# EQUITY FLOW v3 — INVESTOR / CREDIT DILIGENCE REVIEW Reviewer: senior credit + venture/PE lens (adversarial). Date: 2026-08-01. Material reviewed: concepts/EQUITY-FLOW-V3-CONCEPT.md (v0.1, full). Prior positions re-adopted in full: reviews/investor.md (KILL) and reviews/crossexam-investor.md (KILL confirmed on cross-exam). v3 is evaluated fresh on its own mechanics but every legal/tax finding from v1 that v3's own concept sheet concedes is unresolved (Section 6, items 1-5) is treated as open, not solved, until the specialist reviewers close it. Framing per mandate: assume the product has already failed; explain why. Method: all pricing below rebuilt independently in Python (deterministic grid, parameter sweep, and a stylized Monte Carlo). Script logic is described inline; figures are this session's own output, not carried over from v1. --- ## 0. WHAT CHANGED AND WHY IT MATTERS TO THIS LENS v3 kills the v1 findings that were about VALUATION PROCESS (appraisal collusion, exit manipulation, renovation-receipt fraud, pause-AVM disputes, occupancy covenants) by moving settlement onto a third-party metro index. That is a real and correctly-targeted fix for those specific findings. It does nothing to fix the two findings that actually killed v1: (1) the return engine's relationship to cost of capital, and (2) the pricing identity (does risk reallocation create real surplus, or just move the zero-sum). v3 explicitly reopens both as its own Section 6 item 6 ("did risk reallocation genuinely create surplus, or did we just move the zero-sum around?"). This review answers that question with arithmetic. Short answer: no, and the parameter sweep proves it has no answer within any tested region. --- ## 1. PRICING — REBUILD Deal: V0 = $700,000; Advance = 10% = $70,000; Share = 35%; Cap = 3.0x Advance = $210,000; Floor = $0. Notional leverage (Share x V0 / Advance) = 3.5x, exactly as the concept sheet states. Payoff = min( max( Advance + Share x V0 x (HPI_exit/HPI_0 - 1), 0 ), Cap ), further capped at actual net sale proceeds (non-recourse) — the proceeds cap is a second, separate ceiling on recovery that is not modeled below except where flagged; where it binds it can only make investor outcomes worse than shown. ### 1a. Gross investor IRR grid (%, deal-level, constant annual HPI growth g) | exit yr | g=-2% | g=-1% | g=0% | g=1% | g=2% | g=3% | g=4% | g=5% | g=6% | |---|---|---|---|---|---|---|---|---|---| | 3 | -7.4 | -3.6 | 0.0 | 3.4 | 6.7 | 9.8 | 12.8 | 15.8 | 18.6 | | 5 | -7.9 | -3.7 | 0.0 | 3.3 | 6.4 | 9.3 | 11.9 | 14.5 | 16.9 | | 7 | -8.5 | -3.8 | 0.0 | 3.3 | 6.2 | 8.8 | 11.2 | 13.5 | 15.6 | | 10 | -9.7 | -4.0 | 0.0 | 3.2 | 5.9 | 8.2 | 10.4 | 11.6c | 11.6c | | 15 | -15.2| -4.4 | 0.0 | 3.0 | 5.4 | 7.5 | 7.6c | 7.6c | 7.6c | | 20 | FLOOR| -4.9 | 0.0 | 2.9 | 5.1 | 5.6c | 5.6c | 5.6c | 5.6c | | 25 | FLOOR| -5.8 | 0.0 | 2.8 | 4.5c | 4.5c | 4.5c | 4.5c | 4.5c | | 30 | FLOOR| -7.7 | 0.0 | 2.7 | 3.7c | 3.7c | 3.7c | 3.7c | 3.7c | (c = cap already bound at that cell; FLOOR = total loss of Advance.) Two structural facts fall straight out of the formula, both severity CRITICAL: **Cap-bind threshold is duration-independent in cumulative terms.** The cap binds the instant cumulative index appreciation from enrollment hits 57.1% (Advance/(Share x CapMult-1)... algebraically: (1+g)^t - 1 = (Cap-Advance)/(Share x V0) = 57.1%, always, at any t). At t=7 that needs 6.67%/yr; at t=15 only 3.06%/yr; at t=30 only 1.52%/yr. A "boring," trend-line 3-3.5%/yr metro over a 15-20 year hold — exactly the hold length the box's own lock-in/holding-propensity screen selects for — caps the deal. Once capped, IRR becomes a pure function of duration decay with a FIXED numerator: for a g=6%/yr path the IRR falls from 18.6% (yr 3) to 14.7% (yr 8, cap binds) to 7.6% (yr 15) to 3.7% (yr 30) — the exact "duration is not compensated" pathology this reviewer found in v1's ratchet, reproduced without a ratchet. Extension past cap-bind is pure economic loss to the investor with no offsetting mechanism (no ratchet, no re-share, no coupon). **Floor-bind threshold is ALSO duration-independent in cumulative terms.