The $15 Billion Question: Can Tokenization Close the Scale Gap in US Banking?

The regional bank predicament is often misunderstood. It’s not a crisis. It’s something even worse... a slow bleed that doesn’t demand urgent action until it’s too late.

JP Morgan spends roughly $15 billion a year on technology. Out of the 4,000 banks in the US, approximately 3,900 of them have an entire expense base smaller than that single line item.

That number is the gravitational force reshaping American banking. And it raises a question that the tokenization community has been circling for years without quite landing on. Could tokenized financial infrastructure be the thing that finally levels this playing field?

The short answer is probably not for most banks. But the longer answer is more interesting, and more uncomfortable, than either the optimists or the skeptics tend to acknowledge.

The Melting Ice Cube Problem

The regional bank predicament is often misunderstood. It’s not a crisis. It’s something even worse… a slow bleed that doesn’t demand urgent action until it’s too late.

In any given year, maybe 5% of consumer banking relationships are actually in motion with people switching accounts, new entrants joining the system. The largest banks are capturing an outsized share of those new relationships, perhaps 30% to 50% of the flow. But because the addressable pool is small, the gap shows up as just a 1-2% annual growth differential. JP Morgan grows consumer deposits at 5-6%. The average regional bank grows at 2-3%, essentially just holding its existing base and capturing a bit of inflation.

The comparison to traditional asset management is apt. AUM across the industry is still higher than it was a decade ago, but the ability to generate organic flows has been fundamentally impaired. The regionals aren’t dying. They’re running a melting ice cube. Still large, still functioning, but with a trajectory that only goes one direction if nothing changes.

Regional banks can be fast followers on technology. They can watch the mega banks test eight features in their app and copy the one that sticks. But this guarantees you’re always six months behind, always one step back. And the vendor ecosystem, the FIS, Fiserv and Jack Henrys that rent core infrastructure to smaller banks, keeps them alive but doesn’t prioritize their competitive differentiation. Every banker knows they’re running on someone else’s product roadmap.

So when someone proposes that tokenization might be the answer, the instinct from anyone who has actually sat in a regional bank’s technology steering committee is yeah sure, and we still haven’t finished migrating off Windows XP.

That instinct is mostly right. But it’s worth examining where it might be wrong.

What Tokenization Could Change — In Theory

Strip away the hype, and there are three areas where tokenized infrastructure could restructure the cost curves that currently favor scale.

Settlement and clearing. A midsize bank today pays to access the same settlement infrastructure as JP Morgan, but it processes a fraction of the volume. The fixed costs of connecting to payment rails, managing correspondent banking relationships, and reconciling across systems are largely the same whether you’re processing $50 billion or $500 billion a month. Tokenized settlement has the potential to collapse these layers. If the ledger is the settlement, the cost of clearing doesn’t scale with the complexity of your back office. It scales with your transaction volume, which is a more democratic cost structure.

Product manufacturing. When a $3 trillion bank decides to enter private credit alongside its traditional lending book, it can build custom infrastructure, hire specialized teams, and absorb the startup costs across a massive balance sheet. A $30 billion regional faces the same regulatory complexity but can’t spread the investment. If a private credit fund or syndicated loan exists as a standardized tokenized instrument with programmable compliance built in, the marginal cost of offering it drops. The regional bank doesn’t need to build the factory. It needs to connect to one.

Distribution infrastructure. Tokenized protocols are inherently open. If the rails are shared, differentiation shifts from who built the best pipes to who offers the best advice, the best relationship, the best local knowledge. That’s a game regional banks should be able to win.

That’s the theory. But here’s the problem.

The Challenge of Getting There From Here

The average community or midsize regional bank is running on a core system that was architected in the 1990s. Their IT teams are small. Their change management capacity is limited. Their boards are not staffed with people who understand distributed ledger technology, and quite frankly, they shouldn’t be, because that’s not where their fiduciary attention needs to be focused.

These institutions struggle to implement real-time payments. They are years behind on open banking APIs. Many still run batch processing overnight. Asking them to adopt tokenized settlement infrastructure isn’t just ambitious, it fundamentally misunderstands how institutional change works in community banking. You don’t leapfrog from batch processing to programmable money. You migrate to real-time payments, then modernize your core, then maybe, just maybe, you’re in a position to think about tokenized instruments. That’s a five to ten year journey, and it assumes leadership continuity, board support, and vendor cooperation that most banks don’t have.

