Data room / Run 1, stage 2: the nine cross-examinations
Real estate and title
run-1/reviews/crossexam-real-estate-title.md Open the file as written
Stage 2 Cross-Examination — Real Estate / Title Attorney
Reviewer role: Senior real-estate and title attorney (50-state recording, lien priority, mortgage law, foreclosure, homestead, due-on-sale)
My Stage 1 report: reviews/real-estate-title.md (verdict: FIX-BEFORE-LAUNCH; kill-equivalent in Texas; core prescription: "pick the lien")
Reports cross-examined: retail-customer.md, investor.md, tax-attorney.md, estate-bankruptcy.md, securities-consumer-finance.md, product-operator.md, insurance-actuarial.md, red-team.md
Spec: disclosures/equity-flow-invention-disclosure.txt (all 39 sections)
Date: 2026-08-01
Confidence tags: SETTLED LAW / LIKELY INTERPRETATION / STATE-DEPENDENT / AMBIGUOUS / SPECULATIVE
Method note: I dispute only where I have a genuine title-side disagreement on law, severity, or spec-reading. Where another reviewer's title claim is correct, I say so and move on. Several of my "disputes" are calibration disputes — the finding survives, but the tag or the severity does not.
PART I — DISPUTED FINDINGS
Dispute 1 — Estate-Bankruptcy §7.2: the 544(a)(3) strong-arm attack on the memorandum is real but overstated as the "highest-severity open title question in the document"
What they said: A "Memorandum of Equity Flow Agreement" evidencing an interest the drafter insists is not a conveyance, not a mortgage, and not a lien "may be OUTSIDE the recording act in some states — a recorded-but-unrecordable instrument imparts NO constructive notice in a number of jurisdictions. If so, the trustee avoids the interest outright under 544(a)(3)... total loss." Tagged STATE-DEPENDENT/AMBIGUOUS, highest severity.
Why I dispute the severity (not the mechanism):
- Modern recording statutes are capacious. The typical formulation admits any instrument "affecting the title to or possession of real property" (e.g., Cal. Gov. Code §27280) or "conveying or affecting an interest in real estate." A notarized instrument declaring that a named party claims an interest in the property's future appreciation that must be satisfied at any sale, transfer, or cash-out refinance (§34(b)) affects title on its face. Short-form memoranda of unrecorded agreements (leases, options, purchase contracts, rights of first refusal) are recorded and given effect routinely. [LIKELY INTERPRETATION that the memorandum is within the recording act in the large majority of states]
- The "recorded-but-unrecordable imparts no notice" doctrine is real but has mostly been applied to instruments failing formal requisites (defective acknowledgment, missing signatures), not to substantively novel interests. And under 11 U.S.C. §544(a)(3), the trustee's hypothetical-BFP status is measured by state law, which in most states charges a purchaser with inquiry notice from any instrument actually appearing in the chain of title. A title search will find this memorandum; a purchaser who finds it and ignores it is not a BFP in most jurisdictions. [SETTLED LAW as to the inquiry-notice framework; STATE-DEPENDENT as to the minority strict-classification rule]
- The genuinely exposed flank is narrower than the estate reviewer frames it: (a) the minority of states with restrictive instrument-class statutes, and (b) the sufficiency-of-description attack invited by §34's deliberate omission of all financial terms — which the estate reviewer correctly identifies as a secondary risk, and which I flagged in Stage 1 (§1.1). That is a state-by-state opinion item, not a portfolio-wide "total loss" headline.
What survives: The fix is identical either way (record a security instrument; obtain 50-state recordability opinions), so this dispute changes priority-ordering of risks, not the work plan. But diligence readers should not treat 544(a)(3) avoidance as the modal bankruptcy outcome; the modal fight is characterization (their §4.1–4.3), where the estate reviewer is on much firmer ground.
Dispute 2 — Red Team F2/M4: the "industrialized bankruptcy-discharge playbook" is priced as near-certain when it requires winning a fight the rest of the panel says the homeowner loses
What they said: F2/M4: enroll, collect payments, file Chapter 7 in a strong homestead state; the claim is discharged (Johnson v. Home State Bank), the memorandum is avoided (estate 7.2), "I keep the house and 100% of future appreciation." Tagged "SETTLED LAW away from industrialization."
