Foresight for Bank Boards and CEOs

The Tech Company Delusion in Banking.

I have heard the phrase often enough now that it no longer surprises me.

“We are not a bank. We are a technology company with a banking license.”

It is usually delivered with confidence, sometimes with pride, and almost always with applause. It is meant to signal modernity. It is meant to reassure markets and talent that the institution has escaped the gravitational pull of its past.

But every time I hear it, I am struck by the same thought: this is not transformation. It is displacement.

And displacement, when applied to the core identity of a regulated institution, is not just misguided. It is dangerous.


How the Phrase Was Meant – and How It Was Lost

The phrase did not begin as folly. When it first appeared, it was a corrective. A challenge to complacency. A way of telling bankers that digital was not a side project, not a channel, not something to be bolted on after the “real work” was done.

It was meant to say: technology must be taken seriously.

Somewhere along the way, that message curdled. What was once an exhortation became a rebranding exercise. Technology shifted from enabler to identity. And the discipline of banking – slow, cautious, unforgiving – was quietly reframed as legacy baggage.

That is when the trouble began.


Banking Is Not a Sector You Can Re-Label

No serious institution survives by misunderstanding what it actually is.

Starbucks does not claim to be a technology company that happens to serve coffee, even though its supply chains, payments, logistics, and customer analytics are deeply technological. Harvard does not describe itself as a software firm that happens to teach, despite operating one of the most sophisticated knowledge platforms in the world.

They understand something many banks seem to have forgotten: technology is a means, not a metaphysics.

Banking is not defined by interfaces. It is defined by obligation.


What Banking Actually Is

Strip away the apps, the APIs, the user journeys, and the marketing language, and what remains is something stubbornly analog in spirit.

Banking is the disciplined management of other people’s money under conditions of uncertainty. It is the allocation of capital across time, with asymmetric consequences for failure. It is the acceptance of regulatory scrutiny as a permanent condition of existence.

It is Basel, not because Basel is fashionable, but because capital discipline is the price of trust. It is stress testing, not because it is elegant, but because the world breaks in ways models do not anticipate. It is liquidity management under panic conditions, when confidence evaporates faster than spreadsheets update.

None of this feels like technology. And that is precisely the point.


The Category Error at the Heart of the Delusion

Technology companies optimize for speed, iteration, and optionality. Failure is tolerated, sometimes even celebrated, as long as learning is fast.

Banks optimize for resilience, prudence, and continuity. Failure is not instructive. It is systemic.

Conflating these two logics leads to a category error. You begin to treat regulation as friction rather than foundation. You speak of compliance as overhead instead of license. You imagine risk management as something that can be abstracted into code, rather than exercised as judgment.

The slogans sound modern. The consequences are not.


When the Cycle Turns, the Slogan Collapses

I have seen what happens when the credit cycle turns. When liquidity tightens. When asset quality deteriorates. When assumptions that held in benign conditions fail simultaneously.

In those moments, no one asks about user experience. Regulators do not inquire about design systems. Markets do not reward narrative.

They ask harder questions.

Was credit underwritten with discipline? Were provisions conservative enough? Was capital managed with foresight rather than optimism? Were risks understood, or merely modeled?

This is where the fantasy of being a “tech company” dissolves. Because you cannot refactor your way out of a capital shortfall. You cannot A/B test your way through a liquidity run. You cannot iterate your way past a broken trust contract.


Technology Is Indispensable – and Still Not the Point

None of this is an argument against technology. Quite the opposite.

Modern banking without technology is impossible. AI, automation, analytics, platforms – these are now foundational capabilities. They make banking faster, more transparent, more scalable, and in many cases safer.

But they do not change the essence of the institution.

Technology extends banking. It does not redefine it.

The moment a bank confuses the two, it begins to hollow out its own discipline. Risk becomes abstract. Capital becomes theoretical. Trust becomes assumed rather than defended.

That is not modernization. That is erosion.


The Identity Boards Must Defend

Boards carry a responsibility that branding departments do not. They are custodians of institutional truth.

A bank can be digitally sophisticated, technologically advanced, and operationally modern without pretending to be something it is not. In fact, the strongest institutions I know are precisely those that hold this tension consciously. They invest aggressively in technology while speaking soberly about risk. They modernize delivery while deepening governance. They innovate at the edges while protecting the core.

They understand that the power of banking lies not in novelty, but in reliability.


A Final Thought

The most dangerous moment for any institution is when it becomes embarrassed by its own essence.

Banking is not a shameful inheritance to be disguised with tech language. It is a hard-won societal role, built on constraint, accountability, and trust earned over time.

Technology should make banking better. Faster. Clearer. More humane.

But the day a bank truly believes it has transcended banking by calling itself something else is the day it stops taking its own responsibility seriously.

And responsibility – unfashionable, unglamorous, unavoidable – is the one thing no line of code can replace.


Continuing this conversation on Linkedin: see here