A payments vendor has probably told you pay by bank will cut your processing costs by half and make chargebacks a thing of the past. Both claims are doing a lot of work in that sentence, and a high-risk merchant who builds a 2027 processing strategy on the vendor version of that pitch is going to be disappointed by the fine print.
The decision is worth getting right because switching even a portion of your payment mix away from cards changes two things at once: what you pay per transaction, and what happens when a customer disputes one. Those two changes do not move in the same direction, and conflating them is how operators end up with a rail that is cheaper on paper and riskier in practice.
This guide compares pay by bank (the account-to-account, A2A rail built on PISP payment initiation) against card processing on the two dimensions that actually matter for a high-risk merchant: real fee ranges, and what a dispute looks like without a card-scheme chargeback right behind it. For the licensing background behind the rail itself, see our guide to open banking for high-risk businesses.
Direct Answer
Pay by bank typically costs 1.0-2.0% versus 2.5-3.5%+ for cards, a real but variable advantage. The tradeoff: there is no card-scheme chargeback right on an A2A payment. Disputes route through the payment provider's own resolution process instead, a structurally different and often weaker protection model than the one cards provide.
What Is "Pay by Bank," and How Is It Different From a Card Payment?
Pay by bank is the retail name for a payment initiated through a PISP (Payment Initiation Service Provider): the customer authorizes a direct transfer from their bank account to the merchant's, with no card network sitting in the middle. A card payment, by contrast, routes through the card scheme (Visa, Mastercard) and the acquiring bank, both of which take a cut and both of which enforce their own rulebook.
That structural difference is the source of almost everything else in this comparison: fees, dispute rights, settlement timing, and who actually has to approve the transaction in real time.
What to Consider:
- No card network intermediary: a pay-by-bank transaction moves directly between two bank accounts, authorized by the customer through their own banking app.
- Different authorization flow: the customer confirms the payment inside their own bank's app or portal, not through a card form and 3D Secure challenge on your checkout page.
- PISP licensing, not card acquiring: the provider enabling the rail is licensed as a payment institution, not registered as a card acquirer.
- Same underlying bank, different rail: a customer's money still sits in a regular bank account either way — only the movement mechanism changes.
Example
A forex broker added a pay-by-bank option at checkout alongside its existing card flow. Customers who chose it completed funding in under 90 seconds without leaving their own banking app, compared to a card flow that routed through 3D Secure authentication and averaged closer to three minutes end to end.
Final Takeaway: Understand pay by bank as a different rail with a different rulebook, not as "a cheaper card" — the comparison only makes sense once you treat the two as structurally separate.
How Do the Fees Actually Compare?
Bank-transfer and A2A processing fees typically run 1.0-2.0% of transaction value. Card processing runs 2.5-3.5%+ once interchange, scheme fees, and acquirer margin (the MDR, or merchant discount rate) are combined, with the average card swipe fee sitting around 2.36%. That is a real cost advantage for pay by bank, not a marketing invention.
It is also not the "up to 50% lower" figure some vendors quote as a flat guarantee. That number describes the widest possible spread between the cheapest bank-transfer pricing and the most expensive card pricing, not a typical merchant's actual saving. Model your own numbers against the realistic ranges, not the vendor's best case.
What to Consider:
- Use ranges, not headlines: budget on 1.0-2.0% versus 2.5-3.5%+, and treat "50% lower" as the best-case spread, not your expected outcome.
- Card fees are not one number: interchange, scheme fees, and MDR stack differently by card type and region, so your actual blended card rate may already sit below 3.5%.
- Volume-based leverage still applies: both rails typically offer better pricing at higher processing volume — get quotes at your real volume, not a rate card.
- Factor in dispute-handling cost: a cheaper per-transaction fee can be offset by higher operational cost if your dispute-resolution process is not built out, covered next.
Example
An operator processing 250,000 EUR a month at a 2.8% blended card rate modeled a shift to a 1.3% pay-by-bank rate, a theoretical saving of roughly 3,750 EUR a month. The saving applied only to the share of volume that actually completed through the new option — about 18% in the pilot's first quarter.