** Any exit where the index is down 28.6% cumulative from enrollment wipes the Advance to $0, whether that decline happens in one bad year or bleeds out over two decades in a stagnant metro. This is a single fixed number: Advance/(Share x V0) = 70,000/245,000 = 28.57%. ### 1b. The pricing identity — attacked directly Homeowner cost (dollars paid above the Advance, including a 3% origination fee) vs. two HELOC benchmarks: a capitalized/deferred HELOC at 8.5% (fair comparison — no monthly payment required, matching Equity Flow's own "no payments" pitch) and a cash-pay, interest-only HELOC at 8.5% (the realistic comparison — what most HELOC borrowers actually experience as a cash cost, and the one the concept sheet's own homeowner-facing page would be measured against). Base parameters (Adv 10% / Share 35% / Cap 3.0x), full grid of 8 exit years x 6 HPI paths (48 cells; representative sample shown, full grid computed): - g=0%: EF cost $2,100 (fee only) at every exit year — beats both HELOC benchmarks always. This is the "index flat" marketing cell. It is also the cell where the investor's IRR is 0.0% at every exit year — the fund earns nothing. - g=2%, t=7: EF cost $38,528 vs capitalized-HELOC $53,910 (beats) vs cash-HELOC $41,650 (beats, barely) — investor IRR 6.2%. - g=3%, t=7 (roughly trend HPA): EF cost $58,419 vs capitalized-HELOC $53,910 (LOSES) vs cash-HELOC $41,650 (LOSES) — investor IRR 8.8%, clears an 8% hurdle. - g=3%, t=10: EF cost $86,360 vs capitalized-HELOC $88,269 (beats, by $1,909 — 2.2%) — investor IRR 8.22%. This is the single most favorable cell found in the entire 48-cell grid against the generous (capitalized) HELOC comparison. - Long holds (t=20-30) at any g >= 2%: EF cost freezes at $142,100 (cap-bound) while a capitalized HELOC keeps compounding, so EF "wins" on paper — but the investor's IRR in those same cells has decayed to 3.7-5.6%, nowhere near a real-asset hurdle. The homeowner "wins" precisely because the investor has stopped earning anything. **Parameter frontier search.** Swept Advance% in {5,8,10,12,15}, Share% in {20,25,30,35,40,45,50}, Cap multiple in {2.0,2.5,3.0,3.5,4.0} — 175 parameter combinations, each tested against a 30-cell grid (exit years 7/10/13/15/20 x HPI growth 0%/2%/3%/3.5%/4%/5%), homeowner benchmarked against the capitalized HELOC (generous case) and against the cash-pay HELOC (realistic case), investor tested at 6% and 8% hurdles. Results, generous (capitalized-HELOC) comparison, investor >= 8% hurdle: best combination found anywhere in the 175-parameter sweep clears only 2 of 30 cells (6.7% of the grid). Base parameters clear 1 of 30. At a 6% investor hurdle the best combination found clears 13 of 30 (43%) — Adv 15% / Share 40% / Cap 3.5-4.0x — but 6% does not clear the stated 6-9% target's own low end with margin, and is arguably below what a levered, capped, tail- loss-bearing real-asset position should require (Section 2). Results, realistic (cash-pay HELOC) comparison, investor >= 8% hurdle: **zero out of 175 parameter combinations produce a single cell where both sides clear.** Not the base case — none of them, across the entire tested Advance/Share/Cap space. **VERDICT ON THE PRICING IDENTITY:** there is no genuine two-sided surplus. Against the comparison a homeowner would actually experience (paying HELOC interest in cash, which is how virtually every HELOC borrower services the loan), the frontier is empty across the full parameter space tested. Against the most generous plausible comparison (a homeowner who would otherwise let HELOC interest capitalize untouched for a decade-plus, which is not how HELOCs are underwritten or used), the frontier exists but is a hairline — one to two cells out of thirty, worth low single-digit-thousands of dollars of homeowner benefit, evaporating with any real-world friction (higher HELOC rate, higher origination fee, a 1-2% appraisal/AVM cost at enrollment, counseling/attorney fees). The index-settlement redesign relocated the zero-sum from "appraisal/renovation disputes vs. fund" to "HPA-path outcome vs. fund," but it is still a transfer, not a Pareto improvement, in every region of the parameter space this reviewer could find. Severity: CRITICAL — this was the one problem v3 was explicitly built to solve (concept sheet Section 6, item 6), and it is not solved. --- ## 2. THE BUYER The concept sheet's own framing: "pensions/sovereigns/real-asset allocators... 