The residential mortgage market is the largest asset class in American finance. At $13.1 trillion in outstanding balances as of Q4 2025, it dwarfs corporate bond markets and rivals the entire US equity market in aggregate value. For forty years, the dominant securitization architecture — pooling mortgages into agency or non-agency MBS, selling tranches to institutional investors — has remained structurally unchanged since Salomon Brothers pioneered the first mortgage-backed security in 1983. Tokenization represents the first serious architectural challenge to that model, and Figure Technologies has spent six years building the proof of concept.
Figure’s Operational Blueprint
Figure Technologies, founded in 2018 by SoFi co-founder Mike Cagney, built its HELOC product on the Provenance Blockchain from inception — not as a retrofit, but as a native architecture. The result is measurable. Traditional HELOC origination carries fully-loaded costs of $1,200 to $1,800 per loan, reflecting manual title searches, paper-based underwriting packages, wet signatures, and multi-party warehouse reconciliation. Figure’s blockchain-native process has driven that cost to approximately $30 per loan by automating title verification, replacing wet signatures with cryptographic attestation, and settling warehouse lines on-chain within hours rather than days.
By February 2026, Figure has originated more than $15 billion in HELOCs, the majority of which have been funded through warehouse credit facilities that themselves settle on Provenance. JPMorgan extended a $4 billion credit facility backed by Figure-originated loans — the largest single institutional commitment to tokenized residential mortgage collateral by any bulge-bracket bank. That facility is operationally significant: JPMorgan’s participation required internal credit and legal sign-off on Provenance’s smart contract architecture, on the legal standing of digital promissory notes under UCC Article 3 and Article 9, and on the custody model for tokenized loan collateral. The bank’s willingness to extend $4 billion against those structures is the most credible institutional validation the tokenized mortgage market has yet received.
The 1990s MBS Parallel — and Why It Breaks Down
The parallels to the 1990s mortgage securitization revolution are instructive, but the differences matter more. When Freddie Mac issued its first participation certificate in 1971 and Salomon Brothers sold the first private-label MBS in 1977, the innovation was economic: pooling individual mortgage risk to create diversified instruments that institutional investors could price and trade. The operational infrastructure — document custodians, servicers, master servicers, trustees, rating agencies — was built atop those economics over the following decade. By 1993, non-agency MBS issuance exceeded $500 billion annually.
Tokenized mortgage securitization inverts this sequence. The operational efficiency gains come first — the $30 origination cost, the near-real-time warehouse settlement, the automated collateral tracking. The secondary market liquidity that would make tokenized MBS a direct competitor to agency paper does not yet exist. A Figure HELOC that settles on Provenance has a demonstrably lower cost structure than an equivalent traditional HELOC, but a pool of 5,000 Figure HELOCs does not yet trade with the same market depth as a $500 million Fannie Mae certificate.
That gap is where the next phase of residential mortgage tokenization will be decided.
The Fannie/Freddie Question
The most consequential question in residential mortgage tokenization is also the one least discussed outside agency circles: what would it take to tokenize Fannie Mae or Freddie Mac pools?
The two government-sponsored enterprises collectively guarantee approximately $7.5 trillion in mortgage-backed securities. Fannie and Freddie MBS are the benchmark instruments for the $13 trillion residential mortgage market — the securities against which all other mortgage instruments are priced. If the agencies moved to blockchain-based settlement, the secondary market infrastructure would follow, and the efficiency gains would scale to the entire conforming mortgage market.
The structural requirements are achievable. Fannie and Freddie already operate a common securitization platform — the CSP, launched through Common Securitization Solutions in 2019 — that handles data standardization and disclosure for both agencies’ MBS issuance. Tokenizing CSP outputs would require converting CUSIP-identified certificates into on-chain instruments with equivalent legal standing, a process that the UCC Article 12 amendments (enacted in most states by 2024) have substantially clarified.
The regulatory gap sits at the Federal Housing Finance Agency. FHFA, which has supervised both enterprises under conservatorship since 2008, has not issued guidance on blockchain-based mortgage instruments. The agency’s 2023 and 2024 supervisory priorities made no reference to distributed ledger technology. Conservatorship creates a specific constraint: changes to Fannie and Freddie’s operational infrastructure at the scale of a CSP overhaul require FHFA pre-approval and, for changes affecting MBS investor rights, could require congressional action. The FHFA regulatory gap is not a technical barrier — it is a political and institutional one that no private-sector actor can resolve unilaterally.
