In a report on digital finance, IOSG author Benji Siem says a structural shift is taking shape across fintech: major companies are embedding blockchain infrastructure, stablecoins and decentralized finance rails directly into their core products. The report opens with several examples that, in the author’s view, point in the same direction: Stripe’s $1.1 billion acquisition of Bridge, Mastercard’s roughly $2 billion acquisition of Zerohash, and Robinhood’s launch of its own Layer 2. The conclusion is straightforward. Crypto is moving out of the storefront and into the backend.
The paper reviews 10 leading fintech companies, focusing on their business models, revenue drivers, and the ways they are integrating crypto payments and DeFi infrastructure. Its central argument is that the most successful firms are not treating crypto as a speculative asset class. They are using it as backend infrastructure that can improve settlement speed, cut costs and widen global financial connectivity. Stablecoins, in particular, are described as the bridge between traditional financial systems and on-chain markets.
A shared view of digital finance, but different ways to pursue it
The report summarizes the common foundation across the 10 companies in one line: financial services should be borderless, real-time, software-defined and composable, with compliance largely invisible to the end user.
From there, it breaks the market into several groups.
Infrastructure players
Visa, Mastercard, Stripe and Adyen are grouped as infrastructure companies. Their basic view is that the opportunity lies in rebuilding the pipes that move money, without necessarily owning the customer relationship. Every new payment rail, whether stablecoins, account-to-account transfers or instant payments, expands the addressable market. Crypto enters the picture mainly through stablecoins, which can reduce settlement friction and support 24/7 treasury management.
Consumer platforms
Nu, Revolut, PayPal and Cash App are described as consumer platforms. Their strategy is to become the primary financial entry point for users, then cross-sell banking, payments, investing and crypto inside a single app to raise lifetime value. In this model, crypto is less a standalone thesis than a layer that can lift engagement and monetization through trading fees, yield products and cross-border use cases.
Hybrid players
Robinhood, Block and SoFi are placed in a hybrid category. According to the report, these companies are trying to integrate vertically across both consumer products and infrastructure. The goal is to capture profit at multiple layers, from retail engagement to B2B infrastructure and custody. Their crypto strategy is typically two-sided: distribution on the consumer side, ownership on the infrastructure side.
The main patterns the report identifies
Infrastructure first, speculation second
The report says nearly all of the companies studied treat crypto as backend infrastructure rather than a front-end speculative product. Visa, Mastercard, Stripe and Adyen are upgrading settlement layers with stablecoins while keeping the consumer experience largely unchanged. In the author’s framing, crypto works best when it is invisible, and that is the most consistent pattern across all 10 companies.
Stablecoins as the bridge asset
Every company that touches crypto is, in one way or another, betting on stablecoins as the transition layer between traditional finance and crypto, the report says. It lists several examples:
- Visa: USDC settlement on Solana and more than 130 stablecoin card programs.
- Mastercard: support across USDC, PYUSD, USDG and FIUSD on multiple chains.
- Robinhood: a partnership with USDG and shared economics.
- PayPal: PYUSD issuance that internalizes part of the settlement stack.
- Stripe: USDC for cross-border merchant payouts.
The report argues that stablecoins can compress settlement times, reduce foreign-exchange friction and enable programmable treasury management without adding volatility exposure.
Regulated distribution as the moat
Companies with large user bases are positioning themselves as compliant access points rather than protocol builders, the report says. It cites PayPal’s 400 million users, Revolut’s more than 50 million, Nu’s 122 million and Cash App’s 58 million. Their edge, in this view, is not raw technical novelty. It is trust, compliance and scaled distribution.
DeFi as wholesale infrastructure, not a retail interface
The report says none of the companies directly exposes native DeFi protocols to retail users. Instead, DeFi appears as a wholesale backend layer: a source of yield through tokenized Treasuries and money markets, a tool for liquidity optimization through faster settlement and cheaper cross-border transfers, and a way to package regulated savings products backed by DeFi returns. In other words, DeFi supports regulated products behind the scenes rather than serving as the user-facing experience.