6-9% expected... a diversifier, NOT a venture-return story." **Is there a track record for this exact asset class?** Yes, and it is a instructive failure, not a null result. CME Group listed cash-settled Case-Shiller home price index futures and options in 2006 — an exchange-traded, fully liquid, dollar-transparent, no single-name-credit, no-optionality, no-recorded-instrument, professionally-marketed version of exactly the underlying exposure v3 proposes to sell (metro HPI beta). Robert Shiller himself was the intellectual sponsor and spent years advocating for institutional adoption as a hedging and diversification tool ("macro markets"). Volume and open interest never reached levels that supported a liquid institutional market; the contracts survive in name but the thesis that real-money allocators would embrace metro housing-index exposure as a diversifier has had a ~20-year public trial, in a MUCH easier-to-buy form (exchange-traded, standardized, instantly liquid, zero embedded homeowner optionality) than v3's private, illiquid, single-counterparty, prepayment-optioned, non-recourse, 30-year recorded contract. If professional allocators would not clear a liquid CME future on this exact index, the claim that they will underwrite a private, unrateable, single-name-credit- embedded version at scale is SPECULATIVE, not "credible." Tag: SPECULATIVE (no evidenced institutional demand for this exposure in any wrapper over two decades of trying). **What does the historical HPI record actually imply, unlevered?** National nominal HPA has averaged roughly 4-5%/yr over the last 60+ years; REAL (inflation-adjusted) appreciation has been much lower, plausibly 0-1.5%/yr over long spans (Shiller's own long-run US home price index shows near-zero real appreciation for most of the 20th century, with the 2000s-2020s a historically unusual period). Metro-level dispersion around that mean is large: 1990-96 California metros fell 20-27% peak-to-trough over ~5-6 years; 2006-2011 "sand states" (Phoenix, Las Vegas, Miami, the California Inland Empire) fell 45-60% peak-to-trough; even in the same national cycle some metros (Dallas, Denver) barely corrected. This is exactly the dispersion that makes "metro HPI, unlevered" a legitimate diversifier candidate in theory — and exactly why 3.5x leverage on it is dangerous in practice. **What does 3.5x leverage do to that record?** At 3.5x, the floor-bind threshold is a 28.6% cumulative decline (Section 1a) — well inside BOTH historical tail episodes above, and achievable even in a "merely disappointing," non-crisis decade in a stagnant metro (e.g., -1.9%/yr for 18 years = -29% cumulative; Rust Belt and some Sun Belt metros have put up worse over comparable spans without any single "crash" headline). This is not a 1-in-50-year tail for this product; on the historical metro record it is closer to a 1-in-8-to-15-metro-cycle event, correlated within every metro simultaneously. **Does the $0 floor make this venture-risk paper demanding venture-adjacent returns after all?** The stylized Monte Carlo below (Section 3) puts single-deal total-loss probability at roughly 1-4% over a 30-year horizon under an admittedly illustrative regime-switching calibration — that alone is not a venture-return argument (VC-style total-loss rates on individual investments run 30-50%+). But three features push this toward venture-like risk pricing regardless: (i) the loss is NOT diversifiable within a metro-concentrated vintage — a single regional bust extinguishes 30-60%+ of a vintage's capital simultaneously, which is a portfolio-tail statistic, not a single-deal one; (ii) there is no coupon, spread, or running yield anywhere in the structure to absorb any of it — every basis point of loss comes straight off principal, exactly as this reviewer found for v1's 1.0x floor; (iii) the upside is hard-capped at 3.0x while the downside is uncapped down to zero — a payoff shape (long a capped call, short an uncapped-to-zero