There’s also the talent problem. The people who understand both banking regulation and blockchain architecture are rare and expensive. They’re working at JP Morgan, or at crypto native firms, or at fintechs. They’re not moving to a $15 billion asset bank in the Midwest, no matter how compelling the mission statement.

And then there’s the vendor question. The core banking vendors that serve the majority of US banks, Fiserv, Jack Henry, FIS etc, have been cautious about tokenization for understandable reasons. Their business model is built on long term contracts with incremental feature rollouts. They have no incentive to enable a technology layer that could disintermediate them. Until the vendor ecosystem moves, the 3,000+ banks that depend on it are largely stuck.

So Who Actually Benefits?

If we’re honest about the operational constraints, then the tokenization landscape in US banking probably shakes out into three tiers.

The top 20 banks like the PNCs, US Bancorps, Truists, etc, have the scale, the technology budgets, and the organizational capacity to run pilots and build toward tokenized infrastructure. Some already are. These are the institutions that might genuinely close a portion of the scale gap with the mega banks, particularly in capital markets and treasury services. But these banks also have the most to lose from disruption to existing revenue streams, which makes them cautious adopters.

The 4,000 community and small regional banks aren’t going to build tokenized infrastructure. Their realistic path is the same as it’s always been: renting. They’ll plug into tokenized systems that someone else builds, the same way they rent core banking today. The question is whether tokenization vendors build products for them at price points that work, and whether the economics of serving small banks on new rails are any better than the economics of serving them on old ones. History suggests the answer is: not obviously.

The biggest beneficiaries might actually be non-banks. If you’re Apollo or Blackstone and you’re already building tokenized fund structures, the regional bank isn’t your competitor, it’s more likely your distribution channel. Private credit balances have roughly doubled since pre-COVID while bank C&I lending has grown only about 30%. Tokenization could accelerate that shift, not reverse it. The regional bank gets to offer its clients access to tokenized products, but the economics flow to whoever manufactured the instrument. That’s a different story from “tokenization levels the playing field.” It might just change who the regional bank is dependent on, from Fiserv to Blackstone.

The Regulatory Window Is Real, But Narrow

One thing the optimists get right is that the current regulatory environment is unusually open to experimentation. The Basel III endgame package is expected to keep capital requirements roughly flat. The SEC’s posture on digital assets has shifted. The OCC has signaled openness to banks engaging with blockchain based systems.

But none of this is permanent. Administrations change. Personnel change. The six times debt to EBITDA limit that kept banks out of software and enterprise lending for years was a personnel decision, not a statutory one. The same is true for whatever openness currently exists toward tokenized infrastructure.

For the banks that can move, that top tier with real technology capacity, the window matters. For everyone else, the window is largely irrelevant because they don’t have the organizational velocity to exploit it before it closes.

Where This Actually Lands

The honest version of the tokenization thesis for banking isn’t “this levels the playing field.” It’s something more nuanced and a bit less comfortable.

For a small number of large regional banks, tokenized infrastructure creates a genuine opportunity to compete more effectively in capital markets, treasury services, and structured product distribution. Not to match JP Morgan, but to close from 60 cents on the dollar to maybe 75.

For the vast majority of US banks, tokenization is going to be something that happens to them, not something they do. They’ll access tokenized products through vendors or non-bank partners. They’ll distribute instruments they didn’t manufacture on rails they don’t control. Their margins on those products will be thin, and the value will accrue to whoever owns the infrastructure layer.

And for non-banks like the private credit firms, the alternative asset managers, the fintechs, tokenization is a distribution play. It’s a way to reach the $15 billion bank’s client base without building a branch network or a deposit franchise. The regional bank becomes the last mile, the trusted face. Valuable, but not where the economics concentrate.

There’s an old adage in banking that whatever grows the fastest tends to be the problem. It’s worth keeping that in mind before assuming tokenization will be the exception to every other technology cycle that promised to democratize finance and ended up concentrating it further.

The regional bank’s core advantages, the local knowledge, relationship depth, community trust, are real. But technology has never been the thing that made those advantages valuable. What made them valuable was that they were hard to replicate at scale. The question tokenization raises is whether the infrastructure makes it easier for someone else to replicate what regional banks do.

That’s the question that should be keeping regional bank boards up at night.