Final Takeaway: Model your saving against a realistic 1.0-2.0% versus 2.5-3.5%+ range and your own likely adoption share, not against a vendor's best-case percentage.
| Feature | Cards | Pay by bank (A2A / PISP) |
|---|---|---|
| Typical fee range | 2.5-3.5%+ (avg. swipe fee ~2.36%) | 1.0-2.0% |
| Fee components | Interchange + scheme fee + MDR | Provider fee, usually flat or tiered |
| Settlement speed | T+1 to T+3 typical | Near-instant to same-day |
| Authentication step | 3D Secure / card form | Customer's own banking app login |
| Chargeback right | Yes, via card scheme | No — provider dispute process only |
What Happens When a Customer Disputes a Pay-by-Bank Transaction?
This is the part of the pitch that tends to get glossed over. Card transactions carry a chargeback right: a customer can dispute a payment through their card issuer, and the issuer can reverse it, sometimes months after the transaction, under rules enforced by the card scheme regardless of what the merchant's own terms say.
A pay-by-bank transaction has no scheme-level equivalent. Under PSD2, a customer has a right to a refund for an unauthorized or incorrectly executed transaction, but a dispute over goods not received, quality, or a service disagreement routes through the payment provider's own resolution process, not a card-scheme arbitration system. That process is set by the provider's own contract terms, and its strength varies considerably from one PISP to another. The authentication standards behind every initiation — the EBA's technical standards on strong customer authentication — reduce fraud risk but do not create a dispute-resolution right of their own.
Reality Check
Pay by bank is marketed as eliminating chargebacks, and technically it does — because there is no chargeback mechanism to begin with. What replaces it is the payment provider's own dispute process, which is contractually defined, varies provider to provider, and offers none of the standardized, scheme-enforced consumer protection a cardholder gets by default. That is a real, structurally different protection model, not a footnote to skip past in the sales deck.
What to Consider:
- Read the provider's dispute terms line by line: do not assume "no chargebacks" means "no disputes" — it means a different, provider-defined process handles them.
- PSD2 refund rights are narrow: they cover unauthorized or incorrectly executed transactions, not buyer's-remorse or service-quality disputes.
- Your own terms of service need to change: a merchant offering pay by bank should disclose the different dispute path to customers explicitly, not silently.
- PSR fraud-liability rules are still pending: the political agreement on the PSR includes a provision making providers liable for losses when fraud-prevention measures fail, but it is not yet law.
Example
An iGaming operator offering pay by bank alongside cards received a service-quality dispute from a customer who expected a chargeback-style reversal. The claim routed through the PISP's own resolution desk under a 10-business-day review window specified in the provider contract — slower, and procedurally different, from the near-automatic card chargeback the customer had used before.
Final Takeaway: Read and disclose the specific dispute-resolution process behind your pay-by-bank provider before you launch it — do not let "no chargebacks" stand in for "no disputes."
Is Pay-by-Bank Adoption Actually Growing, or Just the Rail?
The honest picture here is genuinely mixed, and a merchant deciding whether to build a pay-by-bank option deserves the real numbers rather than a curated growth chart. In the UK, open banking payment volume grew 53% year-on-year through 2025, a real and substantial increase in rail usage tracked as part of the FCA's ongoing open banking and open finance program. Over the same period, consumer awareness of the term "Pay by Bank" actually fell, from 55% to 38%.
Those two numbers are not a contradiction; they describe different things. Volume can grow through embedded checkout defaults and provider partnerships even while the ordinary customer has no idea what to call the option they just used. In the US, adoption sits much further back: roughly 1.5% of consumer transactions used pay by bank in the twelve months to mid-2025.
What to Consider:
- Volume growth is not the same as awareness: a rail can grow fast through defaults and partnerships while most customers still would not recognize its name.
- US and UK are not comparable markets: roughly 1.5% US transaction adoption versus 53% UK volume growth reflect very different starting points, not the same trend at different speeds.
- Checkout UX drives real-world uptake: how the option is presented and defaulted at checkout affects adoption more than the underlying fee saving does.
- Track your own funnel, not industry averages: the only adoption number that matters for your saving math is the completion rate you actually see in your own checkout.
Example
A payments team projected 40% of card volume would shift to pay by bank within a year based on industry growth headlines. Actual first-quarter completion sat at 12%, closer to the customer-awareness figures than the volume-growth figures — a gap the team had not modeled before committing engineering resources to a full rollout.
