. §1639; 12 C.F.R. §1026.32 — a high-cost mortgage brings §1639's own arbitration ban, counseling, and enhanced liability); emerging state HEI statutes with fee caps (Maine-2026 pattern); and a bigger UDAAP omission problem given that the homeowner page currently discloses no fee at all (my §6.A6). The fee can be raised only inside a rebuilt disclosure stack. [STATE-DEPENDENT / SETTLED LAW contingent]

Convergence note (a fix that helps my domain)

The tax reviewer's minimum redesign — lump-sum funding plus genuine investor participation in depreciation — is the only proposed fix that materially improves the consumer-finance posture: sharing loss of value is precisely what the Reg Z investment-plan exception (per the Roberts amicus) requires, and removing the monthly cadence removes the open-end-credit and reverse-mortgage-tenure analogies (my §3.1). The panel should be clear-eyed that tax, consumer-finance, and title all converge on the same answer — and that answer is Hometap, i.e., the abandonment of both claimed innovations.


DIRECT CONFLICTS

Every direct conflict I see between any two reviewers, including those not involving me.

DC-1. Title/estate vs. tax (and my Stage 1) — the lien fix vs. the characterization posture. Title (verdict fix #1) and estate (fix #1) say: record a true deed of trust; it is the only structure that survives foreclosure, probate, homestead, §6323, and bankruptcy. Tax (verdict ¶2) and spec §25's whole posture say: debt-like form features destroy the equity characterization on which tax and regulatory positioning depend; my §3.1/§3.2 prices the consumer-finance cascade. Incompatible because the same recorded instrument cannot be a mortgage for enforcement and not-a-mortgage for TILA/licensing/tax; estate's own "kill condition" paragraph concedes the incompatibility, and red-team's verdict ¶1 names it. This is the panel's central unresolved fork, and it is a decision, not a drafting task.

DC-2. Estate §4.1 vs. title §5.4 — same instrument, opposite confidence. Estate tags the Chapter 7 discharge-wipeout outcome SETTLED LAW; title tags the bankruptcy characterization of the identical interest AMBIGUOUS (§5.4, cross-referred). Incompatible tags on the same question; the resolution (per my Dispute 1) is that the framework is settled and the application is not, which means red-team's downstream escalation (F2/M4) inherited an overstated premise.

DC-3. Investor A2/B2 vs. insurance-actuarial 1.1/1.2 — is "capital preserved" false or true? Investor: "the capital return is not a floor; it is a junior, non-recourse residual claim... the central risk representation to investors is wrong" (CRITICAL). Actuarial: distressed sales essentially never breach the floor even at −45% (equity cushion + escrow priority); the marketed bear case (1.02x) holds in their model; the real early-cycle loss is duration, not principal. Incompatible severity assessments of the same representation because they model different loss channels (foreclosure/BK wipeout vs. sale-path recovery). My adjudication as securities counsel: both are right on their own channel, and the marketing is still actionable either way — "downside capped at capital already disbursed" is misleading not because loss is probable but because the spec's own identified wipeout paths (Event 4 foreclosure; short sale; bankruptcy) are affirmatively denied on the page. Rule 10b-5 materiality does not wait for the actuarial base case. [LIKELY INTERPRETATION]

DC-4. Investor A10 + actuarial 1.2 vs. retail V-table/T4 — the pause paradox. Investor A10: on the spec's own downside path, the quarterly AVM pause should have stopped disbursements within months, capping deployment near $4.5–9K — so the spec's $54K downside case is impossible if the pause works. Actuarial 1.2 models the pause freezing deployment (consistent with investor). Retail's cost tables and its T4 worked example assume full 36-month deployment through a monotonic decline (consistent with the spec's own §20). Incompatible models of the same contract: either the pause fires (retail's down-market owed-at-exit figures are overstated, and the spec's flagship example is wrong) or it doesn't (the marquee safety feature is inoperative, per ops 1.4's AVM-lag reading). Whichever way it resolves, one of the sponsor's two safety narratives is false — which is itself a §17(a)/UDAAP finding — but the panel should reconcile whose numbers govern Stage 3 scenarios.

DC-5. Actuarial fix (ii) vs. investor C1/C4/B5 — the mandatory restructuring produces an unraisable fund. Actuarial: re-underwrite everything to hold-to-organic-exit (mean exit ~year 24) and market the true duration. Investor: an LP base that accepts a 15–20+ year, negative-carry, ~7%-net consumer-linked vehicle "does not exist" (C1), and the honest version is "a different, smaller, slower product with no right to win." Incompatible because one reviewer's precondition for lawful marketing is the other reviewer's demonstration of commercial impossibility; my Fix-Conflict 2 adds that the restructured vehicle also narrows to QPs-only under the ICA. The panel cannot endorse "fix-before-launch" and "the fixed version is unfundable" simultaneously without saying which governs.

