Building a Bank from Scratch
The bank of the future probably won’t be made of brick or stone, and may not be much to look at. It’s easier to say what it will do.
Entrepreneur and startup mentor Jack Gavigan, who traded tweets with Marc Andreessen and others on the subject back in February, elaborated on his vision for a bank of the future in a blogpost yesterday.
Gavigan believes that a bank charter — i.e. being a bank — is necessary for disrupting banking. Thus Movenbank and BankSimple are derided as being mere “presentation layers” on top of someone else’s banking infrastructure. The full stack is needed, Gavigan wrote.
He then defined banking as the fusion of two businesses:
A bank can be broken down into two distinct businesses. The first is the infrastructure required to offer banking services: the banking licence or charter, branches and call centres (increasingly optional), IT systems, ATMs, AML and KYC procedures, connections to domestic inter-bank networks and SWIFT, the ability to issue credit and debit cards, etc. Some of these are revenue-generating (e.g. fees on SWIFT transfers, payment card interchange fees) but, traditionally, retail banks have offered “free” banking to attract customers so that the second business can offer them financial products like credit, overdrafts, loans, savings, investments, pensions, foreign exchange and insurance. This second business has attracted plenty of competition (including startups like Wonga, Funding Circle and TransferWise), but there appears to be little appetite for investing in the infrastructure required to become an actual bank.
Wonga and Funding Circle offer loans, while TransferWise is a money-movement service.
A company that builds a core banking platform, part of the first business in Gavigan’s scheme above, could truly disrupt banking. Companies using the platform could build out their own services on top — this is more or less what The Bancorp is doing now. Gavigan also envisions a marketplace or app store where customers could take what they need and leave the rest.
The forthcoming Atom Bank in the UK does not meet Gavigan’s requirements because it will use Fiserv’s technology for its backend. The investment needed for building a bank from scratch will be considerable Gavigan warned. He expressed concern that venture capitalists, accustomed to funding small teams tackling a specific problem, will not have the appetite for it.
VCs are just as likely to fund people as ideas, according to this profile of Clinkle and its CEO Josh Duplan, which mentions in passing that Duplan, still sitting on top of $20 million, is “rumored to be mulling over a new round of financing, possibly to try to buy a bank and a payment processor for Clinkle.”
There are a few startups playing in the “API Banking” field, such as Standard Treasury and Corefx, but these companies don’t fit Gavigan’s bill either. Zac Townsend of Standard Treasury recently wrote what he thought the essence banking was, and hence what the bank of the future should do: “The full potential of what banking could be is not even seen in a better retail product experience like Simple. Instead, banking is about making commerce simpler and easier for businesses and consumers alike. It’s about empowering calculated risks and enabling trade in all its forms.”
As for starting a vertically integrated bank from scratch, no one has done that except perhaps GoBank, which has not introduced new features or announced plans to expand its offerings since its launch last summer. With 14,000 banks in the U.S., and many vulnerable to collapse under the strain of rising IT costs and regulatory pressure over the next few years, banking licenses could be available — for a price, naturally.
Perhaps VCs like Andreessen should start by picking up a bank charter, then awarding it to the developer that demonstrates he can build the best bank from scratch.