put) that a derivatives desk would price at a MUCH lower expected value than "6-9%, low correlation" without a demonstrated, liquid hedge market, which per the CME evidence above does not exist. Tag: this is a quantitative/pricing judgment, not a legal one — not tagged under the legal taxonomy. --- ## 3. ADVERSE SELECTION — THE NEW FORM (OPTION-TIMING) Built a stylized 20,000-path, 30-year Monte Carlo (regime-switching annual HPI simulator: "normal" regime mean 5%/yr vol 4.5%, "bust" regime mean -8%/yr vol 6%, average regime durations ~16 and ~3.5 years respectively — an illustrative calibration to the rough order of magnitude of the historical episodes cited in Section 2, NOT an econometric fit; treat all figures below as directional, not point estimates). **Exogenous (uninformed) exit** — exit year drawn independently of the homeowner's own index path, median ~12 years (matching the underwriting box's stayer selection): mean investor IRR 3.6%, median 5.7%, p10/p90 = -1.1%/+16.8%. Floor-bound (total loss) in 3.6% of paths. Cap-bound at time of exit in 43.2% of paths. **Selection exit** — homeowner monitors the index annually starting year 2 and (a) opportunistically buys out when the current formula payoff is cheap (below 90% of Advance, i.e., near-trough — the free American put the concept sheet's own Section 6 item 3 identifies), and (b) never voluntarily settles once the cap is essentially bound (no cost to waiting further — the mirror-image free call), with the underwriting-box baseline hazard operating in between: mean investor IRR 3.3%, median 3.7%, p10/p90 = -7.2%/+15.4%. Floor-bound in only 0.7% of paths (the homeowner exits BEFORE reaching total loss, capturing the "cheap buyout" and, in practice, retaining the option to re-lever with a new investor at the lower reset basis — not modeled here, which means the drag below is understated). Cap-bound in 39.7% of paths. **Option-timing drag: median investor IRR falls from 5.7% to 3.7% (roughly 2 points, ~35% relative) purely from giving the homeowner free, penalty-free, no-appraisal timing control over settlement.** The mean-IRR drag understates this because the selection policy also truncates the worst floor outcomes (the homeowner gets out at 90 cents before the position would have gone to zero) while forfeiting nothing — an asymmetric benefit to the homeowner at every point in the distribution. This is exactly the free-American-option structure this reviewer flagged in v1 (AS3) and the concept sheet's own Section 6 item 3 concedes ("homeowners choose WHEN to settle... a free American option"). v3 does not price this option anywhere in the formula; there is no exercise fee, no minimum-hold penalty beyond an unspecified footnote, and no compensating spread. **Who enrolls?** The concept sheet's underwriting thesis (borrowed from v1: rate-locked, long-tenure, "no near-term sale plans") is in direct tension with the option-timing finding above. A rational, informed homeowner enrolls precisely BECAUSE they hold private information that favors them against the fund: either (a) they believe their specific metro is near a local top and are happy to sell 35% of a levered slice of it to someone else at what they view as full value (adverse selection at the metro-cycle-timing level, not just the household-mobility level flagged for v1), or (b) they intend to exercise the buyout option aggressively at the first sign of a dip and are indifferent to holding through a boom because the cap limits their downside from "giving away too much upside." Both selection channels work against the fund and are unpriced. Severity: CRITICAL. --- ## 4. STRUCTURAL **Cap bind frequency and the upside story.** Per the Monte Carlo, the cap binds in roughly 40-43% of simulated paths by the time of exit — not a tail event, a base-rate outcome for close to half the book under an illustrative but not aggressive calibration. Every capped position then sits earning a strictly decaying IRR for however many additional years it is held (Section 1a table D), with zero mechanism to compensate the investor for that duration. A portfolio where ~40% of positions are earning single-digit-and-falling IRRs on frozen dollar payoffs is not the return profile "6-9% expected" implies; it implies a bimodal outcome (early, high-IRR exits on a shrinking share of the book that clears the cap-bind threshold slowly; a large cluster of capped, decaying-IRR long-holds; and a meaningful floor-bound tail concentrated by metro). Severity: CRITICAL. **Extension risk with no term.