The rise of multi-rail strategy
The paper describes a broader move toward payment-agnostic infrastructure. Stripe’s approach is summarized as owning the programmable money layer regardless of which rail is used. Mastercard is described as a multi-rail company spanning cards, A2A, real-time payments and blockchain. Adyen is framed as a global operating system for enterprise payments. As cards, A2A, stablecoins and BNPL continue to fragment the payments market, the likely winners are the firms that can route intelligently across all of them.
Where the strategies converge
The report distills the winning formula into a simple idea: build the compliant wrapper around programmable money, then capture value through distribution, trust and compliance rather than through protocol ownership or speculative exposure. The companies with the strongest position, it argues, tend to have at least one of three advantages: scaled distribution, as seen with Nu, PayPal, Revolut and Cash App; infrastructure control, as seen with Visa, Stripe and Mastercard; or vertical integration, as seen with Robinhood, Block and SoFi.
Which business models scale, and which ones look fragile
Stronger, more scalable models
The report first highlights the payment-network tollbooth model. Visa and Mastercard have near-zero marginal cost per transaction, large fixed-cost leverage and network effects that are difficult to dislodge. Stablecoins and A2A payments could, in theory, route around card networks, but both companies are responding by positioning themselves as a network of networks that sits above every rail, including crypto.
The second model the report favors is infrastructure-as-a-service. Once Stripe or Adyen is embedded in a merchant’s technical stack, switching costs become high, revenue compounds with merchant growth, and value-added services such as fraud tools, tax and billing can keep lifting ARPU. The report says Stripe processed $1.4 trillion in 2024, equal to about 1.3% of global GDP. It describes Stripe’s Bridge and Tempo bets as the most aggressive infrastructure push in the group, and says Stripe could capture the developer layer of crypto-native commerce if stablecoin payments scale.
A third model is recurring revenue plus float. Stablecoin reserve income, staking rewards and subscription fees are presented as more durable than trading commissions. The report notes that Coinbase generated $332.5 million from USDC in Q4 2025 alone. Revenue tied to AUM or AUC, it says, is more predictable than revenue tied only to trading volume.
The fourth model is low-cost digital banking in underserved markets. Nubank is the example here. The report says its monthly servicing cost is just $0.80, versus $5 to more than $10 at traditional banks, with monthly activity above 83% and a net interest margin of 17.7%. In underbanked markets, that structure can support near-viral customer acquisition, and the company’s 122.7 million customers create a large cross-sell opportunity.
Higher-risk or less scalable models
The report flags heavy dependence on trading fees as one of the weaker structures. If more than 90% of a company’s revenue comes from trading commissions, the business becomes highly exposed to market cycles. In a crypto bear market, trading volume can fall by 70% to 90%. The paper notes that Coinbase’s trading revenue mix has dropped from 96% in 2020 to an expected 59% in 2025, while Robinhood now has 11 business lines.
Dependence on payment for order flow is another risk. The report says PFOF has already been banned in the European Union and remains under scrutiny in the US, leaving companies that rely on it exposed to severe regulatory pressure. The better path, in the author’s view, is to shift toward subscriptions, interest income and institutional clients, with Robinhood Gold and the Bitstamp acquisition cited as examples.
The report also warns against crypto businesses with no recurring or sticky revenue. Exchanges that rely only on spot trading fees, without staking, stablecoin interest, custody, DeFi protocol revenue or subscriptions, are described as deeply pro-cyclical. A stronger crypto model stacks multiple revenue lines together: trading, staking, interest, protocol fees and subscriptions.
Another example is what the report calls a bitcoin revenue line with little profit. Block reported $1.97 billion in bitcoin revenue in Q3 2025, but the cost of bitcoin revenue was $1.89 billion, leaving gross margin at roughly 4%. The report says the strategic value lies more in ecosystem lock-in than direct profit, because users who buy BTC in Cash App tend to be stickier.
How the companies are integrating crypto
Stablecoin integration is the most common play
The report describes stablecoin integration as the most common and fastest-moving strategy, and lists several examples:
- Visa: USDC settlement inside its network.
- Mastercard: a card partnership with OKX and the acquisition of Zerohash.
