Phase A — Understand the business
Lens 1 · Company Overview
Figure is a blockchain-based capital-markets company that makes its money, today, almost entirely from home-equity lending — and is trying to use that cash-generative beachhead to build out trading and investing rails on a public Layer-1 chain. Strip the crypto vocabulary and the core is a lending-technology and loan-marketplace business with unusually good unit economics.
The plain-terms model:
- Loan Origination System (LOS). Figure built an automated HELOC origination engine. It funds a home-equity loan from application in a median of 10 days vs an industry median of ~43 days, at an average production cost of ~$717/loan vs a mortgage-industry average of ~$11,109 (MBA, Sep-30-2025). Application takes as little as five minutes; income verification, valuation (AVM instead of appraisal), lien-matching and closing are automated.
- Two channels. Figure-branded (direct-to-consumer, since 2018) and Partner-branded — a growing roster of third-party lenders originate HELOCs on Figure's technology under their own brand. As of Dec-31-2025 there were 307 active partners; for FY2025 Figure facilitated ~$8.3B of home-equity lending, +62% YoY.
- Figure Connect. A loan marketplace (launched Jun-2024) that connects sellers and buyers of loans on-chain. ~$3.9B of HELOC volume transacted on Connect in FY2025 with 48 marketplace participants. This is the fastest-growing piece — Figure Connect Volume went from $8.1M (FY2024) to $3,842M (FY2025).
Revenue construction (FY2025, $506.9M total net revenue):
| Line | FY2025 ($k) | FY2024 ($k) | YoY |
|---|
| Ecosystem & technology fees | 120,808 | 28,314 | +326.7% |
| Gain on sale of loans, net | 180,024 | 140,353 | +28.3% |
| Interest income | 74,810 | 48,207 | +55.2% |
| Origination fees | 72,536 | 64,867 | +11.8% |
| Servicing fees | 31,540 | 25,245 | +24.9% |
| Gain on servicing asset, net | 24,567 | 32,637 | −24.7% |
| Other | 2,580 | 1,262 | — |
| Total net revenue | 506,865 | 340,885 | +48.7% |
Customer/contract structure. Partner agreements are fixed-term with auto-renewal, non-exclusive, and do NOT obligate partners to use Figure — i.e. no take-or-pay, no lock-in; volume is earned, not contracted. Fees are volume-based (a % of principal originated on the LOS / transacted on Connect). The business is concentrated: HELOCs were >98% of FY2025 loan originations, and LOS-driven revenue (origination + gain-on-sale + servicing + interest) was ~71% of total net revenue (down from ~82% in FY2024 as Connect grew faster). Main "suppliers" are capital-markets buyers of the loans (whole-loan buyers, securitization trusts, and the Sixth Street JV — see Lens 2). Competitors: lending-software providers and other HELOC originators (Lens 3).
Verdict on the model: a genuinely differentiated, capital-efficient origination-and-distribution engine that has reached real profitability — with a single-product, single-asset-class revenue base and a large optionality stack (Exchange/YLDS/Democratized Prime) that is, as of today, almost all promise.
Lens 2 · Supply Chain
Map the loan lifecycle as Figure's "supply chain," naming every stakeholder:
Upstream (capital + inputs) →
- Borrowers / consumers — homeowners seeking HELOCs (FY2025 weighted-avg FICO ~748–755, customer rate ~9.1–9.2%, balance ~$91–93k).
- Warehouse / funding lenders — Figure draws on debt facilities to fund loans held for sale ($6.0B proceeds from debt drawn and repaid in FY2025). Debt, current to related parties of $166.1M at FY2025 (YLDS held by related parties).
The company (transformation) →
- LOS (origination) → DART (on-chain ownership/servicing registry) → servicing (Figure services the majority of loans it facilitates; servicing portfolio UPB $12.9B at Dec-31-2025 vs $8.1B prior).
Downstream (distribution / chokepoint) →
- Whole-loan buyers — realized gain on whole-loan sales $144.5M FY2025; UPB of loans sold rose $4.8B→$6.1B.
- Securitization trusts — Figure earns program fees arranging HELOC securitizations; securitization volume +$1.2B YoY.
- Fig SIX Mortgage, LLC — the Sixth Street JV ("Guarantor Vehicle"). Formed Feb-2025; 95% Sixth Street equity / 5% Figure, up to a $210.5M combined commitment ($200M Sixth Street + $10.5M Figure initial). Buys HELOCs originated on the LOS and sold through Connect, then sells them into Figure securitizations. Not consolidated.. This is the load-bearing demand node — Sixth Street is the anchor balance-sheet buyer that makes Connect's "liquidity" real.
