Author: Mario Stefanidis, Head of Research at Artemis Analytics; Source: Artemis; Compilation: Shaw Golden Finance

Introduction

According to data from the Institute of International Finance (IIF), by the end of 2025, the global debt market hit a historical high of $348 trillion. Of this, government debt is about $107 trillion, corporate debt $101 trillion, household debt $65 trillion, and debt from the financial sector $76 trillion. Digital and financial technology lending platforms account for between $590 billion and $680 billion of total debt—equivalent to less than 0.2%.

This—by far—the largest credit market in human history is still operating on infrastructure designed decades ago (FICO launched in 1989, MERS went live in 1995). According to data from the American Bankers Association of mortgage lenders, the average origination cost of a single mortgage in the U.S. is about $11k. Despite massive technological progress and the widespread adoption of artificial intelligence, this cost is still double what it was in the early 2010s.

Source: Freddie Mac

Clearing and settlement for standard wire transfers still takes about 28 hours, while most banks’ credit approval decisions still rely on a committee process, depending on black-box scoring models built from 20 to 30 variables. All of these are established public facts, but what is less obvious is exactly how the solution is being rolled out.

The credit industry is not being reshaped through a Silicon Valley-style romantic disruption model—no startup can replace systemically important global banks like JPMorgan Chase in one sweep. Real change is more subtle and more structural: the credit end-to-end system that used to be vertically integrated within banks—loan origination, distribution, risk reviews, capital provision, and even the underlying infrastructure— is being dismantled into a horizontal, modular architecture, with each link controlled by specialized institutions.

This architectural shift mirrors what happened in cloud computing, moving from monolithic systems to microservices, and in the media industry, shifting from the studio/factory model to streaming and creator ecosystems. Now, this transformation has finally arrived in the credit space.

In this wave of reintegration, the winners are not the institutions with the largest balance sheets, but core-layer companies in key chokepoints that other participants cannot bypass. Two positions matter far more than the rest: first, the intelligent decision layer, where AI risk reviews and risk scoring determine where capital flows and under what underwriting terms; second, the clearing and settlement channel layer, where blockchain infrastructure is dramatically compressing loan origination costs and settlement time by orders of magnitude.

As long as you occupy these two types of “water-seller” core positions, other lenders will pay you a usage fee. If neither is held, then you can only engage in price competition in a commoditized market, where $3.5 trillion of private credit capital is already chasing yield.

Here, Artemis maps 40 companies across 15 sub-sectors and groups them into five major tiers to analyze where structural value is consolidating.

Five tiers of the new credit architecture

First tier: Loan origination

The loan origination layer is the source of credit business, covering consumer loans, mortgages, small business loans, and crypto-collateralized loans, among other categories. This area is also becoming increasingly commoditized. Having the ability to originate loans is no longer a competitive moat; it is merely the basic threshold for entry. The key differentiator between winners and other participants lies in loan origination costs and approval pass rates.

SoFi, with an estimated valuation of about $24 billion, and the rocket company Rocket (rocket mortgage), with a market cap of $48 billion, both have large-scale loan origination operations, but the core of their profit logic is how to make loan funding at lower cost. Figure, with a market cap of $6 billion, leverages its Provenance blockchain-native origination of home equity lines of credit (HELOCs) and first-lien mortgages, removing the multi-layer intermediary stages that make traditional mortgage funding slow and expensive.

In the crypto space, Aave (market cap $2.7 billion) and MakerDAO/Sky (market cap $1.6 billion) have completely blurred the boundary between financial technology and decentralized finance (DeFi) in the loan origination layer.

Second tier: Channel distribution

The distribution layer is where demand aggregation happens, and embedded finance and “buy now, pay later” (BNPL) models are reshaping this space. The embedded finance market is expected to grow from $156 billion in 2026 to $454 billion by 2031, a CAGR of 24%. The BNPL model is expected to cover 13% of digital transactions, up sharply from 6% in 2021.

Affirm (market cap $15 billion) and Klarna (market cap $5 billion) are well-known players, but the real structural trend is that credit services have been deeply embedded into checkout flows, software platforms, and merchants’ customer experiences. Even though both companies’ stock prices have fallen substantially from historical highs, they are not “water-seller” type businesses that win mainstream market share. Lenders that borrowers don’t perceive are often the real winners.

Today, major software companies are adding financial products. Shopify, Amazon, Square, and Stripe all need API infrastructure layers, and institutions providing these services will extract fees from the size of each added transaction.

