☰Figure Technology Solutions is usually described as a blockchain company that happens to lend, and that framing is backwards. The investable business is the largest non-bank home-equity lender in the United States, running an origination stack that produces a loan for roughly $700 against an estimated $11,000 at a legacy bank — and which has, over eight quarters, converted that cost advantage into a two-sided marketplace that other lenders now originate on top of.
The compounding engine is Figure Connect. In 2024 essentially all volume was Figure's own; by the second quarter of 2026, 65% of a $4.26 billion quarterly marketplace flowed through third-party partners, and the partner count rose from 307 to 489 in six months. That is the difference between a very good lender and a piece of financial infrastructure: the first is capped by its balance sheet and its cost of capital, the second earns a fee on other people's balance sheets and re-rates accordingly. Ecosystem and technology fees became the single largest revenue line for the first time in the June quarter, growing 159% year over year against 113% for the company overall.
What must prove true is that the marketplace keeps widening faster than the take rate compresses. Take rate has already fallen from 4.0% to 3.6% and management guides to the low end of a 3.5–4.0% band. The bull case requires the partner network, the Kiavi first-lien acquisition, and on-chain funding via Democratized Prime to expand volume fast enough that a thinner slice of a far larger pie still compounds revenue in the low-20s percent for a decade — and that the market eventually pays an infrastructure multiple rather than a lender multiple for the result.
First mover in a massive TAM (moderate-strong). Figure sits at the intersection of two large origination markets: US home equity, where tappable equity is at record levels and HELOC credit limits have passed $1 trillion against roughly $200 billion of annual originations, and residential transition lending, a further ~$200 billion annual market entered via the $717 million Kiavi acquisition. Figure was the third-largest HELOC originator in the country in 2024 and the largest non-bank, and its ecosystem has facilitated over $30 billion of loans cumulatively. The honest qualification is that this is share capture in a mature, cyclical credit market rather than the creation of a new one — the runway is real but it is a share runway, not a greenfield runway, and it is levered to housing and rates.
Network effects and data flywheel (strong, and visibly igniting). This is where the thesis lives. Figure Connect is a genuine two-sided marketplace: originators bring loans, capital providers and securitization buyers bring bids, and each side makes the other more valuable. The evidence that the flywheel has caught is unusually clean. Third-party volume went from approximately nothing in 2024 to 46% of marketplace volume in 2025 to 65% in the June 2026 quarter, while active origination partners rose from 307 at year-end to 489 six months later, with 102 added in a single quarter. Deeper supply has tightened execution: AAA securitization spreads compressed to 135 basis points year to date from roughly 255 in 2023, and Figure has obtained AAA ratings from both S&P and Moody's on blockchain-collateralised deals — a first, and a credential competitors cannot simply buy. The data flywheel is thinner than the network effect: $30 billion of origination history and the DART lien registry produce proprietary performance data, but this is not a consumer behavioural loop that improves a product for every user.
Disruptive technology (strong on substance, contested on framing). The load-bearing claim is unit cost: roughly $700 to produce a loan against an estimated $11,000 for a legacy bank, with operations and processing costs falling to 67 basis points of volume from 79 a year earlier on AI-driven automation. That is a roughly sixteen-fold cost compression in a value chain that has resisted change for decades, and it is why banks and fintechs increasingly originate on Figure's rails rather than compete with them. The contested part is attribution. A short-seller report published in April 2026 argued, citing Figure's own filings, that the loan origination system does not itself run on blockchain and that loans are tokenized after the fact as digital twins; the company responded that every loan is represented on chain from funding onward and that all subsequent transfers and pledges execute on chain, with paper documents serving as legal formalities. For an investor the distinction matters less for the economics — a cost advantage is a cost advantage whether it comes from blockchain or from very good conventional software — than for the multiple, because the blockchain framing is what supports a premium to fintech peers.
AI-disruption-resistance (strong — two anchors, one of them hard). Figure clears this sub-criterion decisively on anchor (b), regulated industry. Originating and servicing secured consumer credit requires state lending, servicing and money-transmitter licences across the country; the securities businesses sit inside an SEC-registered broker-dealer and Alternative Trading System; and YLDS is issued by an SEC-registered face-amount certificate company. A general-purpose AI agent cannot originate a HELOC, perfect a lien, or issue a registered security — it faces the same regulatory wall as any new entrant. Anchor (d), proprietary data plus network effect, is satisfied more narrowly through the 489-partner marketplace and the DART registry, which is being used as a functioning MERS alternative. There is a partial claim on anchor (a), physical, in that the value chain terminates in a recorded lien on real property. No AI-disruption risk flag is required.