Why I dispute the certainty: Johnson and §101(5) settle only that the investor holds a dischargeable claim — that half is right and I adopt it. But the debtor keeps the appreciation only if the in rem interest also dies, and that requires the debtor to win two additional contested points: (i) defeating recharacterization of the recorded interest as a consensual equitable mortgage (Cal. Civ. Code §2924; Peugh v. Davis, 96 U.S. 332 (1877)) — the exact recharacterization that tax (D-1), estate (4.3), securities (3.1), and I (Stage 1 §1.2) all rated LIKELY; a lien, once found, rides through discharge (Long v. Bullard; Dewsnup; Caulkett); and (ii) prevailing on the strong-arm/notice attack (Dispute 1 — itself only STATE-DEPENDENT). Recharacterization doctrine is symmetric: the investor will plead the mortgage label in bankruptcy as eagerly as it flees the label before the CFPB, and Butner lets it. The red team's own Section 1 premise elsewhere (M4 "SETTLED LAW away") silently assumes courts take §34 at face value — the opposite of the panel's base case.
What survives: The exposure is real, correlated, and adversary-driven; the drafting (§34's disclaimer) is a gift to debtor's counsel; the first published loss does industrialize. But the honest tag is LIKELY-contested litigation with material loss probability, not settled-law wipeout. This matters for pricing: a 30–50% per-case loss probability across a stress cohort is survivable with reserves; a 100% certainty is not. [LIKELY INTERPRETATION] See also Direct Conflict DC-2.
Dispute 3 — Investor D1 (echoed in A2): "senior foreclosure extinguishes the interest — SETTLED LAW for junior interests" is overstated as uniform
What they said: D1: "Recovery in senior foreclosure: interest extinguished (Event 4 — SETTLED LAW for junior interests); recovery limited to surplus-proceeds claims... Expected recovery 0–50%, mode near 0."
Why I dispute the absolutism: Extinguishment is settled only for a junior interest that is properly joined (judicial-foreclosure states) or properly noticed (non-judicial states). In judicial states, an unjoined junior holder of record is a necessary party whose interest survives the sale (SETTLED LAW); the buyer takes subject to it and must re-foreclose or pay it off. For a novel memorandum that foreclosure counsel and title searchers may not classify correctly, non-joinder is a realistic scenario — and it is investor-favorable, a partial mitigant nobody on the panel credited. In non-judicial states, whether the statutory notice list ("holders of recorded junior liens or interests," e.g., Cal. Civ. Code §2924b) automatically reaches a self-described non-lien memorandum is exactly the wording-dependent question from my Stage 1 §4.1, mitigable for $30 by recording a statutory request for notice. Separately, several judicial states layer statutory redemption periods that further complicate a flat "extinguished, recovery ≈ 0."
What survives: The economic conclusion (senior foreclosure is the worst path; surplus recoveries are thin; surplus standing is clouded by the not-a-lien label — my §4.2) stands. Only the "SETTLED/uniform" framing falls. [SETTLED LAW as to joinder/notice prerequisites; STATE-DEPENDENT as to this instrument]
Dispute 4 — Insurance-Actuarial Finding 1.1 (and the verdict's "escrow-first 1.0x recovery is genuinely robust"): the stress model assumes an escrow priority that will not exist at a real closing
What they said: "The base structure already returns 1.0x from escrow at exit (Section 29, Priority 3), senior even to commissions... Distressed sales almost never breach the floor." The verdict repeats: "its escrow-first 1.0x recovery is genuinely robust in sale exits even at −45%."