- PayPal: PYUSD issuance, with $3.6 billion in circulation, plus DeFi support.
- Stripe: the $1.1 billion acquisition of Bridge for stablecoin orchestration and the buildout of Tempo L1.
- Coinbase: a co-issuer or partner around USDC, with expected 2025 stablecoin revenue of $1.4 billion.
Turning crypto trading into a feature
The report says embedding crypto trading into an existing product suite is a low-cost way to capture retail demand and deepen engagement. Examples include Robinhood integrating crypto trading alongside stocks and options, Revolut offering more than 200 tokens in-app, Nubank’s NuCripto, and Block’s Cash App supporting bitcoin buying, selling and transfers.
Building proprietary blockchain infrastructure
Some firms are going further and building their own chains. The report points to Coinbase’s Base Chain on the OP Stack, Stripe’s Tempo L1 built with Paradigm, and Robinhood Chain on Arbitrum. Owning the chain, the author argues, means owning the economics as well, including sequencer fees, MEV and ecosystem network effects. The analogy used in the report is Visa building VisaNet rather than relying on a third-party network.
Block’s full-stack bitcoin strategy
Block is presented as the clearest full-stack bitcoin bet: consumer wallet through Cash App, merchant acceptance through Square Bitcoin and Lightning, self-custody through Bitkey, mining through Proto, and open-source development through Spiral. The report describes this as a high-conviction vertical bet. If bitcoin becomes an everyday payment rail, Block would own every layer. If it does not, the company will have committed substantial resources to an uncertain outcome.
Custody and institutional services
The report says institutional capital needs trusted, regulated custody. It notes that Coinbase Prime is the custodian for most US spot bitcoin and ether ETFs, while Mastercard and Visa provide compliance, KYC and AML layers for institutional crypto adoption. In the author’s view, this is a high-barrier, high-margin business.
DeFi participation and protocol exposure
On DeFi, the report cites PayPal’s support for PYUSD lending and trading on Ethereum, Coinbase’s Base Chain as a home for DeFi protocols with USDC as the leading stablecoin in DeFi, and staking services from Revolut and Robinhood for ETH and SOL. Even here, DeFi remains mostly an enabling layer rather than a direct retail interface.
Where crypto payments and DeFi can connect
Upstream opportunities: wholesale and infrastructure
The report groups upstream opportunities into several buckets.
The first is stablecoin settlement networks: using stablecoins for interbank settlement, treasury management and merchant payouts, cutting settlement from T+2 to near real time and reducing correspondent banking costs. The likely beneficiaries are Visa, Mastercard, Stripe and Adyen. As an example, the report points to Visa using USDC on Solana to settle VisaNet obligations.
The second is tokenized money markets as liquidity. Tokenized Treasuries such as Ondo OUSG and Franklin OnChain can serve as yield-bearing reserves without sacrificing liquidity or taking credit risk. The report says PayPal, Nu, Revolut and SoFi could benefit, and notes that Mastercard MTN supports tokenized Treasury assets.
The third is cross-border remittance rails. Stablecoins can replace SWIFT and correspondent banking for both B2B and consumer remittances, offering instant settlement, lower fees and better FX rates. Stripe, PayPal’s Xoom, Revolut and Nu are listed as beneficiaries. The report cites Stripe’s USDC payouts for global merchants.
The fourth is programmable compliance. AML, KYC, transaction monitoring and sanctions screening can be embedded into smart-contract-based systems, automating compliance, lowering manual work and enabling real-time risk scoring. The report says all regulated players could benefit, especially those selling value-added services such as Visa and Mastercard. It names Visa Protect and Mastercard Crypto Secure as examples.
Downstream opportunities: consumer and merchant products
On the downstream side, the report lists several product categories.
One is stablecoin-linked cards, which let users spend directly from stablecoin balances while converting to fiat at the point of sale. Beneficiaries include Visa, Mastercard, Revolut and Cash App. The report cites Visa’s more than 130 stablecoin card programs and Mastercard’s OKX Card.
Another is crypto-backed lending, where crypto assets are used as collateral for fiat loans without triggering a taxable event. Robinhood, SoFi, Block and Revolut are listed as potential beneficiaries, with bitcoin-backed loans through Cash App or SoFi given as examples.