- Figure Connect marketplace participants — 48 onboarded buyers/sellers/investors.
Chokepoints / single-source dependencies: (1) Secondary-market demand — the whole model depends on capital-markets appetite to buy the HELOCs Figure originates; in a credit-risk-off environment, gain-on-sale margins compress and inventory (loans held for sale, $404M at FY2025) builds. (2) The Sixth Street relationship is a concentration: one investment firm underwrites a large share of guaranteed demand. (3)
Lens 3 · Competitive Advantages (moats)
The real moat is the integrated, automated origination-to-distribution stack plus a regulatory/licensing apparatus that is genuinely hard to assemble.
- Cost & speed gap. ~$717 production cost and 10-day funding vs ~$11,109 / 43 days industry. A ~15× cost advantage on the core process is a durable structural edge if it holds as competitors digitize.
- Regulatory/licensing moat (the underrated one). >180 lending & servicing licenses, 48 money-transmitter licenses, an SEC-registered broker-dealer running an ATS, and YLDS as an SEC-registered security. Management explicitly frames this as a moat "that limits competitors' capacity to offer our full comprehensive suite" — and it's the part a well-funded fintech can't simply out-engineer; it takes years and lawyers.
- Network effects on Figure Connect. A two-sided loan marketplace where each new originator/buyer thickens liquidity. Early but real: $8.1M → $3.84B Connect volume in one year.
- Tokenized-credit share / standard-setting. ~75% of tokenized private credit by value of outstanding loans, ~$14B RWA TVL (RWA.xyz, Dec-31-2025); >$70B cumulative on Provenance since 2018. Being the de-facto standard (DART) is a positioning moat.
- Switching costs (partners). Once a partner embeds the LOS in its sales workflow, ripping it out is costly — but contracts are non-exclusive and non-obligating, so this is soft switching cost, not contractual lock-in.
Bargaining power: Over borrowers — moderate (consumer credit is competitive; Figure competes on speed/cost). Over partners — moderate and fragile (non-exclusive contracts; partners can leave). Over capital-markets buyers — this is where Figure is the price-taker: it needs buyers more than buyers need Figure in a risk-off tape. The Sixth Street JV is partly an attempt to internalize that demand.
Moat verdict: Real but narrow and asset-class-specific. It is a moat around HELOC origination technology + financial licensing, not (yet) around "capital markets." The crypto-infrastructure narrative is mostly optionality the moat does not yet protect.
Lens 4 · Segments
Figure reports as one operating segment — the CODM (the CEO) reviews financials on a consolidated basis — and is US-only, so there is no product-segment or geographic P&L breakout (no foreign geography to split; do not fabricate one). The closest thing to segmentation is the volume taxonomy and revenue-line mix:
Ecosystem volume bridge ($k):
| Volume metric | FY2025 | FY2024 | YoY |
|---|
| Ecosystem volume | 9,087,631 | 5,879,147 | +54.6% |
| — Consumer loan marketplace | 8,377,133 | 5,128,460 | +63.3% |
| —— Partner-branded | 6,401,396 | 3,447,331 | +85.7% |
| —— Figure-branded | 1,975,737 | 1,681,129 | +17.5% |
| — Digital asset marketplace | 710,498 | 750,687 | −5.4% |
| Figure Connect volume | 3,842,222 | 8,144 | n.m. |
Trend read: the growth engine is decisively Partner-branded + Figure Connect (the technology/marketplace lines), accelerating; Figure-branded (the balance-sheet-heavy DTC origination) is decelerating to +17.5%; digital-asset marketplace volume is shrinking (−5.4%) — the crypto-trading business is not yet working. So the mix shift is healthy (toward capital-light tech fees) but it underlines that the "crypto/Exchange" story is not the thing growing. Q1 2026 confirms the acceleration: ecosystem volume $3,720M vs $1,578M (+136% YoY), Connect volume $1,612M vs $478M, net take rate 3.8% vs 3.6%.
Phase B — Measure performance
Lens 5 · Earnings Result (latest print — Q1 2026, three months ended Mar-31-2026)
A clean beat-and-accelerate.
- Total net revenue $167.0M vs $84.5M, +97.6% YoY — beat consensus ~$161.6M.