Why I dispute the premise: The actuarial model takes §29 Flow 4 at face value. As I found in Stage 1 §2.2 (and as investor A7 and estate 8.2 independently found), the waterfall's placement of investor claims ahead of transfer taxes, title/escrow fees, and broker commissions is unenforceable against the third parties it purports to subordinate: the deed does not record without transfer tax, the broker is paid at closing under the listing agreement, and in bankruptcy those items are administrative or statutory. In every thin-equity sale, 6–9% of gross price comes off the top before the Equity Flow claim sees a dollar. The actuarial finding that a −45% sand-state replay "still nets ~$158K against a $126K capital claim" therefore overstates the cushion by the full costs-of-sale wedge on precisely the marginal deals; the correct modeling order is senior lien → taxes → costs of sale → EF claim. The direction of the error is uniform (optimistic), and it feeds their conclusion that "floor loss ≈ foreclosure rate only." [SETTLED LAW as to closing practice; the modeling implication follows arithmetically]
What survives: Their central Equity Shield findings (underfunded tail, illusory bear-case protection, unlicensed guarantee) are untouched by this and I do not contest them. The fix is the same waterfall reorder I prescribed; the actuarial team should re-run Finding 1.1 with costs-of-sale senior.
Dispute 5 — Retail-Customer Tag Index #4 and Product-Operator 5.1: "title recording / satisfaction-at-closing mechanics — SETTLED LAW (mechanics only)" concedes too much to the spec
What they said: Retail (T12, tag index #4) and ops (5.1, first bullet) both tag §12's escrow-clearance mechanism as SETTLED LAW "as to mechanics," reserving doubt only for the nature of the interest.
Why I dispute the tag: The "mechanics" that are settled are settled for liens with liquidated payoff demands. What §12 actually claims — that the title company will calculate an unliquidated, formula-driven payoff (ratchet years, three-appraisal median, improvement-receipt offsets) and administer it — is not settled anywhere; it is not even current practice for incumbent HEIs, which run servicer-issued binding payoff demands. Without a payoff-demand/estoppel obligation (absent from the spec — my Stage 1 §2.1), the settled escrow machinery has nothing to execute, and the realistic outcomes are exceptions, holds, and interpleader. Ops itself makes this point one clause later ("title companies underwrite off familiar instruments"), so this is a tag-discipline dispute: the correct tag for §12 as written is SPECULATIVE, with the underlying escrow mechanism SETTLED for a different instrument than this one. The distinction matters because "settled mechanics" is the single spec claim most likely to be quoted back at diligence.
Dispute 6 — Estate-Bankruptcy §1.4: "no statutory right in any probate code... to compel sale" is modestly overstated
What they said: Event 3's investor petition for a probate-court sale order is "a remedy the courts do not offer"; probate sales are directed on petition of the personal representative "or, in limited circumstances, secured lienholders."
Why I refine rather than fully adopt: In many states a creditor holding a timely-presented, allowed claim has real machinery: petition for appointment of a personal representative or public administrator where heirs sit idle, petition to compel administration, and — where the estate lacks liquid assets — the PR's duty to sell realty to pay allowed claims, enforceable by court order (see generally UPC art. III; Cal. Prob. Code claims-and-sales framework). So the remedy is not fictional for a timely secured or allowed claimant; it is fictional for (a) a claimant barred by the nonclaim statute (their §1.1 — correct, and the operational miss risk at Tier 2 scale is my Stage 1 §4.3), and (b) a non-lien "equitable interest" holder with no allowed claim, which is what §34 drafts. And in Florida the homestead-descent rule (Snyder v. Davis) moots the machinery entirely for the non-lien version. [STATE-DEPENDENT; the estate reviewer's bottom line survives for the instrument as drafted]
Dispute 7 — Red Team C5: "the first underwriting bulletin declining the instrument is a de facto national injunction" overstates title-industry mechanics
What they said: A competitor circulates Olson and the Roberts amicus to title-underwriter counsel; "the first underwriting bulletin declining the instrument is a de facto national injunction."