** The 30-year outside date is a ceiling, not a plan. Median US homeowner tenure is climbing (mid-teens years and rising with rate lock-in); the box explicitly screens FOR long holders; and Section 3 shows the homeowner has no incentive to settle once capped. Realized portfolio duration should be expected to cluster bimodally between "genuine near-term movers" (a small, adversely-selected-against-the-fund slice per Section 3) and 20-30-year holds. No fund structure is specified in the concept sheet; any closed-end vehicle with a normal 7-12 year life cannot match this duration profile without the same secondary-market dependency this reviewer killed v1 on, and that dependency is not addressed anywhere in v3's mechanics section (only flagged as an open legal/economics question in Section 6). Severity: CRITICAL, carried forward unresolved from v1. **Origination economics.** v3 removes the ongoing servicing-intensive machinery that made v1 expensive to operate (no covenants, no occupancy monitoring, no pause/AVM disputes, no renovation-receipt review) — a genuine cost improvement on the SERVICING side, credited here. It does NOT remove the origination-side cost stack: all-owner/all-spouse execution, independent counseling, plain-dollar disclosure production, state HEI-statute compliance, KYC/AML, legal document preparation, title work, and recording all persist per the concept sheet's own Section 5. Note also that index settlement only removes the EXIT-side appraisal; establishing V0 (the enrollment value used to size the Advance and the notional) still requires some enrollment-side valuation (AVM or appraisal) — the appraisal-collusion risk is reduced (nothing to gain by inflating V0 since the investor, not the homeowner, is short the notional at a higher V0... actually the reverse: a homeowner benefits from an INFLATED enrollment V0, since Share x V0 sets the dollar notional they are selling exposure on relative to a fixed-dollar Advance — mis-marking V0 high effectively UNDER-levers the deal for the investor per dollar of Advance and dilutes the investor's take-up of a given index move; mis-marking V0 low over-levers it) but not eliminated, and the concept sheet does not address who bears this residual valuation risk or dispute cost. On a $70,000 Advance, a 2-3% origination fee is $1,400-2,100 — a smaller ABSOLUTE dollar pool than v1's $3,150 (on a $126,000 disbursement base) against a persisting fixed-cost-heavy origination process this reviewer's prior review estimated at $3,150-8,000 fully loaded (midpoint ~$5,300, adopted from the cross-exam's verified operator figures). Coverage ratio therefore likely WORSENS versus v1 (fee/cost roughly 35-50% vs. v1's ~59%), meaning the fund subsidizes origination more heavily per dollar deployed even though total servicing opex should be lower. Severity: MAJOR. **Basis-risk asymmetry (new in v3, structural, not just disclosure).** The non-recourse- to-proceeds cap means the investor's recovery is capped at actual net sale proceeds if the SPECIFIC house underperforms the metro index (investor absorbs the downside basis-risk gap). But the formula payoff is index-based with no mechanism for the investor to participate in upside if the specific house OUTperforms the metro (homeowner keeps 100% of any house-specific outperformance beyond the formula, by design — Section 2 of the concept sheet frames this as a feature: "everything you build is yours"). This is a third asymmetric option stacked on top of the two identified in Section 3: the investor is short downside basis risk and has no claim on upside basis risk. Correctly disclosed, this functions as designed (it is the honest tradeoff of index settlement) — but it means basis risk is a THIRD one-way transfer against the fund's economics, on top of the