Net: a genuinely defensible, licence-protected business whose marketplace flywheel has demonstrably ignited over the past eight quarters, held below eight out of ten by three things — the underlying market is a mature cyclical credit market rather than an expanding one, the tokenized-asset position is mid-pack by total value locked and overwhelmingly self-generated, and the durability of the blockchain premium itself is an open question rather than a settled fact.
Trait 1 — Missionary vision (20%) — 8/10
Founder and Executive Chairman Mike Cagney has pursued one specific idea since 2018: move the plumbing of secured lending onto blockchain rails and collapse the cost of a loan by an order of magnitude. Almost every product traces back to it — Provenance as the ledger, DART as the lien registry, Figure Connect as the marketplace, YLDS as the settlement asset, Democratized Prime as the funding layer, and a blockchain-native share class as the proof of concept applied to the company's own equity. That is a mission concrete enough to guide capital allocation over eight years, which is rarer than it sounds. The score is held at eight rather than nine because the missionary is the Chairman, not the CEO: Michael Tannenbaum, in seat since April 2024, is an execution operator rather than the mission's author, and Cagney is compensated as a director rather than as an executive officer.
Trait 2 — Radical long-termism & skin in the game (25%) — 5/10
The structural picture is exactly what the framework wants: a triple-class structure in which Class B carries ten votes per share gives the founder roughly 67–68% of voting power on approximately 20–22% of the economics, so public shareholders cannot remove him and he can invest through cycles. The behavioural picture is not. Cagney has sold about $76 million of stock since the September 2025 IPO at an average near $30 with no open-market purchases, the CEO sold roughly $60 million including 908,000 shares two months after listing, the CFO has been selling a small block roughly every two weeks through August 2026, and one director exited his entire position. Against that, an IPO-date Founder Retention Award worth approximately $169 million — equal to 4% of shares outstanding — had its one-year vesting cliff removed by the board two months after grant. The performance tranche, with price hurdles up to 252% of the IPO price, is genuinely aligned, and the sales were made under Rule 10b5-1 plans. But cash-out velocity of this order in the first year post-listing is the single clearest deviation from the long-termism the framework prizes, and it is why this trait carries the largest deduction at the largest weight.
Trait 3 — Product & customer obsession (20%) — 7/10
Management speaks in the right units: partner counts, take rate in basis points, cost per loan, contribution margin by channel, weighted-average FICO of 756 and CLTV of 62%, delinquency by vintage, and weekly application volume. Since July 2026 the company publishes an operating metrics dashboard every Tuesday after the close, which is a higher-frequency commitment to transparency than almost any peer makes. Product cadence is real and continuous — Figure Connect in 2024, YLDS in 2025, the blockchain share class and the Adaptor onboarding agent in 2026. The offsetting fact is selective disclosure: the Ecosystem Volume table was removed from the second-quarter release, and the metric it contained implies the digital-asset side of the business was flat to declining. Dropping a metric when it turns is the opposite of the behaviour this trait rewards.
Trait 4 — Execution velocity (20%) — 8/10
The delivery record over the last twelve months is hard to fault. Second-quarter marketplace volume of $4.26 billion beat the top of guidance by 4%; 102 origination partners were added in a single quarter; third-party volume crossed from a minority to 65% of the marketplace; AAA ratings were secured from both S&P and Moody's on blockchain-collateralised securitizations; and the company completed an IPO, created and listed a blockchain-native share class, priced $600 million of senior notes, and signed a $717 million acquisition — with a three-person executive team. Against that, two initiatives reversed: the national bank charter application filed with the OCC in November 2020 was withdrawn in July 2023, and Figure Markets was split out in 2024 only to be merged back in during 2025. Material weaknesses in internal controls identified at IPO were still listed as unremediated in the August 2026 risk factors.