Why I dispute the magnitude: Title underwriting bulletins are issued per underwriter family, not industry-wide; there are four national families plus meaningful regional underwriters, they compete for closing volume, and no ALTA-level body issues binding refusals. One family declining is a serious impairment (the top two families write a majority of the market), not an injunction; incumbent HEIs obtained workable underwriting treatment despite similar novelty. The realistic competitor-block scenario is friction and delay concentrated in the refi-subordination path (where I already found near-certain refusal absent executed subordinations — Stage 1 §3.1), not a closed sale channel: at sale, insurers happily list the memorandum as a Schedule B exception and demand satisfaction, which is the enforcement the product wants. [LIKELY INTERPRETATION] The GSE-servicing-bulletin half of C5 (due-on-sale agitation) is the more dangerous half and I do not dispute it — it matches my Stage 1 §6.2 caveat 2.
Dispute 8 — Securities/Consumer-Finance Verdict ¶1: "incumbents operate today at scale under materially similar characterization risk" understates how different this chassis is, and the gating list omits the largest title exclusions
What they said: Why-not-kill reasoning: "Lump-sum HEIs (Hometap, Point, Unison) operate today at scale under materially similar characterization risk... recorded-encumbrance enforcement is conventional." Gating items list state exclusions "CT, MD, WA and other HEI-statute states."
Why I dispute: Two title-side corrections. First, the incumbents' enforcement is conventional because they record real security instruments (deeds of trust / mortgages securing performance) — the one thing §34 refuses to do. The risk is not "materially similar"; this spec discards the incumbent chassis and keeps only the incumbent talking points, which is why five of my six checklist domains break (Stage 1, central-defect section). Second, any launch-exclusion list that names CT/MD/WA but not Texas is incomplete as a matter of settled state constitutional law: Tex. Const. art. XVI §50(a) makes the interest void against homestead regardless of consumer-finance strategy, and Florida requires restructure-or-exclude for the death exit (Fla. Const. art. X §4(b); Snyder v. Davis). The securities reviewer's fix list inherits the spec's silence on homestead geography. [SETTLED LAW as to TX; LIKELY as to the FL death-exit failure]
Concessions and updated positions (for the record)
- Estate §7.1 (522(f) unavailable against consensual real-property interests): correct and adopted; it resolves the question I cross-referred as AMBIGUOUS in Stage 1 §5.4. §522(f)(1)(B) reaches only enumerated personalty; the residual risk is classification (judicial-lien argument if a court equitably imposes the lien), which is small. [SETTLED LAW, adopting their analysis]
- Estate §1.6 (Harms v. Sprague joint-tenancy extinguishment) and §3.1–3.2 (joinder/Droeger): correct; strengthens my Stage 1 §1.5. On Texas I would upgrade their "LIKELY void" to SETTLED for the instrument as drafted.
- Ops 3.3 (refi lenders will demand payoff, not subordination): adopted with one refinement — GSE guides do not flatly require payoff of subordinate financing at refi; they permit resubordination of eligible subordinate financing, and the failure point is that an appreciation-share instrument with a ratchet and no stated maximum will not satisfy subordinate-financing eligibility, so payoff becomes the practical condition. Same outcome, cleaner mechanism. [LIKELY INTERPRETATION]
- Ops 6.2 (marketable-title decay tagged "unresearched"): researched in my Stage 1 §1.4 — MRTA extinguishment is SETTLED in the enacting states; their operational fix (re-recording calendar, servicer-of-record) is the right one.
- Tax S-2 (FL doc stamps render unstamped instrument unenforceable until tax paid): correct (Fla. Stat. §201.08) and corroborates my Stage 1 §1.3; note for completeness the bar is curable by payment with penalties, so it is a cost-and-delay defect, not a permanent voiding.