option-timing drag and the cap-vs-floor asymmetry, and it will generate the same disclosure/reputational complaints this reviewer flagged for v1's pause mechanism ("you charged me for gains my house never had") whenever a homeowner's specific block underperforms their metro at exit — a frequent, not rare, outcome given the well-documented dispersion of house-level outcomes within any single metro index. Severity: MAJOR (economic leak) / MINOR-MAJOR (reputational, churn-and-complaint drag on funnel and press coverage). --- ## 5. TAX / CEA / SECURITIES — TAGGED, NOT RE-LITIGATED This reviewer's lane is pricing and portfolio economics; the concept sheet's own Section 6 correctly identifies these as the load-bearing legal risks and defers them to specialist review. Tags applied here reflect this reviewer's read of the concept sheet's own framing, not independent legal research, and should be superseded by the tax/securities/CEA specialist reports: - Prepaid-forward / open-transaction tax treatment for a symmetric, index-settled, single-payment contract: LIKELY INTERPRETATION at best, per the concept sheet's own characterization of Rev. Rul. 2003-7 as an "analogy," not a ruling on point. - Whether INDEX (vs. property) settlement recharacterizes the contract as a notional principal contract / cash-settled derivative: AMBIGUOUS, expressly flagged by the concept sheet as potentially "worse than v1." - CEA/swap classification (CEA Section 1a(47)) and whether real-property recording + disposition-linked settlement brings it inside a forward/real-estate exclusion: AMBIGUOUS, and per the concept sheet's own framing, potentially outcome-determinative ("can kill v3 outright"). - Securities/Howey analysis on the homeowner side, and Investment Company Act Section 3(c)(5)(C) qualification for a fund of index-settled interests: the concept sheet itself flags these as "likely WORSE than v1" because the instrument resembles a derivative more than a real-estate interest once settlement is index-based rather than property-based — AMBIGUOUS, leaning unfavorable on this reviewer's read of the economics (an instrument whose payoff is a pure function of a published financial index, with no claim on the underlying property's actual value, is economically closer to a derivative than v1's property-settled contract was). - Whether a $0-floor, fully symmetric contract passes UDAAP/loss-sharing tests (e.g., the CFPB Roberts framework cited in the prior review): LIKELY INTERPRETATION that it fares better than v1's 1.0x floor, but not SETTLED — no product of this exact shape has been tested. - State HEI-statute capture of an index-settled "shared appreciation agreement": the concept sheet's own Section 6 item 5 calls this "likely yes — footprint question, not structure question." This reviewer treats that as directionally correct and notes it reintroduces the same state-by-state launch-footprint economics problem (TX exclusion, CT/MD/IL/ME capture) that the prior cross-exam found collapses origination volume below platform break-even (Dispute 5 / DC-9 in crossexam-investor.md) — a finding that transfers to v3 with no structural change in this review's domain. None of the above is resolved by anything in the concept sheet's mechanics section; all five are still open per the concept sheet's own Section 6. This reviewer's pricing conclusions (Sections 1-4) hold regardless of how these questions resolve, and get WORSE, not better, under any branch that requires a coupon, spread, minimum-hold penalty, or fixed maturity to fix the CEA/tax exposure — any such fix reintroduces duration compensation, which Section 1a shows is exactly what is missing from the return profile, and which (per this reviewer's v1 finding, still applicable) tends to look like interest for tax purposes the moment it is added. The design is still trapped between its tax/regulatory posture and its economics; index settlement changed which wall it is pinned against, not whether it is pinned. --- ## 6. VERDICT: KILL Not "fix-before-launch." The specific defect v3 was built to cure — "did risk reallocation genuinely create surplus, or did we just move the zero-sum around?" (concept