- Operating income $41.9M (op margin 25.1%) vs $8.1M; net income $45.0M vs a $(0.6)M loss; adjusted EBITDA $82.7M, margin ~49.5% vs $28.3M / 33.5%.
- EPS $0.21 basic / $0.18 diluted (vs $(0.01) PY); diluted weighted shares 248.8M. EPS $0.18 ~in line with ~$0.19 consensus.
- Drivers: ecosystem & technology fees $47.3M vs $15.6M (+203%); gain on sale $49.4M vs $29.8M; 113% YoY Consumer Loan Marketplace growth and a record 80 new partners added in the quarter.
- Margin moves: the tax line swung to a $6.9M benefit (vs a $1.2M provision) — a chunk of the net-income jump is a deferred-tax benefit, not operations. Worth flagging: GAAP net income flattered by tax.
- Balance sheet: cash & equivalents $1,464.6M vs $1,198.1M at FY2025; total assets $2,730.9M. Strong, post-IPO-fortified.
- Market reaction: muted-positive — shares +~3.7% after-hours / +1.36% on the session. A near-100% revenue beat that moves the stock <4% tells you a lot is already priced in.
Vs the company's own history: the trajectory is striking — op income (loss) of $(49.4)M (2023) → $9.2M (2024) → $117.5M (2025); revenue $209.5M → $340.9M → $506.9M. This is a business that crossed into durable GAAP profitability in 2024 and is now compounding it. FY2025 adjusted EBITDA $251.2M (49.6% margin). Q3 2025 (first quarter as a public co) was also a beat: EPS $0.34, revenue $156M (+42%), adj-EBITDA margin 55%.
Lens 6 · Earnings Calls (sentiment trend)
No transcripts on disk (transcripts=0). From web coverage of the available prints (Q3 2025 → Q4 2025/FY → Q1 2026):
- Consistent themes management keeps hammering: partner growth ("record 80 new partners" in Q1'26), Consumer Loan Marketplace / Figure Connect volume, adjusted-EBITDA-margin expansion (mid-50s%), and the "future of capital markets on blockchain" framing.
- Tone shift: from IPO-debut confidence (Q3'25, stock +109% off IPO) toward a more defensive, "the operating numbers justify the multiple" posture as the stock fell from ~$78 to high-$20s and BofA downgraded on valuation. Management leans harder on margins and volumes (the strong, defensible facts) and less on the speculative Exchange/YLDS lines (which still produce ~no revenue).
- What they've gone quiet-er on: near-term monetization of Figure Exchange / Democratized Prime — the filings repeatedly say "no material revenue" and "we expect [it] to grow as users recognize its benefits," which is investor-relations for not yet.
(Sentiment read is -grounded + filing-language, not a transcript NLP run — flagged as lower-confidence than the rest.)
Lens 7 · Comps
FIGR has no clean peer. It straddles fintech lending (SoFi, Rocket, Upstart, LendingClub) and tokenized-RWA / crypto-infra (Ondo, Maple, Centrifuge — mostly tokens, not equities). Table below; multiples are `` with date or n/a. Do not read precision into them.
| Company | Ticker | Mkt cap | P/E | P/S | EV/EBITDA | Notes |
|---|
| Figure Technology | FIGR | ~$6.3B | ~69× (norm.) | ~8.3× | ~25× | ~50% adj-EBITDA margin; +98% rev YoY |
| SoFi Technologies | SOFI | n/a | ~24–27× (fwd/NTM) | n/a | n/a | 30%+ rev growth '26; bank charter |
| Upstart | UPST | n/a | ~11× (NTM) | n/a | n/a | AI underwriting; ~40% rev growth '26 |
| Rocket Companies | RKT | n/a | n/a | n/a | n/a | price ~$13.63, PT ~$23 |
| LendingClub | LC | n/a | ~8.3× (fwd) | n/a | n/a | slower growth |
| Ondo / Maple / Centrifuge | (tokens) | n/a | n/a | n/a | n/a | RWA-native; not equity comps |
EV/EBITDA ~25×: mkt cap $6.3B, net cash positive ($1.46B cash, ~$0.56B debt → EV ~$5.4B), FY2025 adj-EBITDA $251M → ~21×; on a forward ~$330M+ adj-EBITDA run-rate (Q1'26 $82.7M ann.) → ~16×. So mid-teens-to-low-20s forward EV/EBITDA depending on annualization — rich for lending, reasonable for software if the margins and growth hold.