PART II — PROPOSED FIXES THAT CREATE NEW PROBLEMS IN MY DOMAIN
FC-1 — Tax attorney's implied fix (accept the SAM/loan characterization; their "minimum redesign") does not open Texas, and it triggers the mortgage-tax regime everywhere else
The tax review concludes the most natural characterization is a shared appreciation mortgage (Rev. Rul. 83-51) and that the least-broken path embraces loan mechanics (lump sum, genuine downside sharing). On title, embracing that means the recorded instrument must be a mortgage/deed of trust securing a nonrecourse obligation with contingent appreciation-based interest. Consequences:
- Texas stays closed. A voluntary lien on homestead must fit Tex. Const. art. XVI §50(a). As a loan, the product must comply with §50(a)(6) (closed-end home equity) or §50(t) (HELOC). The monthly-disbursement structure is an open-end/advance structure, so §50(t) governs — and §50(t)(6) requires each advance to be at least $4,000; a $1,500/mo draw is constitutionally noncompliant on its face. The 2–3% origination fee also collides with the §50(a)(6)(E) 2% fee cap, and the 80% LTV aggregate cap and one-equity-loan rule bind. A noncompliant homestead lien is void, with no limitations bar to quieting title (Wood v. HSBC Bank USA, 505 S.W.3d 542 (Tex. 2016)). So the tax fix converts "unavailable in Texas as equity" into "unavailable in Texas as debt." [SETTLED LAW as to the constitutional text; LIKELY INTERPRETATION as to §50(t) capture of this structure]
- Mortgage recording taxes become unambiguous. NY Tax Law art. 11 (tax measured by the amount secured "or which under any contingency may be secured" — an uncapped ratcheting appreciation claim forces a stated maximum-amount clause or risks tax on an unbounded base), FL doc stamps + nonrecurring intangibles, MN/AL/GA/TN/OK/VA regimes. The $200–$500 recording budget (§29 Flow 2) is off by an order of magnitude in these states. [STATE-DEPENDENT; SETTLED as to the existence of the regimes]
- The compensating title benefits are real and should be said aloud: Garn-St Germain §1701j-3(d)(1) squarely shelters a recorded subordinate lien (my Stage 1 §6.2 caveat 1 dissolves); surplus, notice, cure/reinstatement, §6323 federal-tax-lien priority, and FL homestead enforceability (with both-spouse joinder, Art. X §4(c)) all snap to settled law. The tax fix and my fix converge on the same instrument — the price is TILA/licensing (their domain) and items 1–2 above (mine).
FC-2 — The estate/bankruptcy "record a true security instrument" fix (which I share) must be priced honestly on the title side
Since I prescribed the same fix, this is a candid costing, not an objection: (a) recording/doc-stamp taxes per FC-1.2; (b) an indeterminate secured amount forces a stated maximum claim in several states — a drafting decision with consumer-disclosure consequences the retail reviewer will need to reconcile with "no balance" marketing; (c) a deed of trust needs a named beneficiary and trustee, and if combined with ops 3.4/6.2's "MERS-analogue registry / permanent servicer facade," the nominee-beneficiary structure walks into the MERS case law — e.g., Bain v. Metropolitan Mortgage Group, 175 Wn.2d 83 (Wash. 2012) (nominee that never holds the obligation cannot be deed-of-trust beneficiary) — so the registry must hold by assignment of record, not nominee status, in the strict states [STATE-DEPENDENT]; (d) a power-of-sale instrument contradicts the "no foreclosure right" consumer promise, and a deliberately judicial-only performance deed of trust gives back speed; someone must choose; (e) if consumer-finance recharacterization later lands anyway, TILA §1635 rescission voids the security instrument itself (securities 3.2) — i.e., the fix concentrates all enforcement into an instrument that the three-year rescission tail can dissolve. None of this changes my verdict that the lien is the right call; all of it belongs in the price.