Third tier: Risk underwriting review and risk pricing

This is the first core layer in the entire credit architecture. The institution that controls borrowers’ credit scoring controls how value is distributed across the entire credit industry.

Today, the credit bureaus market is dominated by three major incumbents forming an oligopoly: Experian, TransUnion, and Equifax. Together, they generate about $18 billion in annual revenue by scoring borrowers using 20–30 variables.

AI risk models can evaluate more than 1,600 variables (data from Upstart). Upstart’s published data also shows that, while maintaining the same bad-loan rate as traditional models, its approval volume increases by 44%, default rate falls by 53%, and its annualized interest rate (APR) drops by 36%. With mortgage interest rates surging to nearly 7% today, each basis point is critical for first-time homebuyers.

Upstart currently achieves 92% of lending decisions through full automation, completing approvals within minutes, whereas traditional risk underwriting reviews take 3 to 5 days. The U.S. Consumer Financial Protection Bureau (CFPB) is pushing for alternative FICO models with lower discrimination. The EU’s “Artificial Intelligence Act” also classifies credit scoring as a high-risk scenario, requiring explainability. These regulatory developments favor explainable machine learning models and give them an advantage over traditional credit bureaus that rely on black-box models.

The value of this tier is extremely high because whoever owns the scoring engine controls the entire revenue curve up the stack. But at the same time, the moat in this space still needs ongoing verification—rapid AI progress means that, as long as there are enough resources and time, “any institution” can build scoring models.

Fourth tier: Capital and funding supply

In the post-pandemic era, capital overall is abundant. Even though the current environment is challenging, the scale of private credit management has expanded to $3.5 trillion, and Morgan Stanley expects it to reach $5 trillion by 2029. The total value locked (TVL) in decentralized finance (DeFi) lending protocols ranges from $5 billion to $78 billion, accounting for about half of all DeFi activity. The size of non-transactional perpetual assets (NPE) has grown from zero growth in 2021 to over $200 billion.

In an age of abundant capital, the most core capability is the intelligent allocation of capital flows. Therefore, even if the capital layer is enormous in volume, its structural position still remains subordinate to the intelligent decision layer above it and the infrastructure layer below it.

Private credit institutions such as Ares, Blue Owl, and Golub are important allocators of capital, but they heavily rely on upstream scoring systems and downstream clearing channels to execute efficient loan funding. In the DeFi space, Ape has an absolute dominant position in liquidity, accounting for more than half of lending scale; protocols such as Maker, Morpho, Maple, and Kamino compete for the remaining market share.

Fifth tier: Infrastructure

Infrastructure is the second core layer in the entire architecture. Whoever holds a financial license or a clearing and settlement channel, everyone has to pay “tolls” to them. According to management disclosures, the bank license held by SoFi reduces its cost of funds by 170 basis points, and its annualized interest expense is reduced by more than $500 million. Figure, leveraging its Provenance blockchain, has processed over $50 billion in total transaction value; its cost to originate a single loan is below $1,000, while the average cost of traditional channels is about $11k. Final settlement confirmation on blockchain requires only seconds, whereas traditional wire transfers take about 28 hours.

SoFi’s Galileo and Technisys technology stacks, as well as platforms like Blend Labs, form the supporting base technologies for the remaining loan-as-a-service (LaaS). Cross River Bank, as a hidden partner bank behind dozens of fintech companies, has issued more than 96 million loans totaling over $140 billion through its collaborations.

Companies that can win long term will either occupy a key chokepoint and become indispensable to all participants, or vertically integrate multiple tiers to form compounded competitive advantages. Companies that lose will get trapped in commoditized business layers, lacking structural voice, and will have to compete on price until profits approach zero.

Winners: core-layer holders and multi-tier compounded-advantage enterprises

SoFi: an all-stack compounded tool

SoFi is the only company covering four of the five tiers:

• Directly origination consumer loans and mortgage loans.

• Through the Galileo platform, it provides lending infrastructure to third parties, supporting about 160 million activated accounts.

• Leveraging its in-house risk underwriting models to conduct loan reviews, with core evaluation dimensions being repayment willingness, repayment ability, and stability.

• Holds a bank license and, in the infrastructure layer, has the core banking technology system of Galileo and Technisys.

SoFi’s 2025 revenue hit a record $3.6 billion, up 38% year over year. The platform has 13.7 million members and a financial products scale of 20.2 million. Management guided that 2026 revenue will reach $4.7 billion, with EBITDA of $1.6 billion. This business not only has strong revenue growth, but its profitability is excellent as well, with a profit margin of 34%. Just the bank license alone enables SoFi to finance loans via deposits rather than wholesale markets, directly lowering its cost of funds by 170 basis points.