Read: On earnings, FIGR trades at ~2.5–3× the P/E of SoFi and ~6× Upstart — it is priced as a high-growth software/crypto-infra name, not as a lender. The bull defense is the ~50% EBITDA margin + ~100% revenue growth + tokenized-credit dominance; the bear is that every fintech lender that got a software multiple eventually got re-rated to a lender multiple when origination volumes wobbled.
Lens 8 · Stock-Price Catalysts (the >5% moves)
FIGR has only ~9 months of tape (IPO Sep-11-2025), but it has been violent:
- Sep-11-2025 — IPO. Priced $25 (raised ~$787.5M, ~$5.3B valuation), opened/closed first day ~$31.11 (+24%). Also marketed as the first on-chain IPO on Solana.
- Sep–Nov 2025 — melt-up to ~$78. Stock surged ~109% in the five months post-IPO to a 52-week high of $78.00. Catalyst: Q3'25 beat (EPS $0.34, +42% rev), crypto-narrative euphoria, index inclusion chatter.
- Late-2025 — the unwind. BofA downgraded Neutral→Underperform on valuation; insider selling in Q4; a secondary offering of blockchain common stock priced at $32; and shares dropped ~10% on preliminary Q4 operating metrics.
- Jan-28-2026 — data breach disclosed (ShinyHunters, ~967K records, SSNs) — overhang.
- Feb-19-2026 — tokenized stock debut + upsized $150M offering at ~$32.
- Mar-10-2026 — IPO lock-up expiry. Supply released.
- May-12-2026 — Q1'26 beat, +~3.7% after-hours (muted).
What the tape reveals: the market reacts to (1) crypto-narrative beta (the melt-up and the breach both moved it hard), (2) valuation calls (the BofA downgrade mattered because the multiple is the whole debate), and (3) supply events (secondary, lock-up). It reacts surprisingly little to operational beats now — the story is contested, so good numbers get discounted. This is a stock where the multiple, not the print, is the variable.
Phase C — Judge people & books
Lens 9 · Management
Founder vs operator split is the defining governance fact.
- Michael Tannenbaum — CEO (since Apr-2024), director (since Jul-2024). The operator running the public company. Track record: COO of Brex (2021–24), CFO/CBO of Brex (2017–21); earlier Chief Revenue Officer of SoFi (2014–17); H&F private equity and JPMorgan IBD before that; Columbia summa cum laude. A credible fintech operator with a revenue/finance pedigree — and notably another SoFi alum.
- Michael (Mike) Cagney — co-founder, director, Controlling Party. Not CEO of the public entity — he is CEO of "FMH" and President of "Figure REIT" (affiliated entities), and a board member. He co-founded Figure in 2018 with his wife June Ou. Before Figure he was co-founder/CEO/Chairman of SoFi (2011–2017) — built it into a multi-billion-dollar lender — and was forced out in Sep-2017 amid sexual-harassment lawsuits and a NYT "frat-house" culture exposé. He is a genuinely talented, blockchain-native financial builder with a real personal-conduct cloud in his history.
- Skin in the game / control: dual-class structure — Class B = 10 votes/share concentrates voting control with Cagney; the 10-K flags this as a risk that "may depress the trading price". Founder controls the vote without running day-to-day ops — an unusual and governance-relevant separation.
- Capital allocation: so far disciplined — grew from $209M→$507M revenue while turning a $49M operating loss into $118M operating income, capital-efficiently (capex is just capitalized software, ~$17–21M/yr). IPO raised ~$663M net; cash hoard $1.46B. No buyback of scale yet (some buyback backdrop chatter pre-lock-up ). The Sixth Street JV is a smart way to source guaranteed demand off-balance-sheet.
- Red flags (governance/related-party): the entity structure is a thicket of affiliates — FMH, Figure REIT, FLC, FT — all in Cagney's orbit. Related-party items in Q1'26: Reflow Services (Figure holds 17.3%, contributed by the Controlling Party); the Domestic Solana Fund (acquired SOL via the FTX bankruptcy / Alameda Research); Hastra/Signum software-license royalty (50bps); executive & Controlling-Party YLDS holdings in Democratized Prime; Controlling-Party travel costs run through G&A. None individually large, but collectively they describe a company deeply intertwined with its founder's other vehicles — exactly the structure that demands scrutiny.