FC-3 — Consumer-side fixes of "termination with forfeiture on investor default" / estate fix 6's "springing release": an off-record termination is invisible to title and unadministrable at closing
Event 10 already says the encumbrance is "modified (not released)" upon forfeiture. Estate fix 6 proposes escrowed/LC-backed commitments with a springing release on fund default; the retail review demands the forfeiture remedy be real. On title, a right that terminates or shrinks by off-record events (three missed ACHs; a fund's insolvency) cannot be seen by a title searcher, so: (a) the memorandum continues to cloud title after the homeowner has "terminated," and clearing it requires investor cooperation the defaulted/insolvent investor will not give — reproducing exactly the stale-encumbrance problem (§34 release-failure; estate 6.2; ops 6.2) the fix was meant to solve; (b) title underwriters will no more insure over self-executing release language than over self-executing subordination language (my Stage 1 §3.1) — symmetric defect; (c) partial forfeiture (capital-return claim survives) means a release is never the right instrument; an amendment of record is, which doubles the cooperation problem. Workable machinery exists but must be built: pre-executed releases and amendments held by an independent escrow agent with objective delivery triggers, a recordable homeowner affidavit-of-default protocol where state law permits, and a statutory-style court-petition backstop with fee-shifting (§34 already has the petition; it presumes a findable respondent). Absent that build, "termination on investor default" is a contract right that dies at the recorder's office. [LIKELY INTERPRETATION as to underwriting practice; the drafting gap is on the face of the documents]
FC-4 — Red team M7's privacy fix (minimum-content memoranda) directly undermines the constructive-notice and recordability fixes
M7 proposes "minimum-content memoranda where statute permits" to stop the recorded instrument functioning as a scam-target list. But the memorandum is already content-light (§34(b) omits all financial terms), and the estate reviewer's §7.2 attack and my §1.1 finding both run through insufficient content: several states condition recording of a memorandum on statutory minimum contents, inquiry-notice sufficiency depends on what the instrument discloses, and (post-FC-1/FC-2) mortgage-tax states require a stated maximum secured amount — the opposite of content minimization. You cannot simultaneously optimize the recorded instrument for privacy and for notice/priority; the resolution is a fuller memorandum plus the operational protections (servicer-of-record masking of the beneficial holder, payment-change authentication), not a thinner one. [STATE-DEPENDENT as to memorandum-content statutes]
FC-5 — Actuarial fix (ii) (re-underwrite as hold-to-organic-exit, 25–30-year evergreen) leans the whole fund on the title feature most likely to lapse
Honest duration means the base case now holds recorded interests through years 25–35 — squarely into Marketable Record Title Act extinguishment windows (Fla. ch. 712 — 30 years; MI/OH/MN at 40; a dozen-plus states) and lien-expiration statutes (Cal. Civ. Code §§882.020–.030; Mass. obsolete-mortgage statute), none of which the spec addresses (my Stage 1 §1.4). The fix is only viable with a funded, calendared re-recording/preservation program and a solvent, findable holder in year 29 — which collides with the same reviewers' observation that the originator dissolves at year 7–8 and servicing is priced at zero (ops 1.3, 6.1–6.2). Their fix (iv) alternative — "a small fixed-dollar accrual senior in the waterfall" — they concede reintroduces debt character; add from my side that an accruing balance also (a) re-opens the growing-balance/reverse-mortgage optics (Olson), (b) inflates the "amount secured under any contingency" for mortgage-tax states, and (c) collides with emerging state HEI statutes banning escalating shares (Maine-2026 pattern). [SETTLED LAW as to MRTA/lien-expiration operation in enacting states]
FC-6 — Investor verdict's "settlement-dated instrument (defined maturity)" starts lien-expiration clocks the ride-along then ignores
A stated maturity is the investor reviewer's fix for the no-maturity problem. On record title, a stated maturity starts the ancient-lien clock: e.g., Cal. Civ. Code §882.020 bars enforcement 10 years after recorded maturity absent a recorded extension. A holder that states a year-10 maturity and then forbears (the ride-along behavior, or its negotiated successor) walks the interest into statutory expiration unless extensions are recorded on calendar — one more entry for the preservation regime, and a trap for a passive Tier 2 holder with no servicer. Fixable, but only if designed in with FC-5's machinery. [SETTLED LAW in the enacting states]
FC-7 — Estate §2.4's implied fix (homeowner-elective payment suspension for Medicaid protection) worsens the tax defect and the recorded-claim mismatch
Estate 2.4(a) implies the contract needs a homeowner right to decline or suspend disbursements. Tax P-5 identifies variability of total consideration as a core recharacterization defect ("fixed committed amount" vs. Flow 3a). An elective-suspension fix makes consideration more path-dependent and homeowner-steered, strengthening the lender-credit-line reading — and on my side, it widens the gap between the recorded claim basis and cash actually advanced, which must be tracked off-record and certified at payoff (the dual-ledger problem ops 1.2 flagged, now with a third input). If Medicaid protection is required, the cleaner instruments are term-restructuring at origination or a hardship-deferral held at the investor's obligation level with recorded-basis certification — designed jointly with tax counsel, not bolted on. [AMBIGUOUS on the Medicaid side per the estate reviewer; the tax-collision is on the face of P-5]
(Noted in passing, sub-major: the retail fix demanding APR-equivalent disclosure is right for consumers but is admission-adjacent evidence in any credit-recharacterization fight — coordinate the disclosure language with counsel; and the securities fix list's "reserved TILA-compliant fallback disclosure stack" should include a fallback recorded instrument, because Reg Z compliance with a non-lien memorandum is incoherent.)