SoFi is building the “Amazon Web Services (AWS)” for lending—a platform that both competes with other lenders and empowers them. Galileo itself has already been built into a $3.48M-plus revenue engine. Technisys, acquired for $1.1 billion in 2022, provides core banking system layers for third parties. A bank license forms a structural moat that most fintech lenders cannot replicate; even though many in the industry are trying to imitate it, the U.S. Office of the Comptroller of the Currency (OCC) received 14 applications for new bank charters in 2025 alone in a single year, signaling that the race for the infrastructure tier is accelerating.

Upstart and Pagaya: intelligent decision layer

It’s somewhat ironic that winning in lending may not require doing the lending business itself. Upstart and Pagaya both center on a risk underwriting engine whose risk performance is better than lenders’ proprietary models, without needing to rely on their own balance sheets to operate. This is exactly the “water-seller” logic taking shape in the credit decision domain.

Compared with traditional FICO-based risk models, Upstart’s model can approve 44% more borrowers at the same bad-loan rate, while reducing the default rate by 53%, and it also provides borrowers with a significantly lower annualized interest rate. Currently, nearly all new loans initiated on its platform are fully automated, sharply reducing the need for human intervention. This is fundamentally different from traditional consumer credit risk underwriting models.

Pagaya operates on the same track but faces tougher market realities. The company does not directly fund loans; instead, it authorizes banks to use its AI risk underwriting engine. Since its founding in 2016, Pagaya has supported 31 partner banks to evaluate approximately $2.6 trillion in loan applications in total. Its structural positioning is very clear: it doesn’t need borrowers to know the brand—banks only need to rely on its scoring system. But the market currently does not validate this logic. In the fourth quarter of 2025, online business volume grew only 3% year over year; revenue missed market consensus expectations, and forward-looking performance guidance was also below expectations. Its stock price plunged on a single day by nearly one quarter. The value of the intelligent decision layer is entirely constrained by the credit cycle—when the bad-loan rate rises across the partner network, even excellent AI cannot withstand the pressure from deteriorating asset quality.

But the core logic remains valid: FICO builds a single cross-sectional score based on only a small set of historical variables, and as consumers’ financial situations become more complex and diverse, AI risk underwriting systems become even more critical. Unlike FICO, these systems continue to learn and improve after every scoring run.

Figure: the next-generation clearing and settlement channel

Origination cost for a single loan using traditional channels and the Mortgage Electronic Registration System (MERS) is $11,000, while using Figure’s technology stack—including the Provenance blockchain and the DART system—can reduce the cost to $717. This new channel infrastructure enables lending costs to drop by an order of magnitude.

Figure has used the Provenance blockchain to originate over $21 billion in home equity products (primarily home equity lines of credit), and has processed more than $50 billion in cumulative on-chain transaction volume. In the fourth quarter of 2025, loan origination reached $2.7 billion, up 131% year over year. The company holds over 180 lending licenses and has U.S. SEC broker-dealer registration qualifications, providing a compliance foundation for scalable operations. It also has over 300 white-label lending partners; since filing its S-1 for an IPO last September, it has added partners at a pace of about 1 per day. Revenue has grown from $28.5 million quarter annualized in the first quarter of 2023 to $146.8 million today.

Figure’s core business is not strongly tied to crypto assets, but its stock price trend is highly similar to Bitcoin. The company’s settlement system reflects the logic of cost-structure reconfiguration: final settlement confirmation takes only seconds, whereas traditional methods take over a day; loan origination cost is only a fraction of the traditional approach. Across the entire loan lifecycle, costs related to asset securitization are saved by more than 100 basis points—within the $3 trillion annual asset securitization market, this implies potential cost reductions of more than $30 billion.

Aave: the core controller in the DeFi space

Aave holds more than half of the DeFi lending market share. Liquidity breeds more liquidity; borrowers continuously cluster toward platforms with the deepest liquidity pools (network effects). Its cumulative loan origination volume has surpassed $1 trillion, and the protocol officially crossed the $1 trillion milestone in total loans just last month.

Beyond its dominance in DeFi, Aave is structurally most interesting in its institutional lending business line Horizon. Horizon has attracted $580 million in deposits, aiming to exceed $1 billion by 2026. It acts as a bridge connecting DeFi liquidity with traditional credit demand. If Aave can bring on-chain funds into institutional-grade lending products, it would become the capital supply layer for traditional lenders, unlocking a potential total market far larger than the retail-focused DeFi market (TAM).