Archetype: founder-controlled (Cagney's vote) but professional-operator-run (Tannenbaum). Implication: for a high-velocity fintech this is arguably the right split — adult in the CEO chair, visionary on the board — if governance holds. The Cagney history + the related-party web are the offsets.
Lens 10 · Forensic Red Flags
Acting as a forensic analyst — the accounting here is unusually fair-value-heavy, which is where the bodies could be.
- MATERIAL WEAKNESSES IN INTERNAL CONTROL — disclosed, as of Dec-31-2024. The 10-K states: "we did not design or maintain an effective control environment… did not maintain adequate resources in the accounting and finance functions with requisite knowledge, skills, and experience to accurately apply GAAP… did not design and maintain effective risk assessment and monitoring," which "contributed to several additional material weaknesses" over technical-accounting transactions, manual review of loan assets on the balance sheet, and review of valuation specialists' deliverables. Management says no material misstatement has been identified and remediation began in FY2025 — but for a company whose revenue is dominated by fair-value marks, a weak accounting function is the single most important red flag in this dossier. This is not theoretical; it is the company's own admission.
- Fair-value-dependent earnings. A large share of net revenue is gain on sale of loans ($180M) and gain on servicing asset ($24.6M) — both fair-value-driven. Loans held for sale ($404M) and servicing assets ($113M) are Level-3 (model-priced; discount rates 5.3–26.9%, CPR/CDR assumptions). Earnings quality is therefore assumption-sensitive — a haircut to MSR or whole-loan marks flows straight to the P&L. (Adjusted Net Revenue exists precisely to strip MSR fair-value noise — telling.)
- Cash flow vs earnings divergence. FY2025 net income $134.3M but CFO only $62.6M; non-cash adjustments of $(128.3)M. The gap is the fair-value gains (non-cash) and working-capital swings from loan origination/sale flows. Not alarming for a loan-flow business, but it means GAAP net income overstates cash generation, and the prior year had negative CFO of $(127M). There was also a cash-flow-statement presentation correction (retained securitization beneficial interests reclassified operating→investing; deemed immaterial) — a presentation error, consistent with the control-weakness theme.
- Tax benefit flattering net income. FY2025 included a $20.6M income-tax benefit and Q1'26 a $6.9M benefit (deferred-tax recognition), inflating GAAP net income vs a normalized tax rate.
- SBC. $62.4M stock-based comp in FY2025 (a meaningful add-back in adjusted EBITDA) — standard for the cohort but dilutive (diluted share count 248.8M vs 217.3M basic in Q1'26).
- Digital-asset / crypto exposure on balance sheet — $96.6M digital assets (FY2025), SOL via the FTX-bankruptcy fund, distressed-asset bankruptcy claims (Level 3). Mark-to-market volatility risk.
Regulatory findings (required sub-section).
- SEC Litigation Releases / AAERs: None. "No LR found" and "No AAER found" for Figure Technology Solutions, 2021-06-20→2026-06-20, per SEC EDGAR EFTS.
- 10-K Item 3 (Legal Proceedings): the index routes Item 3 to p.98; the 10-K's risk-factor and legal language describes ordinary-course litigation/regulatory exposure inherent to lending (TILA/HOEPA/Reg Z, FDCPA, RESPA, CFPB UDAAP, state money-transmitter regimes) but no disclosed material adverse pending matter is flagged as a stated liability. (Note: the FY2025 10-K predates the Jan-2026 breach litigation below.)
- Non-SEC enforcement (web): No FTC/DOJ/CFPB enforcement action found. BUT a material event: Figure disclosed a data breach on Jan-28-2026 — ~967,200 records exposed (names, SSNs, loan details), claimed by ShinyHunters via social engineering of an employee; multiple class actions / investigations are pending. Not (yet) a regulator action, but a real legal/regulatory and reputational liability for a lender holding sensitive PII.
- Summary: No SEC enforcement and no AAER as of 2026-06-20 (verified via EDGAR EFTS LR+AAER + 10-K Item 3 + web). The two live forensic/legal concerns are (a) the self-disclosed material weaknesses and (b) the 2026 data-breach litigation — neither is a fraud finding, both are real.
Phase D — Project & stress-test
Lens 11 · Forward Projection (FY2026 / FY2027 / FY2028 EPS)
Built bottom-up off the latest actuals. All inputs labeled; output `` with arithmetic. Do not treat as consensus — no published EPS consensus beyond near-term was sourced.
