B-Hive strives to boost collaboration between banks and fintechs


Peter Lee
Published on:

The rules of engagement between incumbents and fintech suppliers of new products, which banks then white-label to their customers, need review.

The rise of potential fintech disruptors and the challenge for incumbent banks of digitalizing their businesses and staying abreast of these innovators – while still keeping their ageing operating systems running – has changed the job of chief technology officers.

It comes with higher pressure and more scrutiny, but greater excitement.


Fabian Vandenreydt,

“There are two jobs in banking that I see as particularly interesting,” says Fabian Vandenreydt, executive chairman of B-Hive, a Brussels-based member association that promotes cooperation between large incumbent insurance companies and banks ­– such as ING, BNP Paribas, KBC ­­– and fintechs.

“First, architects today in charge of, say, payment engines or risk-management systems at banks are rarely just buying and monitoring a single, large solution. Rather, they want to ensure that they can integrate new components that may be coming in all the time from different recent and maturing start-ups. That’s very cutting edge.

“And the second exciting job is that of partner manager inside the bank for these fintech providers.”

Vandenreydt was head of capital markets and innovation at Swift for more than 10 years before joining B-Hive, and before that a consultant, after previously working at JPMorgan.

He says: “The innovation scene today focuses a lot on how fintechs should pitch their solutions to banks.

“But we’re seeing more inquiry coming the other way, with banks outlining a business problem to select potential vendors – perhaps fraud detection, regulatory compliance, payments, data analytics, identity management ­– and asking them to present solutions.”

He returns to the component manufacturing analogy, saying:“Twenty years ago, a bank might buy a whole solution from one large, established vendor. Now it’s more about identifying, assessing and assembling best-of-breed software components from multiple providers.”

Strained relations

There are very few genuine tech-enabled disruptors on the scene ambitious to take substantial share from established banks. Their backers might point to Revolut, Monzo or N26, but most fintechs are selling to the established players.

However, relations between vendors and buyers are strained today. The old IT procurement function inside banks was geared around due diligence before signing multi-year contracts to take a whole system from IBM, Oracle or SAP.

Both sides were big and could handle these long-drawn out review cycles: not so new start-ups keen for a big, first bank customer to validate an offering.

“The timelines are a nightmare,” Jamie Campbell, head of awareness at Bud – a plug-and-play financial services platform, told the Innovate Finance global summit last year, in a discussion on the slow progress with open banking.

“Just to go through procurement is one thing. Then to do the actual implementation, you are looking at more than a year.”

The situation may even be getting worse. Others say it can now be closer to 18 months.

And then, when a small fintech finds its much-sought-after first, large bank customer, that customer may demand so much attention – handholding and training through initial implementation then subsequent amendments and updates – that the fintech has no remaining capacity to find its second customer.

It has been owned without selling equity.

Cynical agenda

Worse, there is now increasing suspicion that some large banks may be pursuing a cynical agenda through so-called due diligence.

Another source recounts a fintech’s suspicions over recently being asked to provide its artificial-intelligence (AI) service for free to a bank to review before perhaps buying.

“So, they want to take our toys, play with them for two months, figure out how they work and then build their own versions?” asks the source. “They might be dangling a potentially big contract after that, but… no thanks.”

Fintechs are often concerned that they should not be tied into exclusive arrangements with the first big bank to take their new offering, but they also need to think about banks’ casual brutality towards fintechs.

“Why does it even matter who the fintech is,” one banker responds to Euromoney’s question on a new offering the bank will white-label to its customers. “The one we’re working with right now is the best.

“But if we find a better one, we’ll unplug from it and go with the new fintech.”

Algorithms always make assumptions on that data. And if the algorithm is obscure – for example, written in a language that only the writer understands – that might hide a bias inside the algorithm 
 - Fabian Vandenreydt, B-Hive

It remains to be seen how banks and the fintech community might establish protocols around intellectual-property protection and payment while trialing new products.

At B-Hive, Vandenreydt is looking to compress the time taken over due diligence between traditionally slow-moving banks and fintechs looking to shorten time to market and time to revenue.

“Fintechs need to understand the bank’s review cycles,” he tells Euromoney. “It would be great if we could take an 18-month process down to say six months and we are trying to do that through a project called ‘know your vendor’, which obviously echoes know your customer.

“One concrete step fintechs can take is improving their own cyber security. A lot of start-up solutions now are cloud based and we’re looking to train them and test their cyber defences to see whether we might produce a label on cyber that might reassure banks.”


Banks need to be especially wary over whatever latest hype attracts senior management and board attention. Three years ago, it was blockchain. Right now, it is AI. Many sources Euromoney speaks to agree that 2019 will be the year of widespread AI implementation, especially around bank transaction data analytics.

“A big issue in AI is around obscurity,” says Vandenreydt. “AI solutions take different sets of structured and unstructured data, apply an algorithm to those and then generate an output – an insurance pricing or credit underwriting decision, for example.

“But remember that algorithms always make assumptions on that data. And if the algorithm is obscure – for example, written in a language that only the writer understands – that might hide a bias inside the algorithm.”

He suggests: “We will be hearing a lot more about ethical AI. The people inside banks that approve AI tools need to think about the outputs. Do they always suggest charging high prices for insurance or credit for residents in a certain district, for example?

“In the end, questions will be asked about whether regulators should have a role in approving these algos.”

No one wants that, do they?