DeFi lending also has one structural risk advantage that is often underestimated. In DeFi, the overcollateralization ratios are typically between 150%–180%, while in traditional peer-to-peer lending they are only 50%–70%. Bad loans in DeFi mainly stem from oracle failures or technical malfunctions, rather than creditworthiness defaults.

Affirm: channel lock-in

Affirm has a leading position in the buy now, pay later (BNPL) space by deeply embedding merchant payment settlement infrastructure. Critics focus on its consumer credit risk, but they overlook the core structural logic: Affirm is not a consumer lending company in the traditional sense; it is a credit distribution channel for sales terminals. Its moat is the systems integration with merchants. Given that BNPL is expected to cover 13% of all digital transactions, large-scale platforms that embed into checkout flows will charge a structural “channel fee” directly from the commercial transactions themselves.

The losing pattern: four structural failure modes

We intentionally do not name the companies that fit these patterns. If you are an investor or operator in the credit space, you naturally know who they are. More important than specific names is understanding why these structural positions are destined to fail—because the same patterns will create new victims in the next cycle as well.

Lenders that only focus on their balance sheets

Their only competitive advantage is access to capital. They originate loans using traditional risk models, fund them with their own balance sheet assets, and have no dedicated technology layer. They are merely “mindless pipelines” for capital.

In a world where private credit management scale has reached $3.5 trillion and is moving toward $5 trillion, capital is not scarce; what is scarce is intelligent decision-making and infrastructure. These enterprises can only rely on price competition, compressing profits to zero in each interest-rate cycle and forcing them to take on excessively high risk. In the end, these lenders will extend credit to high-risk borrowers, and when the cycle turns, they will suffer losses.

Most of these participants are traditional consumer lending institutions, small-scale banks, and fintech lenders that have never built a technology moat beyond their initial loan products. When capital becomes commoditized, and with no technology advantages and lending only off their own balance sheets, it is essentially a slow transfer of shareholders’ equity to borrowers.

CeFi lending victims

The centralized crypto lending (CeFi) platforms that collapsed in 2022 were not casualties of the bear market. They failed under the oldest lending failure mode in the industry: maturity mismatch, misappropriation of customer funds, lending against illiquid assets, and a lack of transparent risk management.

Decentralized finance (DeFi) protocols that automatically enforce collateral discipline through smart contracts, with on-chain collateralization ratios that are publicly visible, did not blow up. What really went wrong were those CeFi platforms that rely on human judgment and have opaque balance sheets. Any lending platform—whether in crypto or traditional finance—if it only asks you to believe its balance sheet but does not show you the collateral, is simply repeating the same structural old road that has already failed.

Ghost protocols

There is a class of DeFi lending protocols that is still alive technically but structurally dead. After going live, they attracted initial locked liquidity through token incentives, but once incentives faded, they stalled. The code can run and TVL is not zero, but usage curves have leveled off or continued to decline, and there is no clear path for natural demand growth.

The reason is that DeFi lending shows extreme power-law distribution characteristics: liquidity concentrates on platforms that have network effects. Aave dominating an absolute share of the market is proof. Protocols that cannot break through critical scale fall into a structural “no-man’s-land”: too small to attract natural liquidity and supporting integrations; not small enough to shut down decently. As profit-seeking capital flows toward top platforms, their locked value continues to bleed slowly and the process is irreversible. These are zombie protocols maintained only with the sunk costs of governance tokens.

Lenders that missed the platformization transition

Some companies built strong loan origination capabilities in the last cycle but never developed platform capabilities. They have no API distribution channels, no embedded finance partnerships, and no technology licensing model. They can originate loans very well, but they cannot export that capability externally.

As the credit industry moves toward modularization, whether you can become a component within someone else’s system is just as important as directly originating loans. Enterprises that can only lend directly to end borrowers will have growth constrained by the coverage of their own channels. In contrast, enterprises that can provide lending capability support to other institutions have an unlimited potential market space (TAM). Pure loan originators usually have good unit economics on a per-customer basis, but their growth curves are flat because their addressable market is limited to their own brand and channels. In modular architectures, being an excellent lender is a necessary condition, but being the kind of lender that other lenders can plug into is the real winning position.

Worth watching

The winners above have become market consensus or near-consensus, while the following companies have not. They have structural traits to become core-layer controllers, but have not yet been validated at the scale level. These are the targets worth continuous tracking.