Anchors: FY2025 net revenue $506.9M, net income $134.3M, adj-EBITDA $251.2M; Q1'26 net revenue $167.0M (+98% YoY), net income $45.0M, diluted shares 248.8M.
Run-rate sanity check: Q1'26 annualized = ~$668M revenue / ~$180M net income / ~$331M adj-EBITDA. But Q1 had a tax benefit and gain-on-sale strength; normalize down.
- Base case (FY2026). Revenue ~$760M (+50% YoY; volume +136% in Q1 decelerating as comps toughen and rate environment normalizes). Net margin normalizes to ~22% (less tax benefit, some gain-on-sale mean-reversion) → net income ~$167M; on ~255M diluted shares → EPS ~$0.66. Adj-EBITDA ~$370M.
- Bull (FY2026). Revenue ~$840M (+66%), partner ramp + Connect + first real Exchange/YLDS contribution; net margin 25% → net income ~$210M → EPS ~$0.82.
- Bear (FY2026). Revenue ~$610M (+20%, HELOC volume stalls on a credit/rate shock + secondary-market demand softening), margin compresses to 16% on gain-on-sale haircut → net income ~$98M → EPS ~$0.38.
- FY2027 (base): revenue ~$990M (+30%), EPS ~$0.90.
- FY2028 (base): revenue ~$1.24B (+25%), EPS ~$1.15.
Valuation implication: at ~$28.5, base FY2026 EPS ~$0.66 → ~43× forward P/E; FY2028 ~$1.15 → ~25× on numbers two years out. The growth is real but the price already pays for two-to-three years of flawless execution. The bear EPS ($0.38) would re-rate the stock toward a lending multiple (~15–20×) = ~$6–8 of value on earnings — hence the $33 low-end analyst target and the asymmetry.
(Per --watchlist rules, NOT logging a Brier forecast via forecast.ts — no committed base case in breadth mode.)
Lens 12 · Bull vs Bear
Bull case. Figure is the tokenized-credit standard (~75% share, $14B RWA TVL) wrapped around a genuinely superior, ~15×-cheaper HELOC origination engine that already throws off ~50% adjusted-EBITDA margins and ~100% revenue growth — a combination almost no public fintech can match. The Partner-branded + Figure Connect flywheel is compounding (307 partners, +80 in one quarter), it's capital-light (capex = software), net-cash-rich ($1.46B), and it has a regulatory/licensing moat (180+ licenses, broker-dealer/ATS, SEC-registered stablecoin) that rivals can't quickly replicate. The Exchange / YLDS / Democratized Prime stack is free optionality on a multi-trillion-dollar tokenization wave that the market is not paying much for today. If even one of those becomes a real revenue line, the "lender multiple" bear thesis breaks and FIGR re-rates as crypto-capital-markets infrastructure. Analysts agree directionally: Buy consensus, avg PT $53.88 (high $75), Mizuho $55 Outperform.
Bear case (permanent-impairment risks).
- It's one product. >98% of originations are HELOCs; ~71% of revenue is HELOC-derived. A US housing/credit downturn or a HELOC-demand normalization (rates fall, cash-out refi returns) hits the whole P&L at once. The moat is around HELOC tech, not "capital markets."
- Gain-on-sale + Level-3 marks are the earnings. Earnings quality depends on secondary-market demand and model assumptions; a risk-off tape compresses gain-on-sale margins and could turn fair-value gains into losses — and the company just disclosed material weaknesses in the accounting function that prices those marks.
- The multiple is the bet. ~69× P/E / ~8× sales prices flawless growth; the stock already fell from $78 to ~$28 once the market questioned that. Lock-up expired, ~7% short interest, insiders sold.
- Governance + founder risk. Super-voting founder with a real conduct history, a web of related-party affiliates, and a 2026 breach exposing ~1M SSNs.
Pre-mortem (it's Dec-2027, the thesis broke — what happened?): Rates fell through 2026–27, banks re-entered HELOC aggressively, and cash-out refis came back — Figure's volume growth halved. With gain-on-sale margins compressed and the Exchange/YLDS lines still immaterial, the market re-rated FIGR from a software multiple to a lender multiple. A second control/restatement issue (or breach-litigation settlement) accelerated the de-rating. Stock $12.
Are multiples too high? Yes, on earnings — ~69× trailing / ~43× forward base is a software multiple on a business still ~98% lending. The bull defense (margins, growth, optionality) is legitimate but requires both sustained HELOC volume and a new revenue line to validate the price.