Trait 5 — Capital efficiency & financial discipline (10%) — 6/10
On the headline numbers this looks excellent: GAAP profitable with roughly $237 million of trailing net income, a 54.6% adjusted EBITDA margin, $1.44 billion of cash, and roughly $1.1 million of trailing revenue per employee across about 640 people (about $1.4 million on the June-quarter run rate). Three things pull the score back to six. Trailing operating cash flow is negative $86.5 million — structurally normal for a warehouse-funded originator, but it means the reported profit is not yet cash at the operating line. Dilution is heavy: roughly 26% of shares outstanding sit in options and RSUs at a weighted-average strike of $6.60, and quarterly stock compensation rose from $2.8 million to $26.1 million year over year. And the capital allocation signals conflict — a $200 million buyback authorised in February has been executed at only about 14%, while $600 million of debt was issued at 8.5% and insiders sold well over $150 million.
Trait 6 — Talent magnetism & organisational scaling (5%) — 5/10
There has been no executive turnover since the IPO, which is a genuine positive, and roughly $1.1 million of trailing revenue per employee indicates a high-calibre, leverage-heavy organisation. But the bench is extraordinarily thin for an $8.5 billion company: three Section 16 officers, with no named President, COO or CTO, running simultaneously a national lender, a public blockchain, a broker-dealer and ATS, a registered stablecoin issuer, and the integration of a $717 million acquisition. Governance is concentrated in the founding family, with the co-founder and the founder's spouse sitting on the compensation committee. Glassdoor sits at 3.7 out of 5 with 67% recommending — adequate, not a differentiator. Culture appears personality-dependent rather than encoded.
Valuation — FLAG, but the flag needs qualifying
On the most recent full fiscal year, FY2025 revenue of $506.9 million, Figure trades at approximately 16.8x sales. On trailing twelve-month GAAP revenue of $709 million the multiple is 12.0x, and on consensus 2026 revenue of roughly $820 million it is about 10.4x. Every one of those is well above the framework's sub-5x entry discipline, and the asset-light exception — which Figure genuinely qualifies for on a 54.6% adjusted EBITDA margin — does not stretch far enough to make 12x a disciplined entry. The necessary qualification is that this company is profitable, which most names screened at these multiples are not: roughly $237 million of trailing net income puts the stock at about 36x earnings, and 36x for a business that grew revenue 113% year over year is not, on its face, expensive. The P/S discipline was built for pre-profit platforms and mechanically over-penalises a profitable one. The honest reconciliation is that the entry is expensive on the framework's primary metric and merely full on the secondary one, and that the gap between them is where the argument sits. One screener note: S&P reports trailing revenue of $619 million under a financial-sector netting definition, which produces a 13.7x multiple — the figures here use GAAP net revenue as reported.
Revenue and margin trajectory — excellent, with quality caveats
Second-quarter net revenue of $225.6 million grew 113% year over year, operating income rose 180% to $77.7 million for a 34.5% margin, and net income rose 192% to $87.4 million. First-half revenue is up 106%. Adjusted EBITDA margin expanded 7.4 points to 54.6%. Three quality caveats belong alongside those numbers. Gain on the servicing asset jumped from $1.8 million to $29.1 million year over year and now represents 12.9% of revenue — a non-cash fair-value mark on newly created MSRs, not collected cash. Adjusted net revenue of $218.4 million came in below GAAP revenue for the quarter, meaning GAAP was flattered by those marks rather than depressed by them. And net income exceeded pre-tax income because of a $4.4 million tax benefit. Separately, the company-reported net take rate - a company-defined revenue measure over marketplace volume, not GAAP net revenue divided by volume, which is a higher 5.3% - compressed from 4.0% to 3.6%, and management guided to the low end of its 3.5–4.0% band for the third quarter; the company attributes this to the deliberate mix shift toward capital-light Figure Connect volume rather than to partner pricing pressure, which is credible but means revenue will compound more slowly than volume.
Balance sheet and path to profitability — funded, but the shape is changing
At 30 June 2026 Figure held $1.44 billion of cash against $963 million of total debt, roughly $475 million net cash, on $1.41 billion of equity, with warehouse facilities drawn at only about 18% of $2.39 billion of capacity. That is a comfortable position. It is being spent: $600 million of unsecured senior notes closed in July at 8.5% to fund the $717 million Kiavi purchase, which moves the pro-forma balance sheet to roughly net-neutral and adds a fixed 8.5% coupon to the cost structure. Profitability is not a future milestone here — it has been achieved and is expanding — so the relevant question is not path to profit but durability of profit through a credit cycle. Trailing operating cash flow of negative $86.5 million reflects loans held for sale consuming operating cash, which is normal for the model but means reported earnings and cash generation are not the same thing. One further caveat: $424.6 million of current debt is owed to related parties and grew 2.6x in six months.