Morpho

Morpho’s total value locked (TVL) is already $6.6 billion, up 164%, and its market cap exceeds $800 million. Its structural logic is completely different from Aave’s: Aave is like a commercial bank within decentralized finance (using a unified lending liquidity pool model), while Morpho is building a modular lending layer that allows institutional participants to customize proprietary lending markets based on their own risk parameters, collateral types, and interest-rate models. If the lending system truly moves toward modularization, Morpho will become a loan-as-a-service protocol at the on-chain layer.

Maple Finance

In 2025, Maple originated total loans of $11.3 billion, serving 65 active borrowers. Assets under management (AUM) grew sharply from $516 million to $4.6 billion, a 767% increase. The company targets $100 million in annual recurring revenue (ARR) in 2026. Maple is one of the few protocols genuinely dedicated to bringing real-world business lending to blockchain infrastructure—connecting institutional credit demand with on-chain capital and settlement systems to deliver the business. Its explosive AUM growth indicates that institutional interest in the on-chain credit market is shifting from theoretical concepts to real execution.

Cross River Bank

Since 2008, Cross River has issued more than 96 million loans through partnerships, totaling over $140 billion. It is the partner bank behind Affirm, Upstart, and dozens of other fintech lenders. Reports say the bank is preparing for an IPO. Cross River is a “hidden bank,” supporting a substantial portion of fintech lending activity as an infrastructure-layer provider. As the partner-bank model matures, the leverage its market position brings is something no single fintech lender can replicate. Cross River’s winning key is making it so fintech companies can’t conduct lending without its support.

License battle

The U.S. Office of the Comptroller of the Currency (OCC) received 14 applications for new bank charters in 2025 alone—almost equal to the total of the prior four years. The total number of charter applications submitted by fintech institutions has also hit a record high of 20. Affirm, Stripe, and Nubank are all actively applying for licenses. These companies view charters as the core competitive advantage that will determine the endgame of credit business rebuilding.

Companies that started as technology service providers are now capturing economic value across the entire industry chain by obtaining regulatory qualifications. In lending, a bank charter’s position is comparable to regional nodes in cloud computing, because:

• Building costs are extremely high;

• Industry participants cannot bypass it;

• Once obtained, it forms a permanent structural advantage.

The business logic is very clear: optimizing funding costs by 1 basis point can improve pre-tax net asset returns by several percentage points. For scaled enterprises, the advantages brought by charters are very significant. But for small and mid-sized institutions, a charter may become a trap: they bear all compliance costs, regulatory exam pressure, and capital requirements, yet lack sufficient operating scale to cover these expenses. Only enterprises that already have large business volumes can make charters into an accelerator of growth.

Credit architecture in 2030

If there is one core analytical framework to remember from this article, it is the following three questions. They apply to all lending enterprises, whether public or private, and whether on-chain or off-chain.

First: Which tier does the company occupy? Loan origination and commoditized capital supply are red-ocean tracks, where profit margins will continue to compress with industry cycles. AI risk underwriting, blockchain settlement, and bank licenses are core chokepoints, where value can compound over time. If a company is trapped in a red-ocean track and cannot enter core tiers, no matter how strong the team is, its long-term profitability will be steadily eroded.

Second: Is it a platform or a single product? A single product serving end borrowers scales linearly with the reach of its own channels; a platform empowers other lenders, and growth depends on the ecosystem’s total size rather than only its own business. SoFi has both attributes, while Pagaya is a pure platform-type business. For enterprises that only lend directly to their own customers, growth will hit a ceiling, whereas platform-type businesses are not subject to that limitation.

Third: Does it have a regulatory moat? This includes bank charters, 180 lending charters across states, or programmed compliance achieved through smart contracts. In the lending industry, regulation is not an extra cost—it is core infrastructure. Companies that recognize this early will build advantages that competitors would need years and large amounts of capital to catch up with.

By 2030, the credit industry will no longer look like traditional banking; it will be more like the cloud computing industry. A small number of full-stack platforms will cover multiple tiers, creating compounded advantages across each layer. The most typical representative in traditional finance is SoFi, while in the on-chain space it’s Aave. Around these core platforms, many specialized tier service providers will connect via APIs and on-chain channels, focusing deeply on specific functions and charging service fees.

In the global $348 trillion debt market, fintech penetration is still below 0.2%. This market is not for hundreds or thousands of lending institutions to divide up—it will be dominated by a dozen-plus platforms, becoming the foundational support of the entire industry.