Build, buy or partner? Banks overlook the fourth option
A decision framework for fintech initiatives at banks: build, buy, partner or kill
Most bank discussions about a new fintech capability start as a choice between building it and buying it. That framing leaves out two options: partnering with someone who already has the capability, and stopping the initiative altogether. Both are often the cheapest answer, and both rarely appear on the steering committee slide.
On this page
The default reflex
Build appeals because it keeps control and feels easy to govern. Buy appeals because it looks like speed. Partner tends to be treated as a stopgap until the internal build is ready, and kill is treated as an admission of failure rather than as a decision. That is our observation of a common pattern, not a measured finding. We are not aware of a published dataset on how banks distribute these decisions, so the four cases below illustrate the options; they do not prove a trend.
Build: only where you have the right to win
JPMorgan announced Chase UK in January 2021 as a UK-licensed subsidiary, and launched it in September 2021. In May 2022 an executive said the operation would lose about $450 million a year "for the next few years", according to Banking Dive. By November 2023, the same outlet reported, JPMorgan president Daniel Pinto put it at roughly two million customers and nearly $20 billion of deposits, and expected it to break even within 12 to 18 months, at least two years earlier than the previous estimate of 2027 to 2028. This piece does not assess results since then.
Chase UK is a US bank building in the UK, so it is not a template for most European institutions. It is useful because it shows what a serious build demands: a balance sheet that can absorb several years of losses, a brand and a licence to start with, and a board that agreed the loss curve in advance. Our view is that a build is justified when the capability is one you expect to own for a decade, when nobody can sell you something better, and when the loss curve has been accepted before the money is spent. The third condition is the one most likely to go unchecked.
Buy: you are paying for time
On 19 November 2025 Lloyds Banking Group announced it would acquire Curve, a London digital wallet founded in 2015 that says it has more than six million customers, in order to put wallet features in front of Lloyds' roughly 28 million customers. Lloyds did not disclose the price; Finextra reported about £120 million, less than half of the £250 million Curve had raised, so treat that as a press estimate. Completion was expected in the first half of 2026, subject to regulatory approval. Curve's largest external shareholder, IDC Ventures, with roughly 12%, publicly objected to how the sale was run.
Two lessons follow. The price is the visible cost of a buy, but the shareholder dispute shows that a seller's cap table can turn an acquisition into a negotiation with people outside the deal. And what a bank actually acquires is a product and a team. Whether they stay productive inside a group of Lloyds' size is what decides if the price was right, and that is too early to judge from outside. A buy makes sense when the time to capability matters more than control, and when there is an integration plan that protects what was paid for.
Partner: when someone else is better placed
United Fintech, founded in 2020, describes itself as an industry-neutral platform that connects financial institutions with fintech solutions. When Barclays invested in December 2025 it became the fifth global bank shareholder, joining BNP Paribas, Citi, Danske Bank and Standard Chartered. Four of the five are headquartered in the UK or continental Europe.
We read this as banks that compete hard choosing to share the cost of integrating fintechs rather than each building its own plumbing. That is an interpretation. The evidence so far consists of announcements, not outcomes, and a shared platform brings its own dependency. The partner test is whether the capability is one that does not differentiate you, whether the partner is genuinely better placed, and whether you can live with the dependence. That last point needs specifics: who owns the customer data, what the exit terms are, and how the arrangement meets third-party risk expectations. For EU-regulated financial entities those include the Digital Operational Resilience Act.[1]
Kill: the option nobody puts on the slide
BBVA paid $117 million for Simple in 2014. Within three years it had written down $89.5 million, according to Fintech Futures. BBVA USA announced Simple's closure on 7 January 2021, weeks after PNC agreed to buy BBVA's US business for $11.6 billion. BBVA is a Spanish group and Simple was a US operation, but the governance question travels: the write-down flagged a problem years before anyone acted on it.
The closure was tied to the sale, so this is not a clean case of a bank ending a project on its merits. That is partly the point. The decision arrived when an outside event forced it, not when the numbers first suggested it. We can't know what internal reviews concluded in between. What we can say is that carrying an underperforming initiative has a cost every quarter, and that the cheapest time to agree the stopping criteria is before the build or buy is approved.
Four questions, in order
Option: Build
Choose it when: The capability differentiates you and you expect to own it for a decade.
Ask before committing: Has the board accepted the loss curve? Can we hire and keep the team?
Case in this article : Chase UK (JPMorgan)
Option: Buy
Choose it when: Time to capability matters more than control.
Ask before committing: What protects the product and team after completion? Who else has a say in the deal?
Case in this article : Curve (Lloyds Banking Group)
Option: Partner
Choose it when: The capability does not differentiate you and someone is better placed.