A third is yield-bearing savings accounts backed by tokenized Treasuries or DeFi lending protocols, delivered through a regulated wrapper. Nu, Revolut, PayPal and SoFi are named here. The report points to Revolut’s crypto earn-style products, with returns close to staking yields.
A fourth is merchant stablecoin settlement, allowing merchants to settle in stablecoins instead of fiat to reduce FX risk and speed up payouts. Stripe, Adyen, Square and PayPal are listed as beneficiaries. The report says Mastercard, through Nuvei and Circle, enables merchants to settle in USDC, PYUSD or USDG.
A fifth is instant cross-border P2P. Stablecoin-powered remittances can arrive in seconds and cost less than 1%, the report says, putting pressure on Western Union and MoneyGram in Latin America and Asia. PayPal’s Xoom, Revolut, Cash App and Nu are named as beneficiaries. The report gives one example: Nu enabling transfers in Latin America from BRL to USDC and then into local currency.
Niche opportunities with higher margins
The report also points to several narrower opportunities. One is self-custody wallet-as-a-service for institutions and high-net-worth clients, a white-label model that can generate custody fees while meeting self-custody compliance needs. Robinhood, Block and Stripe are listed as beneficiaries, with Block’s Bitkey hardware wallet cited as an example. Another is blockchain-based loyalty programs, where points are issued as tokens and can move across ecosystems, with Mastercard, Visa and PayPal named as likely participants. A third is institutional DeFi integration through compliant middleware, with SoFi, Stripe and Mastercard listed as beneficiaries and SoFi Galileo’s white-label crypto staking for banks given as an example. The report also mentions privacy-preserving payments using zero-knowledge proofs to support stablecoin transfers that are both private and compliant, especially in B2B settings.
A structural view: who wins the rebundling phase
The report connects its argument to another study, The Architecture of Value: Structural Evolution of Financial Innovation and Deep Dive Report on Venture Capital Strategies, and frames the current moment through the lens of unbundling and rebundling.
Its conclusion is that the most durable and scalable fintech businesses in the rebundling phase are usually infrastructure providers, not consumer platforms trying to do everything themselves. The comparison in the report is stark:
- Infrastructure players such as Visa, Mastercard, Stripe and Adyen are described as operating with 90% to 98% gross margins and 50% to 62% operating margins, near-zero customer acquisition cost, and moats built on network effects or developer lock-in. Their revenue grows as the ecosystem grows.
- Consumer platforms such as Robinhood, Nu, Revolut and PayPal are described as operating with 30% to 50% gross margins and 10% to 25% operating margins, customer acquisition costs of $200 to $450, and a need to reach adoption across at least three products before the model becomes attractive. They are also more exposed to regulation and market cycles.
The report uses one comparison to make the point: Visa earns 97.8% gross margin on $17 trillion in payment volume, while Robinhood’s crypto trading revenue is pro-cyclical.
Three dependencies that determine the outcome
The report says the success or failure of rebundling depends on three key factors.
The first is funding. A banking license, in the author’s view, is a compounding moat. Nu and SoFi are listed as winners: Nu has $19 billion in deposits, a 3% to 4% cost of funds and a 17.7% lending net interest margin, while SoFi can gather deposits through its bank charter. PayPal, by contrast, lacks a license and cannot gather deposits, while Revolut’s delayed UK license leaves it unable to compete fully on the lending side. Without a banking license, the report says, consumer fintechs are either dependent on BaaS partners, as highlighted by the 2024 Synapse collapse, or unable to internalize the spread between deposits and loans.
The second is cross-sell economics. With customer acquisition costs of $200 to $450, a single-product user generating $50 in annual revenue and $150 in LTV is unprofitable, the report says. A three-product user generating $180 in annual revenue and $540 in LTV is where the model starts to work. The report lists Robinhood, with 11 products and ARPU of $191, up 82% year over year, Revolut with wealth revenue up 298%, and Nu with 83% cross-product monthly activity as successful examples. PayPal is listed as the weaker case: despite 400 million users, most use it only for checkout, and cross-selling into Venmo, crypto and savings has been limited.