The blockchain premium is the valuation, and it is contested
A short-seller report published on 16 April 2026, by a firm disclosing a short position, argued that Figure is a conventional home-equity lender carrying an approximately 100% premium to fintech peers on a blockchain narrative — citing Figure's own filings to argue the origination system does not run on chain, that roughly 99% of YLDS was held by Figure and one passive investor at end-2025, and that Figure and affiliates control a majority of the Provenance governance token. Figure disputed the token concentration figure and stated that every loan is represented and transferred on chain from funding onward. Neither side has closed the question. The market's answer matters more than the philosophical one: if the premium compresses toward a lender multiple, roughly half the equity value is at stake regardless of how the loans are recorded.
Rate and credit cycle exposure runs both ways
Fed funds sat at 3.50–3.75% in August 2026 after five consecutive holds, with three FOMC members dissenting in favour of a hike and the ten-year Treasury near 4.65% and rising. The current setup favours Figure: high first-lien rates intensify the lock-in effect, pushing homeowners toward second liens instead of cash-out refinancing. But the exposure is genuinely two-sided. A sharp fall in long rates would unlock cash-out refis and cannibalise HELOC demand; a sharp rise would compress gain-on-sale spreads, raise funding costs against a freshly issued 8.5% coupon, and pressure borrower affordability. Delinquent balances on loans held for sale rose from 3.91% in 2024 to 5.46% in 2025 on the whole book, against a securitized-pool weighted average the company puts at 0.80% — different denominators, both capable of being true, and worth monitoring as vintages season.
Take-rate compression against volume growth
Revenue capture has fallen from 4.2% in 2024 to 4.0% in 2025 to 3.6% in the June quarter, with guidance to the low end of the band. Management frames this as the intended consequence of shifting to capital-light marketplace volume and notes that expanding first-lien volume — precisely what Kiavi does — is a further modest headwind. The thesis therefore requires volume to grow materially faster than the take rate falls, indefinitely. It has done so far. If volume growth normalises toward market rates while take rate keeps sliding, revenue growth decelerates faster than the volume headline suggests.
Governance concentration and insider selling
The founder controls roughly 67–68% of voting power on approximately 20–22% of the economics, so public shareholders have no mechanism to influence direction. That structure cuts both ways in this framework — it enables long-term decisions, and it removes the check if judgement fails. In the first year post-IPO the founder sold roughly $76 million, the CEO roughly $60 million, the CFO has been selling on a fortnightly cadence, and one director exited entirely, while the company executed only about 14% of its buyback authorisation. There is an explicit anti-hedging policy but no anti-pledging policy disclosed.
Kiavi integration and execution capacity
The $717 million Kiavi acquisition is strategically the right move — it adds first-lien residential transition and DSCR lending to a business that is otherwise almost entirely second-lien, opens roughly $200 billion of annual addressable origination, and is structured so that a joint venture with Sixth Street holds the balance-sheet assets while Figure buys the platform. It is also a large, debt-funded diversification into cyclical investor-property credit, not yet confirmed closed, to be integrated by a three-person executive team that has no COO or CTO. The stated economics are attractive — over $250 million of revenue and over $100 million of EBITDA, accretive to EPS with under a four-year unlevered cash payback — but they are the company's figures and are unproven under Figure's ownership.
Internal control material weaknesses
Risk factors in the August 2026 release still list the company's ability to remediate previously identified material weaknesses in internal control over financial reporting. For a business whose reported revenue includes large non-cash fair-value marks on servicing assets, unremediated control weaknesses are a more consequential item than they would be at a simpler company.
Read off the actual price series rather than the headline drawdown, Figure has just completed a textbook framework setup — and it has already been arbitraged. The stock reached an intraday high of $78.00 on 20 January 2026 (closing high $73.91 on 16 January), then declined for six and a half months to an intraday low of $24.11 on 3 August 2026 and a closing low of $24.91 on 31 July. That is a peak-to-trough decline of 69.1%. At the low the shares traded 3.6% below their $25.00 IPO price while revenue had more than doubled year over year — the precise signature of a Pattern D narrative collapse: a short-seller report in April, a blockchain premium unwinding, analyst price-target cuts through July, and an 8.5% debt raise landing into a hostile tape. Pattern E contributed the initial break, with a 25.7% single-day fall on 27 February after a fourth-quarter earnings miss.