Final Takeaway: Pilot pay by bank with a modest adoption assumption, measure your own completion rate for a full quarter, and only then scale the rollout or the marketing budget behind it.
| Market | Adoption / growth metric | Consumer awareness trend |
|---|---|---|
| United Kingdom | +53% open banking payment volume, YoY through 2025 | Awareness of "Pay by Bank" fell from 55% to 38% |
| United States | ~1.5% of consumer transactions, 12 months to mid-2025 | No comparable national tracking figure available |
When Does Pay by Bank Make Sense for a High-Risk Merchant?
Pay by bank makes the most sense as a second rail alongside cards, not a replacement for them, for a high-risk merchant specifically. It sidesteps card-scheme mechanics that fall hardest on high-risk verticals — rolling reserves, chargeback ratio monitoring, and MATCH list exposure — without inheriting a chargeback right the customer may still expect.
The provider question matters as much as the fee math: not every PISP onboards high-risk verticals, and crypto, iGaming, forex, and adult operators should confirm sector appetite before building a checkout flow around the option. Our guide on open banking payments for iGaming covers the sharpest version of this friction, where some providers decline the vertical outright.
What to Consider:
- Add it, don't swap it: run pay by bank alongside your existing card and EMI rails rather than replacing either outright.
- Confirm vertical appetite first: get explicit written confirmation that the PISP onboards your specific business type before integrating.
- Disclose the dispute path: update customer-facing terms so the different dispute process is stated plainly, not discovered mid-complaint.
- Track completion, not just cost: a cheaper rail that customers do not finish checkout through saves nothing.
Example
A crypto exchange added pay by bank as an alternative to card funding specifically for customers who had previously been declined at the issuer level. It did not replace the card flow, and after two quarters it accounted for roughly 15% of total deposit volume, concentrated among exactly the previously-declined customer segment it was built for.
Final Takeaway: Deploy pay by bank as a targeted second rail for the specific friction it solves — declined cards, reserve exposure — rather than a blanket replacement for your primary processing.
Making the Pay-by-Bank Decision Honestly
The discipline that matters is holding both halves of this comparison at once. Pay by bank genuinely costs less per transaction than cards, on average, and it genuinely removes the card-scheme chargeback right your customers are used to. Both of those are true, and neither cancels the other out.
A high-risk merchant who adds the rail without updating dispute-resolution disclosures, or who models adoption off a UK volume-growth headline instead of their own checkout data, is optimizing for the vendor's story instead of their own numbers.
Run it as a second rail, confirm the provider actually wants your vertical, disclose the dispute process honestly, and measure your own completion rate before scaling — the same sequencing that separates a real cost saving from a rail nobody finishes checking out through.
How BankMyCapital Helps
Adding a pay-by-bank option without losing ground on dispute handling or vertical acceptance means confirming a PISP's real appetite for your business type, keeping card and EMI rails live as the primary path, and rewriting customer-facing terms before launch rather than after the first dispute. BankMyCapital structures that sequencing rather than presenting A2A as a plug-and-play card replacement.
See our payment processing services for how card, EMI, and open banking rails are structured together for high-risk merchants.
Frequently Asked Questions
Is pay by bank cheaper than card processing?
Usually, yes — typical pay-by-bank fees run 1.0-2.0% against 2.5-3.5%+ for cards. Treat "up to 50% lower" marketing claims as the widest possible spread, not a guaranteed outcome, and model your own numbers against your actual card rate.
Does pay by bank eliminate chargebacks?
It eliminates the chargeback mechanism specifically, because there is no card scheme involved. Disputes still happen; they route through the payment provider's own resolution process instead, which is contractually defined and generally weaker than the standardized, scheme-enforced protection cardholders get by default.
Why is pay-by-bank adoption growing but awareness falling?
In the UK, open banking payment volume grew 53% year-on-year through 2025 largely through embedded checkout defaults and provider partnerships, while consumer awareness of the term "Pay by Bank" fell from 55% to 38% over the same period — customers are using the rail more without necessarily recognizing its name.
Should a high-risk merchant replace cards with pay by bank?
No. The realistic approach is running pay by bank as a second rail alongside cards and EMI-issued accounts, targeted at specific friction like declined cards or reserve exposure, after confirming the provider actually onboards your vertical.
What does BankMyCapital charge to add pay-by-bank processing?
Engagements are scoped individually because provider selection, dispute-process review, and integration work all change the scope, but structured support starts from a fixed floor rather than a percentage of turnover. Ask for a scoped quote once you can describe your current rail mix and target vertical.