DC-6. Investor's spread-restoration alternative vs. tax D-1/P-2 (and my §3.1). Investor's verdict floats "a priced spread or premium restoring loss-absorption" as a fix component; tax holds that a certain return on advances is interest and the 1.0x floor is already the confession (D-1). Incompatible — and investor itself concedes the trap in the same sentence ("resurrects the debt/tax problem... trapped between its tax posture and its economics"). Flagged so Stage 3 does not treat the spread as an open option; it is not.

DC-7. Estate fix #6 vs. investor C3 / ops 4.1 — consumer protection vs. fund economics. Estate requires escrowed/LC-backed committed capital so homeowners are not unsecured creditors of the fund (6.2). Investor C3 shows the fund already suffers negative convexity from holding reserves against pause-resumption, and ops 4.1 shows the platform cannot fund its own operating machine. Incompatible in degree: the homeowner-protection fix converts ~100% of undisbursed commitments into trapped, yield-dragging collateral, deepening the exact defect investor rates CRITICAL. One side's mandatory minimum is the other side's kill-deepener; the panel should say explicitly that homeowner escrow is non-negotiable and the fund economics must absorb it (my view), or the product dies here.

DC-8. Ops 5.1 / investor O2 vs. actuarial 2.1 / my §1.6 — Equity Shield's classification tag. AMBIGUOUS (ops, investor) vs. LIKELY-unlawful-in-every-characterization (actuarial, me). Incompatible launch postures follow from the tags (risk-memo-and-proceed vs. withdraw). Per my Dispute 6, the actuarial/my framing is the correct one: the ambiguity is only about which regime applies.

DC-9. Red-team F2/M4 vs. estate §§4.2/7.2 — unconditional vs. conditional discharge outcome. Red-team treats the discharge wipeout as settled and industrializable; the estate report's own §4.2 (conveyed-interest fork) and §7.2 (strong-arm/constructive-notice contingency: STATE-DEPENDENT and AMBIGUOUS) make it doubly conditional. Red-team built a HIGH-likelihood correlated-loss scenario on premises its own source document tagged as open. Per my Disputes 1–2.

DC-10. Retail P4 vs. red-team M5 / estate 6.2 — does the class waiver actually close the courthouse? Retail: the class waiver "removes the only economically viable remedy" (and tags the waiver SETTLED-enforceable). Red-team M5: McGill makes public-injunctive-relief claims non-waivable in California — the largest target market — and estate 6.2 notes AGs are never bound. Incompatible because retail's remedy-foreclosure premise fails in exactly the jurisdiction where the deception catalog (my §6.A) supplies a ready-made public-injunction predicate. The waiver is load-bearing only where it is weakest — and per my Dispute 5, the §38 clause has an independent AAA-administration defect no reviewer had flagged.

DC-11. Title §§3.1–3.2 vs. ops 3.3 — is the subordination promise fixable by drafting? Title: the subordination machinery (binding covenant, servicer POA, definiteness limits) is "feasible and mandatory" — a drafting build. Ops: mainstream refi lenders and the GSEs will simply condition the refi on payoff because no GSE-recognized subordination framework for HEI-type instruments exists, and none can be willed into existence by the sponsor's own covenant (LIKELY INTERPRETATION). Incompatible timelines: title treats it as pre-launch drafting; ops treats it as a multi-year market-adoption dependency outside the sponsor's control. Consequence for my domain: until resolved, the "never traps you in your mortgage" promise (§13) cannot be made in consumer marketing at all — the UDAAP exposure of promising a subordination the closing table won't honor is mine, whichever reviewer is right about the machinery.


BOTTOM LINE FOR STAGE 3

My Stage 1 verdict stands: fix-before-launch for the core contract, with embedded kills for the current marketing pages and for Equity Shield as structured. Nothing in the other seven reports weakens it; Olson, the Roberts amicus, and the tax report's SAM analysis strengthen the recharacterization half. The cross-examination adds three things the panel should carry forward: (1) the discharge-playbook and "payday-priced" narratives are overstated and should not drive the aggregate verdict — the accurate findings are damaging enough without them; (2) the two flagship fixes (record the lien; hold-to-maturity fund) each trade a solved problem in one domain for a new, priced problem in mine — TILA/§1639c(e)/licensing/§1641 on one side, ICA/Advisers Act/ERISA on the other — and the panel owes the sponsor a single reconciled fix-set, not eight incompatible ones; and (3) the characterization fork (DC-1) is the only decision that matters: until the sponsor picks one characterization and pays its full price in every domain simultaneously, every other fix is drafting on sand.

End of Stage 2 cross-examination. Findings tagged LIKELY INTERPRETATION or below require confirming opinions of licensed counsel in the relevant jurisdictions before reliance.