The critical assessment is what has happened since. From $24.11 on 3 August the stock reached $38.02 by 27 August — a 57.7% gain in eighteen trading sessions, the best month in its history. The June-quarter print on 13 August initially gapped up and fully reversed, and the market only re-rated it four sessions later. Short interest stands at 12.84 million shares, 9.21% of float and 3.33 days to cover, and rose into the rally — so a meaningful share of the move is short covering rather than re-rating. Price now sits 23.4% above the 50-day average of $30.80 and just above the 200-day of $36.97, but the 50-day remains far below the 200-day: the trend structure has not repaired, the price has simply run ahead of both averages.
So the current level, 51.3% below the January high, reads as a dip only against the peak. Against the stock's own recent path it is an extended bounce off a capitulation low that the framework would have wanted to buy three weeks ago at $25. The discipline that applies here is the same one that says buy into dislocation, applied in the other direction: the dislocation has been substantially removed, and chasing a 58% eighteen-session move on a name with two-sided rate exposure and an unresolved short thesis is FOMO wearing a drawdown costume. The setup is worth waiting to re-form — a rate scare, a Kiavi integration stumble, a take-rate print below 3.5%, or a credit-quality disclosure would each plausibly return the stock toward the mid-twenties, where the ten-year math becomes genuinely compelling rather than merely adequate.
Monopoly potential scores 7.5/10. The marketplace flywheel is real and has visibly ignited — 489 origination partners from 307 in six months, third-party volume from roughly nothing in 2024 to 65% of a $4.26 billion quarter, AAA ratings from both S&P and Moody's on blockchain-collateralised paper — and the regulated-industry anchor makes the business structurally resistant to AI-agent intermediation. It falls short of eight because the underlying market is mature cyclical credit rather than a new one, and because the blockchain premium that supports the multiple is genuinely contested rather than settled.
Founder leadership scores 6.7/10 and financials and entry score 6.0/10. Leadership is a split picture: an eight-year missionary arc, exceptional execution velocity, and a founder with 67–68% voting control, set against roughly $136 million of combined founder and CEO selling in the first year post-IPO, a $169 million retention award whose vesting cliff was removed two months after grant, and a three-person C-suite with no COO or CTO running a lender, a blockchain, a broker-dealer and a $717 million acquisition. On entry, the business is genuinely profitable — roughly $237 million of trailing net income at a 54.6% adjusted EBITDA margin, which is about 36x earnings — but 12x trailing sales is more than double the framework's entry discipline, and the earnings carry non-cash servicing marks worth 12.9% of quarterly revenue.
The result is a better business than its critics allow, at a worse price than its 51% drawdown from the January high suggests. The stock bottomed at $24.11 on 3 August — below its IPO price, with revenue having doubled — and has since squeezed 57.7% in eighteen sessions on 9.2% short-of-float. That was the entry; this is the bounce. The ten-year math returns roughly 8x from $38.02 and above 12x from the mid-twenties, and the difference is entirely the entry dial. WATCHLIST: revisit for entry on a return toward $25-28 — trailing P/S near 8 — driven by a rate scare, a take-rate print below 3.5%, a Kiavi integration stumble, or a credit-quality disclosure. Position sizing should stay modest even then, given the two-sided rate exposure and the unresolved question of how much of the blockchain layer is load-bearing infrastructure versus packaging on a very good, very cheap conventional lender.
Not financial advice
The analyses published on Triportfolio are for informational and educational purposes only. Nothing on this site constitutes financial advice, investment advice, trading advice, or a recommendation to buy or sell any security. Triportfolio is not a licensed financial advisor, broker, or investment professional.
All investment analysis reflects the personal views and independent research of the author at the time of publication. Markets change rapidly and analyses may become outdated. Past performance of any security discussed is not indicative of future results.
Investing in equities — particularly early and mid-stage growth companies — involves significant risk, including the possible loss of the entire amount invested. The companies discussed on this site are typically high-volatility, high-risk investments that may not be suitable for all investors.
Before making any investment decision, you should conduct your own research and consult a qualified financial professional who understands your personal financial situation, risk tolerance, and investment objectives.