The third is developer and enterprise lock-in. Stripe’s moat, the report says, comes from integration depth. Once a merchant has adopted Billing, Tax, Connect and Radar, ripping Stripe out can require more than six months of engineering work. Each additional product compounds switching costs. Visa’s advantage is different but just as strong: 20,000 banks compete for customers, yet all use Visa’s protocol. Banks cannot leave because the network itself is the product. The report sums that up as zero CAC, 97.8% gross margin and toll collection on $17 trillion in payment volume.
By contrast, the report says consumer crypto faces structural challenges: custody responsibility, compressed margins, weak lock-in and strong cyclicality. It gives several examples, including Uniswap’s 0.3% fee versus Coinbase’s 1% to 2%, the ease with which users can withdraw to self-custody, and Coinbase’s 75% revenue decline in 2022 and 2023.
Three investment themes the report highlights
1. Stablecoin orchestration
The report says Stripe’s $1.1 billion acquisition of Bridge proves one thing: enterprises want stablecoin settlement, but they do not want to run nodes, manage wallets or deal with licensing across 50 states. The winning product, in the author’s view, would be an API that handles multi-chain routing, liquidity optimization, compliance screening and tax reporting in one place, so the enterprise only needs to call a transfer endpoint and the infrastructure handles the rest.
On economics, the report sketches a model with a 0.5% to 1% take rate on large transaction volume, near-zero marginal cost per transaction and high switching costs once integrated. It frames this as a 90%-plus gross-margin opportunity applied to a stablecoin settlement market of more than $2 trillion.
2. Vault-as-a-Service
The second theme is Vault-as-a-Service. The report says PayPal, Robinhood and Nu each custody billions of dollars in crypto assets, and that custody requires OCC compliance, MPC or HSM security, more than $100 million in insurance and disaster recovery capabilities. The opportunity, it argues, is to build the AWS of crypto custody: an API-based service that handles key management, policy engines, compliance reporting and OFAC screening.
The report says every fintech that adds crypto needs this layer, the B2B sales logic is clear, and retention is high because switching custody providers can mean more than 12 months of security review. It names Fireblocks, valued at $8 billion, OCC-chartered Anchorage Digital and Copper.co as examples.
3. DeFi middleware, or the vault curator layer
The third theme is DeFi middleware. The report argues that the market still lacks a Stripe for DeFi yield: a layer that can aggregate returns across Aave, Compound, Morpho and tokenized Treasuries such as Ondo OUSG, while abstracting gas fees, generating IRS-compliant 1099 tax reporting, and adding insurance and compliance wrappers.
It points to Ondo Finance, Backed Finance, Maple and Goldfinch as examples across tokenized Treasuries, tokenized corporate bonds and institutionally underwritten DeFi lending. The report gives a specific illustration: if SoFi holds billions of dollars in deposits and traditional accounts pay 0% to 1%, then Galileo, its B2B platform, could in theory offer a DeFi Yield API connected to tokenized Treasuries yielding 5%. SoFi could then offer customers 4% APY, keep a 1% spread, and avoid putting the assets on its own balance sheet. In the author’s view, this middleware layer between protocols and regulated fintechs, handling compliance wrappers and tax reporting, is one of the clearest gaps in the market.
The report’s closing view
In its conclusion, the report says the future of fintech lies in the fusion of traditional financial platforms and programmable financial infrastructure. Blockchain is not replacing existing systems. It is increasingly being integrated as a settlement and liquidity layer that runs in the background. Stablecoins, tokenized assets and on-chain markets can deliver faster settlement, cheaper cross-border payments and new financial products, while remaining largely invisible to end users.
The firms best positioned to capture value in digital finance, the report says, will be the ones that combine scaled distribution, regulatory trust and infrastructure control. Whether they come through payment networks, developer platforms or consumer financial ecosystems, the winners will be the platforms that abstract away financial complexity while orchestrating value across multiple payment rails. As programmable money sees broader adoption, the fintech companies that successfully connect traditional finance with blockchain infrastructure will shape the next generation of global financial services.