DIRECT CONFLICTS
Every direct conflict I see between any two reviewers, including those involving me.
DC-1 — The lien fork (the panel's central conflict). Real-estate-title (Stage 1 fix #1) and estate-bankruptcy (fix #1) say: record a true security instrument; every enforcement path (discharge survival, homestead, surplus, §6323, Garn (d)(1)) depends on lien status. Tax-attorney (D-1, D-4, verdict) and securities/consumer-finance (3.1–3.2) say: debt/security-instrument characterization detonates the tax posture, TILA/licensing exemption, usury shield, and the arbitration clause; the estate review's own kill-condition concedes it. Incompatible because the same recorded instrument cannot be a security device for enforcement and not-a-security-device for tax/regulatory purposes; one domain's mandatory fix is another domain's kill condition. (Red team verdict ¶1 flagged this; I confirm it from the title side and note the only coherent resolution: take the lien, fight "not a loan" on economics and disclosures — accepting that the fight now runs through consumer-finance law, not property law.)
DC-2 — Discharge playbook vs. recharacterization base case. Red-team F2/M4 says the homeowner discharges the claim and keeps house plus appreciation because the interest is not a lien (near-certain, "SETTLED LAW away"). Tax D-1/D-2, estate 4.3, title 1.2, securities 3.1 say the LIKELY judicial outcome is recharacterization as a mortgage/equitable lien — which would ride through discharge. Incompatible because both cannot be the modal outcome of the same litigation; each report priced the branch worst for the product. (Both branches are individually possible across a portfolio — different courts, different states — but the panel should not sum worst cases as if each were certain.)
DC-3 — Escrow floor robustness vs. junior-residual reality. Insurance-actuarial 1.1/verdict says the 1.0x capital return is "genuinely robust in sale exits even at −45%," senior even to commissions. Investor A2/A7, estate 8.2, and title 2.2 say the capital return is a junior non-recourse residual whose contractual seniority over costs of sale, commissions, and transfer taxes is unenforceable against third parties (and inverted in bankruptcy). Incompatible because the actuarial stress arithmetic assumes a waterfall priority three other reviewers (and closing practice) reject; the floor-breach frequency is understated by the costs-of-sale wedge.
DC-4 — Honest disclosure: survivable or fatal? Securities/consumer-finance verdict says the marketing is a kill as written but the product survives once both pages are rewritten from the disclosure. Retail-customer verdict ¶4 says a page that honestly disclosed the pause trigger, ratchet, fees, covenants, and APR-equivalents collapses the rational customer base, so redescription cannot save it. Incompatible because sec-cf's fix presupposes a viable honestly-marketed product and retail's independent cost model (V1–V4) says the honestly-marketed product is dominated by a HELOC for everyone who can get one.
DC-5 — Death: cleanest path or broken exit? Tax H-5 calls death "the one clean path" (basis step-up; claim deductible under §2053 if enforceable). Estate 1.1–1.5 and title 4.3/5.2 find the death exit is where the interest is most likely to be barred (nonclaim statutes), unenforceable (FL homestead descent — Snyder v. Davis), or stranded (trust-held property; undefined mid-term death economics). Not a contradiction of doctrine — the reviewers answer different questions — but the conclusions conflict at the product level, and tax H-5's §2053 deduction presupposes exactly the enforceability estate/title deny: the death exit cannot be simultaneously "clean" and uncollectible.
