Key takeaways
- Neobanks are scaling faster than their infrastructure. The market is projected to grow from $261.4B in 2025 to $385B in 2026, heading toward 850 million users by 2030 - while the number of neobanks that actually control their issuing stack has barely moved.
- FX spread, settlement timing, authorisation reliability, and reserve requirements are not product decisions most neobanks make consciously. They are inherited from whoever sits between them and the card network.
- Every intermediary in that chain takes basis points and adds a dependency. Consolidating issuer and processor into one licensed partner is how neobanks reclaim both.
- Every intermediary in that chain takes basis points and adds a dependency. Consolidating issuer and processor into one licensed partner is how neobanks reclaim both.
- Reap is a Visa Principal Issuer in Hong Kong and Mexico, with stablecoin- and fiat-funded programs live in 2-3 months post-KYB. If your growth is capped by your issuing stack.
Two neobanks can look identical from the outside and still deliver very different economics to their users. Recent third-party research tested this by moving real money through five leading neobanks and putting a number on it: on the same €5,000 round-trip, one cost about $2 and another over $130, an FX spread ranging from near zero to roughly 2%. Some of that spread is the stack each neobank runs on, some is its own pricing on top. Either way, it is set behind the app, not in it.
And FX is just the account layer. The bigger economics sit on the card, where the issuing stack shapes the FX spread on every foreign swipe, how fast it settles, how reliably it authorises, and how quickly a neobank can ship new features. Most neobanks treat that stack as plumbing and never look closely, even though it sets the ceiling on what the product can do. This piece breaks down the card-issuing stack, where the margin leaks, and what changes when a neobank consolidates it.
The typical layers of a card program
A card program has at least four functions that must be filled, whether the neobank fills them itself or buys them: the card network, the issuer, the processor, and the program manager. The pricing question only makes sense once you can see the whole system.
The default path for a new neobank is to buy each function from a separate provider:
A BIN sponsor supplies the issuer layer: it is a scheme-licensed principal member that owns the BIN, and with that BIN comes the neobank's access to the card network such as Visa or Mastercard.
- A BIN sponsor supplies the issuer layer: it is a scheme-licensed principal member that owns the BIN, and with that BIN comes the neobank's access to the card network such as Visa or Mastercard.
- A processor handles authorisation, clearing and settlement.
- A program manager runs the day-to-day, or the neobank runs that layer itself.
- Operationally, a card manufacturer produces and personalises the physical plastic.
The above model works but it carries significant structural costs.
Costs of building your own card program from scratch
Every intermediary prices its own risk. FX handling, settlement float, and authorisation each pass through a party that adds a margin. Those margins compound into the spread the end user eventually pays, which is precisely the variance third-party testing measured. A neobank running four vendors deep often cannot fully explain its own FX pricing, because it did not set all of it.
Every intermediary is a dependency. Feature roadmaps get gated on someone else's release cycle. A vendor's compliance issue becomes your outage. The industry has already lived through this: when a major banking-as-a-service provider collapsed in 2024, the fintechs downstream discovered that their user balances and their reconciliation were only ever as reliable as a vendor they had no control over.
Consolidating issuing and processing under a single licensed partner, known as full-stack issuing, is the market's answer to both problems. Fewer parties between you and the network means fewer margin takers, fewer roadmap dependencies, and one accountable counterparty when something breaks. We've written up the mechanics of the licence layer specifically in Understanding BIN Sponsorships, and the broader model in What is Card-as-a-Service (CaaS)?
Where does the neobank margin actually leak?
The margin in a neobank card program leaks in three places, in descending order of impact on a scaling neobank: FX spread, settlement timing and idle capital, and authorisation reliability.
1. FX spread. This is the leak third-party testing surfaced, and the one users feel most directly. When conversion happens through a chain of parties, each applying its own buffer, a spread that looks reasonable per transaction becomes a meaningful revenue eater and trust question at volume. Users increasingly compare rates across apps, and a roughly 2% spread against a competitor at near mid-market is a churn risk, not just a pricing choice.
2. Settlement timing and idle capital. Traditional fiat programs require pre-funded reserves against multi-day settlement windows, typically several days of projected spend sitting idle. That capital is unavailable for growth. Stablecoin-funded programs can compress this materially, because funding and settlement don't wait on banking hours or weekends. At low volume this is a rounding error. At scale it is one of the largest line items in the program's economics.
3. Authorisation and reliability. Declines are invisible in a demo and brutal in production. Whether the neobank can implement real-time authorisation logic against its own ledger, deducting a user's digital asset balance at the moment of spend, rather than reconciling after, determines both the user experience and the operational risk it carries.
The rise of neobanks and where they are heading
Neobanks have moved from fintech experiment to the default way a large share of the world expects to bank:
- The global neobanking market is projected to grow from $261.4B in 2025 to roughly $385B in 2026, with forecasts pointing to continued compounding growth through the early 2030s.
- Nubank has passed 110 million users. Revolut and Klarna together serve more than 135 million.
- Neobanks reached about 1.4 billion accounts globally in 2026, adding roughly 300 million customers in two years.
- A newer wave - stablecoin-native neobanks - is building on top of this base, letting users hold, spend and convert digital assets as fluidly as fiat. For where stablecoin adoption stands overall, see Reap's Stablecoin Statistics & Data 2026.
Three shifts are shaping what comes next, and each raises the bar on infrastructure:
From growth to unit economics. The era of subsidised acquisition is closing. Neobanks have the users but not the profit: digital-first banks still account for only about 5% of global retail banking revenue, at roughly $70 to $80 per customer a year. That makes interchange, FX margin, and idle capital strategic rather than back-office. A neobank cannot optimise economics it does not control.