Comparing Cost Structures
| Item | Traditional HELOC | Figure Blockchain HELOC |
|---|---|---|
| Origination cost per loan | $1,200–$1,800 | ~$30 |
| Title search method | Manual / third-party | Automated on-chain |
| Warehouse line settlement | T+2 to T+5 | Same-day |
| Document custody | Physical / PDF escrow | Smart contract |
| Investor reporting | Monthly servicer reports | Real-time on-chain |
| Secondary market liquidity | Limited (whole loan market) | Emerging (Provenance DEX) |
The settlement speed differential is operationally material for warehouse lenders. A warehouse line that turns over in hours rather than days requires less capital to support the same origination volume. For a lender originating $1 billion per month in HELOCs, reducing warehouse line duration from four days to four hours frees approximately $130 million in capital. At current warehouse rates of 6.5 to 7.5 percent, that represents $8 to $10 million in annual interest savings independent of any origination cost reduction.
Non-QM and Jumbo: The Near-Term Opportunity
The near-term tokenization opportunity in residential mortgages lies not in agency paper but in non-QM and jumbo origination. Non-qualified mortgages — loans that do not meet the CFPB’s Ability-to-Repay qualified mortgage safe harbor — carry higher origination costs, shorter warehouse periods, and less liquid secondary markets than conforming loans. Jumbo loans above the conforming limit ($806,500 in most markets for 2026) similarly lack the GSE backstop that gives agency MBS their structural advantage.
Both segments are natural candidates for blockchain-based origination and securitization. Non-QM and jumbo borrowers are typically higher-net-worth individuals with more complex income documentation, making automated underwriting less applicable but blockchain-based document management more valuable. Non-QM securitization volumes exceeded $40 billion in 2024; institutional appetite for yield-premium instruments in a rate environment that has kept 30-year fixed rates above 6.5 percent is substantial.
Securitize and Figure are both active in exploring non-QM securitization on blockchain rails. The legal structure — a Delaware statutory trust or special purpose corporation issuing digital notes backed by non-QM pools — is structurally equivalent to traditional non-agency securitization and does not require new regulatory frameworks. The question is whether institutional buyers will accept tokenized note certificates as equivalent to traditional CUSIP-identified securities for purposes of their own investment mandates and custody arrangements.
What Institutional Adoption Actually Requires
JPMorgan’s $4 billion facility is the headline, but the structural prerequisites for broad institutional adoption of tokenized mortgage instruments are more demanding. Investment-grade insurance companies operating under NAIC statutory accounting rules can only hold residential mortgage instruments that qualify as admitted assets. Admitted asset status for tokenized mortgage certificates requires state insurance commissioner approval in each state where the insurer operates — fifty separate regulatory determinations, each of which may require legislative action.
ERISA-covered pension funds face analogous constraints under the Department of Labor’s prudent investor standard. While no DOL guidance explicitly prohibits tokenized mortgage holdings, plan fiduciaries bear personal liability for investment decisions that cannot be defended as commercially reasonable. Until a critical mass of large plans hold tokenized mortgage instruments — creating a track record and a peer-group benchmark — plan sponsors face asymmetric liability: limited upside from operational efficiencies relative to the downside of being the test case in a DOL enforcement action.
The tokenized real estate market will not reach institutional scale in residential mortgages through technology alone. It will reach scale when custody banks offer tokenized mortgage certificates as standard admitted assets in their prime custody platforms, when NAIC develops a valuation methodology for tokenized MBS, and when the first ERISA plan successfully defends a tokenized mortgage holding in a DOL examination. Figure’s $15 billion is proof that the economics work. The remaining distance is institutional infrastructure, not technology.
Looking Forward
The FHFA question will be resolved by one of two pathways: either the agencies exit conservatorship under legislation that explicitly authorizes blockchain-based securitization infrastructure, or a non-agency tokenized MBS market grows large enough to force the agencies to respond competitively. The second pathway is slower but does not require congressional action. Given the current legislative environment, it is probably the more realistic of the two.
Figure’s immediate trajectory points toward expanding its warehouse credit facilities — the company has disclosed conversations with multiple additional bulge-bracket lenders beyond JPMorgan — and toward launching a secondary trading platform for Provenance-based HELOC certificates. If that platform achieves sufficient liquidity to support institutional portfolio management, it will demonstrate something the 1990s MBS revolution could not: that a tokenized secondary market can function for consumer mortgage instruments within a regulatory framework that did not anticipate blockchain technology.
That demonstration would change the analytical baseline for every other real-world asset tokenization initiative in the residential mortgage space.