DC-6 — Does the rate-and-term carve-out work at all? Retail C5/T5 analyzes the harm as asymmetric lock-in: rate-and-term refis are permitted and functional while cash-out is blocked. Ops 3.3 and title 3.1–3.2 say even the permitted rate-and-term refi fails at the closing table (lender/insurer demands executed subordination or payoff; no machinery exists). Incompatible descriptions of the same closing: retail models a homeowner who can refinance down-rate; ops/title model one who cannot. (The truth is ops/title's version, which makes the consumer harm worse than retail's model.)
DC-7 — Foreclosure extinguishment certainty. Investor D1 says extinguishment of the junior interest in senior foreclosure is SETTLED, recovery ≈ 0. Title 4.1 (mine) says extinguishment requires joinder/notice; an unjoined recorded junior survives judicial foreclosure, and non-judicial notice capture is wording-dependent. Incompatible at the tag level (SETTLED vs. STATE-DEPENDENT); materially changes recovery modeling in judicial-foreclosure states.
DC-8 — Memorandum content: thinner or thicker? Red-team M7 prescribes minimum-content memoranda (privacy/anti-targeting). Estate 7.2 (and title 1.1) treat the memorandum's thin content as a live avoidance/notice defect, and mortgage-tax states force a stated maximum amount if lien-ized. Incompatible prescriptions for the same instrument; must be resolved toward fuller content plus operational masking (see FC-4).
DC-9 — Binding payments as protection vs. payments as Medicaid poison. Investor C2 and ops 1.5/5.2 treat the binding disbursement obligation (with three-miss forfeiture) as the homeowner's core protection and a fund-side risk to be engineered around; estate 2.4(a) treats the inability to decline those same mandatory payments as a defect (Medicaid disqualification) implying an elective-suspension fix; tax P-5 says any additional payment variability deepens the recharacterization defect. Three-way incompatibility in the design objective for the payment covenant (rigid vs. suspendable vs. fixed-total).
DC-10 — Core-product viability verdicts on the same record. Insurance-actuarial ("the core Equity Flow contract does not, from this domain's vantage, require killing"; escrow recovery robust; not insurance) and securities/consumer-finance ("the product concept can survive diligence") vs. red team ("kill — the fix-before-launch verdicts each assumed their domain's repairs could be made without detonating another domain, and they cannot"), with tax/retail/investor also at kill. Incompatible because the fix-before-launch verdicts (including mine) are each conditioned on cross-domain fixes that DC-1, DC-3, and DC-4 show are mutually exclusive or unpriced; the red team's meta-claim is a direct challenge to the conditional structure of the four fix verdicts, mine included.
NET POSITION AFTER CROSS-EXAMINATION
My Stage 1 verdict stands: FIX-BEFORE-LAUNCH, kill-equivalent in Texas, with the lien decision as the load-bearing choice. Cross-examination sharpens it in three ways. First, the panel's worst cases are being double-counted: the discharge-wipeout scenario (red team) and the mortgage-recharacterization scenario (tax/securities) are opposite branches of one fork, and the strong-arm total-loss scenario (estate 7.2) is a minority-state tail, not the modal outcome — risk pricing should reflect the fork, not the sum. Second, the actuarial validation of the 1.0x escrow floor — the most product-favorable finding on the panel — rests on a waterfall priority that closing practice will not honor and should be re-run. Third, the fixes now on the table are not free in my domain: the tax fix does not open Texas (§50(t)(6) kills monthly advances as surely as §50(a) kills the equity version), the lien fix imports recording taxes and a rescission-voidable instrument, the honest-duration fix walks into marketable-title extinguishment without a preservation machine, and the consumer termination fix is unenforceable at the recorder's office as currently imagined. Whoever runs Stage 3 should force a single characterization decision first (DC-1) and re-price every other reviewer's findings on the chosen branch only.
— Real Estate / Title Attorney, Stage 2 cross-examination