Ask before committing: Who owns the customer data? What are the exit terms and third-party risk controls?
Case in this article : United Fintech (five banks)
Option: Kill
Choose it when: Agreed criteria are missed and there is no credible recovery plan.
Ask before committing: Were the criteria written before funding? Who has the authority to call it?
Case in this article : Simple (BBVA)
Order matters. Ask first whether the capability differentiates you. If it does not, partner or buy will usually beat build. If it does, ask whether you can win and fund the losses. Whatever the answer, write the kill criteria at the same meeting.
What a CEO should do on Monday morning
List every live fintech initiative and assign each one to a single option: build, buy, partner or kill. Name an owner for each. For anything without written kill criteria, agree them before the next steering committee, and put a date on the first review. For each build, ask what loss curve the board has approved. For each partner, read the exit clause.
Aulay's view. Aulay works with banks, insurers, asset managers and the portfolio companies of private equity firms on exactly this kind of decision, and on making it stick once it has been made. If you are a COO, CTO or head of strategy with a fintech portfolio that has never had a structured review of its kill criteria, get in touch.
References
1. Banking Dive, "JPMorgan pushes up UK digital bank's break-even timeline," 29 November 2023. Web article, full text. Supports the Chase UK customer, deposit, break-even and May 2022 loss-guidance figures. Secondary source reporting JPMorgan statements.
2. JPMorgan Chase, "JPMorgan Chase to launch digital consumer banking in the U.K.," 27 January 2021. Press release, full text. Supports the Chase UK announcement and UK-licensed subsidiary status.
https://media.chase.com/news/jpmorgan-chase-to-launch-digital-consumer-banking-in-the-uk
3. Lloyds Banking Group, "Curve acquisition," regulatory announcement, 19 November 2025. Full text. Supports the announcement date, 28 million customers, expected completion in the first half of 2026 and the regulatory-approval condition. The announcement does not disclose a price.
4. Curve, "Lloyds Banking Group acquires Curve," company press page. Full text. Supports Curve's founding year (2015) and its stated customer base of more than six million.
https://www.curve.com/blog/press/curve-is-joining-lloyds-banking-group/
5. Finextra, "Lloyds Bank completes controversial acquisition of Curve." Web article, full text. Supports the reported price of about £120 million, the £250 million raised, and the IDC Ventures objection. Trade-press estimate, not a Lloyds figure.
https://www.finextra.com/newsarticle/46934/lloyds-bank-completes-controversial-acquisition-of-curve
6. FinTech Weekly, "Lloyds Confirms Curve Acquisition as Investor Dispute Intensifies," 19 November 2025. Web article, full text. Supports the announcement date and the roughly 12% IDC Ventures stake.
7. Fintech Global, "Barclays backs United Fintech with strategic investment," 11 December 2025. Web article, full text. Supports United Fintech's description, its founding year (2020) and its five bank shareholders.
https://fintech.global/2025/12/11/barclays-backs-united-fintech-with-strategic-investment/
8. Fintech Futures, "BBVA shutters Simple after paying $117m for it in 2014." Web article, full text. Supports the $117 million price and the $89.5 million write-down. Secondary source.
https://www.fintechfutures.com/paytech/bbva-shutters-simple-after-paying-117m-for-it-in-2014
9. TechCrunch, "BBVA says that it is shutting down banking app Simple, will transfer users to BBVA USA," 7 January 2021. Web article, full text. Supports the closure announcement date and the link to PNC's $11.6 billion acquisition of BBVA's US business.
10. European Union, Regulation (EU) 2022/2554 on digital operational resilience for the financial sector (DORA), Official Journal of the EU, 27 December 2022. Article 28 (general principles on ICT third-party risk). Supports the note on third-party risk.
https://eur-lex.europa.eu/eli/reg/2022/2554/oj
[1]DORA: Digital Operational Resilience Act, Regulation (EU) 2022/2554, in application since 17 January 2025. It sets requirements for managing risk from ICT (information and communication technology) third-party providers.
Frequently asked questions
- What is the difference between building, buying and partnering in fintech?
- Building means developing the capability internally and owning the outcome and the risk. Buying means acquiring a company that already has it. Partnering means using another firm's capability under a contract or shared platform, without owning it.
- When should a bank stop a fintech initiative?
- When it misses criteria agreed in advance, such as customer, revenue or cost targets by a set date, and there is no credible plan to recover. Agreeing those criteria before funding is far easier than arguing them after.
- Is partnering with a fintech riskier than building?
- It is a different risk rather than a larger one. Partnering reduces build cost and time but creates dependence on a third party, so data ownership, exit terms and third-party risk controls matter more.