The merchant of medieval times, perhaps venturing from Venice to Paris to conduct trade of great benefit to both sides as well as to the intermediary, took not bullion but letters of credit. It wasn't his good looks that got him through the door of the finest merchants, but his letter of introduction from a respected banking house. He had money on deposit with his banker, but it was his reputation that he took with him.

Today, we don't even have the money on deposit. Fifty years ago, around two-thirds of Americans' wealth was in the form of demand deposits. Now it's less than a third. We have mutual funds, cash management accounts, Starbucks cards and many other places to leave our money. The vaults of our great banking houses are not stuffed with cash or gold bars or negotiable securities.

Fifty years from now? In her excellent and thought-provoking Long Finance report on the future of financial services, Gill Ringland rather memorably said that the citizen of the future would need the critical resources of an identity, a credit score and a parking place in order to function. If that's true – and I certainly believe it to be the direction of travel – the bank's critical role will be built on the customer identities, not their deposits. The vaults will not be stuffed with material valuables, but with the most valuable asset of all: personal data.

The Society for Worldwide Interbank Financial Telecommunication's Innotribe initiative has already flagged just such an infrastructure as a possible banking business for the future, labeling it the "Digital Asset Grid". It seems to me that if an organization as conservative as SWIFT is thinking about this proposition, then it must be taken seriously.

We can see the building blocks: the separation of banking and payments businesses, the emergence of identity management frameworks (and protocols such as OpenIDConnect), the standardization of digital signatures and so on. On top of these basic blocks we can arrange new thinking around vendor relationship management, the personal data store and, in particular, the smartphone as a secure remote control to identity-in-the-cloud and a bank-delivered secure vault for personal and private data to shift us into the next phase of evolution in the financial services world.

In this version of the future, where the child-of-M-Pesa interconnects with the future-of-Simple and the child-of-Zopa to create a "bank" that is actually a bundle of services provided by separately regulated bodies, each one technologically optimal, what is the role of the retail bank? If it is not place to store your money, what is it?

In the same way that I can give iPhoto permission to access my Flickr albums and log in to the Daily Telegraph using my Twitter name, why can't I give Wal-Mart access to my checking account to pay for my groceries (as Merchant Customer Exchange is proposing) and then friend my bank account on Facebook so that I get useful status updates from it ("Hey, I just paid the electricity bill for you"). And if the phone company needs certain data to rent me a smartphone, why can't I just give them permission to get the relevant data from my bank instead of filling out their stupid form? The service provider would not need to trust me, but the financial institution. And if the banks look after my data properly, and don't abuse it, then they can play with it using their big data tools and find some new financial services to sell me, which is fair enough.

The bank isn't a place to store your money; it's a place to store your data. It's the place to safely and securely leave your reputation. Under lock and key, but available to all those with permission to use it.

Dave Birch is the director of Consult Hyperion.