From plastic to credentials everywhere. Users expect virtual issuance in seconds and instant provisioning into Apple Pay and Google Pay, and increasingly into wearables and connected devices. Every one of those form factors depends on tokenisation and certification depth in the issuing stack. Fragmented stacks ship these features late, if at all.
From cross-border as a feature to cross-border as the product. Dollar-denominated spending is in structural demand in markets where local banking cannot easily provide it. Serving that demand through conventional fiat infrastructure means rebuilding, re-underwriting and re-funding in each market. Stablecoin-backed programs are the more scalable path, and the newest frontier, agentic payments, assumes programmable, always-on rails from the start.
Why neobanks are the natural fit for Reap's card issuing
Neobanks are one of the core use cases Reap's card issuing program is built for: connecting a user's digital asset or fiat balance to a card that works anywhere Visa is accepted. Concretely, here's what changes.
One licensed issuer, not a chain of vendors. Reap is a Visa Principal Issuer in both Hong Kong and Mexico. Neobanks don't need to source and coordinate a separate BIN sponsor and program manager - the exact multi-party arrangement that produces FX pricing variance and roadmap dependencies that slow feature delivery.
Fund and settle in stablecoins or fiat. Programs can be funded in whichever form suits the business, and users can spend digital assets or fiat interchangeably. Neobanks get stablecoin settlement efficiency without being forced into a treasury model they don't want.
Real-time authorisation against your own ledger. Reap supports both standard and real-time authorisation. With real-time authorisation, each transaction is referred to you at the point of authorisation, and you apply the debit to the collateral balance recorded on your own ledger and held with your own custodian, in real time, rather than reconciling afterwards.
Credit, not prepaid. Reap issues Visa credit cards, which means stronger fraud protection, proper chargeback support, and acceptance at merchants and MCCs that routinely block debit and prepaid BINs. For a neobank, declined transactions at hotels, airlines and subscription merchants are a retention problem, not an edge case.
Live in 2-3 months, not 12-18. Most programs go live within 2-3 months post-KYB. Where a client's engineering team moves fast, we've seen 2 weeks. Compare that to the 12-18 months a traditional multi-vendor issuing build typically absorbs.
Compliance and fraud built in. PCI DSS compliance via our hosted widget (so you avoid your own certification burden), plus tokenisation, 3DS, two-factor authentication, Visa Risk Manager and Chainalysis monitoring as standard. These are the layers most likely to be underbuilt when a neobank assembles its own stack.
Your brand, your rules. Card design, spend controls by time, location and merchant category, and every UI touchpoint stay yours. The end user sees your neobank, not your issuer.
Scale without rebuilding. Unlimited virtual and physical cards, USD or HKD denomination, and international logistics tuned to where your cardholders actually are.
What consolidating your card-issuing stack doesn't change
Worth stating plainly, because credible infrastructure claims come with honest limits.
Compliance obligations are identical. Consolidating your stack does not reduce KYC, KYB, AML monitoring, sanctions screening or reporting requirements. It changes who operates the tooling, not whether the obligations apply.
Local regulation still applies. A principal issuing licence in one jurisdiction is not a global passport. Market entry still requires jurisdiction-specific work; the gain is that you're not rebuilding the entire stack each time.
The user experience should feel unremarkable. Merchants receive fiat through the normal network flow. Cardholders tap or swipe. If your users have to understand settlement mechanics, something has gone wrong in the design.
FAQ
Are neobanks actually banks?
Most neobanks are not licensed banks. They rely on licensed partners, an issuing bank or principal member, a processor, sometimes a program manager, to hold the licence, move the money and authorise transactions. Those vendors, not the app, determine FX pricing, settlement speed and reliability.
Why do FX spreads differ so much between neobanks?
Because each intermediary in the chain applies its own margin. A neobank running several vendors deep inherits the accumulated spread. Testing on leading apps found this ranged from roughly 2% down to near mid-market.
Is full-stack issuing only relevant to crypto-native neobanks?
No. The benefits of fewer intermediaries, lower idle capital and faster feature delivery apply to any card program. Stablecoin funding is an option, not a prerequisite; programs can be funded in fiat.
How long does it take a neobank to launch a card program with Reap?
Most clients go live in 2-3 months after KYB, covering onboarding, integration, testing and deployment. Fast-moving engineering teams have launched in as little as 2 weeks.
Can our users spend both stablecoins and fiat on the same card?
Yes. Programs are denominated in USD or HKD, and users can settle in either digital assets or fiat.
If you're a neobank, let's talk
Independent research is a useful reminder that in a crowded market, the infrastructure decisions your users never see are the ones that decide whether they trust you with their next transaction. Growth is good. Growth on a 2% spread and a fragile vendor chain is a liability compounding quietly in the background.
If you're building or scaling a neobank and want a single, transparent partner for card issuing - one Visa Principal Issuer, stablecoin and fiat native, live in months instead of years - get in touch with Reap. We'll pressure-test what your program needs, then build it with you.
Disclaimer
The information provided in this material is for general informational purposes only and does not constitute legal, financial, tax, or business advice. It should not be interpreted as a recommendation, offer, solicitation, or inducement to engage with Reap’s products or services. Any use of Reap’s services is at the user’s sole risk and discretion.
Reap makes no representation or warranty, express or implied, regarding the accuracy, completeness, or reliability of the information provided. Services are governed exclusively by Reap’s applicable legal agreements. Service availability, features, and eligibility may vary by jurisdiction and are subject to regulatory, card network, and operational limitations.
All trademarks, logos, and brand names are the property of Reap and/or their respective owners. References to third-party platforms or services are for descriptive purposes only and do not imply endorsement, partnership, or affiliation.
Reap’s services and information are provided on an “as is” and “as available” basis, without warranties of any kind. Reap shall not be liable for any loss or damage arising from the use of, or reliance on, this information or its services.