sheet's own Section 6, item 6) — is empirically NOT cured. The 175-parameter frontier search in Section 1b found zero parameter combinations, across a wide and reasonable Advance/Share/ Cap space, that clear both a homeowner-beats-realistic-HELOC test and an 8% investor hurdle simultaneously. Even against the most generous plausible homeowner comparison, the frontier is one to two cells out of thirty — a rounding error, not a business. Layered on top of the pricing-identity failure: 1. The buyer thesis is unevidenced, not merely early-stage. The closest real-world analog to this exact exposure (CME Case-Shiller housing futures) has had a ~20-year public trial in an easier-to-buy, fully liquid, exchange-traded form and never attracted durable institutional real-money demand. Selling a private, illiquid, 3.5x-levered, single-counterparty, prepayment-optioned version of the same underlying at "6-9%, diversifier" pricing to pensions/sovereigns/insurers is SPECULATIVE. 2. The floor is a real, uncompensated, undiversifiable tail. A 28.6% cumulative metro decline — well inside both the 1990s California and 2006-2011 sand-states historical record, and achievable via slow bleed with no "crash" headline at all — wipes 100% of the Advance across every deal in that metro simultaneously. There is no coupon anywhere in the structure to absorb it. 3. A structural feature the concept sheet itself identifies as unpriced — the homeowner's free, penalty-free, no-appraisal timing option to buy out at troughs and let winners ride to the cap — costs the fund a directionally real and possibly understated chunk of realized return (illustrative Monte Carlo: median IRR 5.7% to 3.7%) and is nowhere compensated in the formula. 4. The cap binds in roughly 40% of simulated outcomes, converting close to half the portfolio into frozen-dollar, decaying-IRR long-duration paper with no fund-life mechanism specified to match it and no secondary market evidenced to exit it — the identical liquidity-thesis failure this reviewer killed v1 on (investor.md Verdict bullet 2), reproduced under new mechanics. 5. Every one of the five open legal/tax/CEA/securities questions the concept sheet itself poses in Section 6 remains open, and on this reviewer's read of the concept sheet's own framing, several are conceded to be equally or more dangerous than v1's (items 1-3 explicitly "worse than v1" in the sheet's own language). Any fix path for those questions that adds a coupon, spread, or minimum-hold compensation to repair the duration-non-compensation finding in Section 1a/4 reintroduces the debt/interest characterization risk the symmetric, floor-to-zero structure was chosen to avoid — the same trap this reviewer identified in v1, relocated rather than escaped. **If forced to name a parameter set and buyer profile that could work:** none was found in the tested space (Section 1b) against a realistic HELOC comparison. The closest approach — Advance 15% / Share 40% / Cap 3.5-4.0x — clears 13 of 30 cells only against a 6% investor hurdle and the generous capitalized-HELOC comparison; it requires the fund's buyer to accept ~6% nominal, pre-fee, on a levered, capped, uncompensated-tail, illiquid, single- name-credit-embedded 20-30-year instrument, which is a worse risk-adjusted return than investment-grade credit with none of the liquidity — not a viable institutional pitch at any scale this reviewer could construct. **The single make-or-break assumption:** that a levered (3.5x), collared, no-coupon, prepayment-optioned claim on a single-family metro HPI index is an asset institutional real-money allocators will price at a real-asset-diversifier discount rate (6-9%) rather than at a derivatives-desk / distressed-optionality discount rate (mid-teens or worse). Twenty years of the closest available real-world evidence (CME housing futures) says the institutional bid for this exact exposure, in a far more convenient wrapper, has not materialized. If that assumption is wrong — and this review finds no evidence it is right — the entire "genuine two-sided surplus via risk reallocation" thesis, and the fund economics built on it, do not exist. --- end of report ---