Contrarian view (what the market refuses to see, both ways): Bears dismiss it as "a HELOC lender in a crypto costume" and miss that the licensing + DART standard + on-chain settlement is a genuinely defensible distribution moat that could compound for a decade. Bulls dismiss the single-product, fair-value, founder-control fragility and pay a multiple that only works if the optionality monetizes on schedule. The truth is a great business at a demanding price — the disagreement is entirely about what you pay, not whether it's good.
Lens 13 · Devil's Advocate (short-seller)
Dismantling the bull case:
- Where revenue is concentrated: one asset class (HELOC), one country (US), one demand structure (secondary-market buyers + the Sixth Street JV). Shift any one and the model wobbles. If Sixth Street pulls or re-prices its commitment, Connect "liquidity" thins.
- Why the moat is weaker than bulls think: partner contracts are non-exclusive and non-obligating — there is no contractual lock-in. The cost/speed edge is real but replicable as legacy lenders digitize (the 10-K admits well-resourced competitors with data-science teams could build competing platforms). The "75% of tokenized private credit" stat is share of a tiny, nascent market — impressive optically, thin economically.
- Most dangerous competitor bulls underestimate: not the crypto-RWA names (Ondo/Maple/Centrifuge) — it's a re-engaged depository or a SoFi/Rocket-scale fintech bringing balance sheet + brand + its own digital HELOC. Banks de-emphasized HELOC in 2023; if they re-emphasize as rates fall, Figure's whitespace shrinks.
- Worst capital-allocation / governance moves: the related-party lattice (FMH, Figure REIT, Reflow contributed by the Controlling Party, the FTX/Alameda-sourced Solana fund, Hastra royalties, Controlling-Party travel through G&A) — every one is small, but the pattern is a company run alongside the founder's other interests under a super-vote. Add insider selling into the post-IPO spike and a secondary at $32 while telling public holders to believe in the multiple.
- Assumptions that must hold for ~$28: HELOC volume keeps compounding ~30%+; gain-on-sale margins hold; the accounting weaknesses get remediated with no restatement; the breach litigation stays contained; and at least one of Exchange/YLDS/Democratized Prime becomes material. That's a lot of ANDs.
- If growth disappoints 20–30%: on bear EPS ~$0.38 and a lender-like 15–18× multiple, the stock is ~$6–7 — a >75% drawdown from here.
- Single permanent-impairment scenario, and plausibility: a material restatement stemming from the disclosed control weaknesses in the fair-value loan/MSR marks — destroying the "trustworthy on-chain truth" brand that is the entire pitch. Plausibility: low-to-moderate (no misstatement identified yet; remediation underway), but the consequence is severe because Figure literally sells "displacing trust with truth."
Lens 14 · Management Questions (ordered by information value)
- The FY2024 10-K disclosed material weaknesses in the accounting/finance function — including over fair-value loan and MSR marks that drive most of your revenue. Exactly what remediation is complete, what remains, and when will the auditor attest controls are effective?
- What share of net revenue is fair-value gain (gain-on-sale + MSR + Level-3 remeasurement) vs cash fees, and how do those marks move in a 100bp credit-spread-widening / risk-off scenario?
- Concentration: what % of Figure Connect demand and securitization exit runs through the Sixth Street JV and your top three loan buyers — and what happens to gain-on-sale margins if they step back?
- HELOCs are >98% of originations. In a scenario where rates fall and cash-out refis return, what is your realistic HELOC-volume growth, and what's the bridge to revenue from non-HELOC products?
- Figure Exchange, YLDS, and Democratized Prime still produce ~no revenue. What are the specific 12–24 month revenue milestones, and at what point do you stop funding them if adoption lags?
- Walk through the related-party architecture (FMH, Figure REIT, Reflow, the Domestic Solana Fund, Hastra). How do you ensure arm's-length terms, and why are these adjacent to — rather than inside — the public company?
- The dual-class super-vote concentrates control with Mr. Cagney, who does not run the public company day-to-day. What governance guardrails protect minority holders, and is there a sunset?
- What is the status and estimated financial exposure of the January-2026 data-breach litigation, and what changed in your security posture after an employee was socially engineered?
- Your average production cost is ~$717/loan. How much of that gap vs the ~$11,109 industry average is structural (automation) vs cyclical (low volume periods), and where does it go at 3× scale?