Tokenization Is Not a Product

Tokenization is an infrastructure revolution that rewires the fundamental economics of banking. It changes what settlement means, what deposits are worth, and what collateral can do.

Mar 08, 2026

Many banks learning about tokenization are making a common mistake. They’re treating it as a product innovation. A new way to move money, a faster rail, a shinier pipe. It is not. Tokenization is an infrastructure revolution that rewires the fundamental economics of banking. It changes what settlement means, what deposits are worth, and what collateral can do.

I think the distinction matters enormously. If tokenization is a product, you adopt it on your timeline. You form a working group, hire a consultant, run a pilot in 2028. If tokenization is infrastructure, it doesn’t wait for your roadmap. It reprices your deposit franchise, disintermediates your settlement revenue, and commoditizes your lending relationships. Whether you’re ready or not.

Why Tokenized Assets Still Don’t Work Like Real Ones

Most banks are planning for the first version. They’re going to experience the second.

Here are three things they’re not seeing.

Blind Spot #1: Atomic Settlement Kills Your Revenue Model

When banks encounter atomic settlement in their tokenization education, it’ll often show up as a bullet point about “reducing counterparty risk” in securities settlement. Then I’m sure everyone moves on. I doubt many bankers are taking the time to model the P&L impact.

They really should. Atomic settlement doesn’t just reduce risk. It eliminates the settlement window that has generated bank revenue for decades. When cash and securities both exist as tokens on the same ledger, a smart contract executes simultaneous delivery and payment. No float. No T+1 delay. No overnight batch processing. The transaction completes instantly or reverts as if it never happened.

This is not theoretical. Broadridge’s Distributed Ledger Repo platform processed $365 billion in average daily volume in January 2026, a 508% year over year increase. JPMorgan’s Kinexys processed over $1.5 trillion in tokenized transactions in 2025. DTCC received SEC authorization in December 2025 to tokenize custodied assets, launching in the second half of this year. Fnality raised $136 million in September 2025 to build institutional onchain settlement infrastructure, backed by Bank of America, Citi, and DTCC.

Now think about what this does to a typical commercial bank’s income statement. Settlement float income is likely gone. Correspondent banking spreads on cross-border payments that take 3 to 5 days and cost 2 to 7%, compressed to minutes at under 2%. Back office processing that McKinsey and Accenture estimate DLT could cut by 50%, I expect those savings accrue to the market, not to the bank. Netting services that banks orchestrate among customers will be obsolete when every trade settles atomically, one by one.

The BIS’s 2025 Annual Economic Report describes a “next-generation monetary and financial system” built on tokenized unified ledgers. The New York Fed’s research notes that onchain transactions settle with “objective cryptographic guarantees” that replace subjective trust in intermediaries.

That intermediary role, the trusted counterparty standing between two parties who can’t yet trust each other directly, has been the gravitational center of commercial banking for centuries. It is extraordinarily valuable. And it is precisely what atomic settlement makes optional.

Blind Spot #2: The Yield Awakening Will Empty Deposit Franchises

Here’s the question that everybody in banking should be asking: When a corporate treasurer can allocate every idle dollar across a spectrum of yield bearing tokenized instruments, things like T-bills, money market funds, short duration credit, and structured products, calibrated to their risk appetite, with instant settlement, 24/7, why would a single dollar remain in a non-interest bearing demand deposit account?

The concept of “idle cash” is a product of friction. Corporate treasurers have always known they were leaving yield on the table. They accepted it because moving money was slow, expensive, and operationally burdensome. Sweeps ran overnight. Money market fund subscriptions settled T+1. Optimizing across instruments meant coordinating multiple custodians, each with their own settlement windows and minimum thresholds.

Tokenization will likely eliminate all of that. When every asset class, from treasuries, money market funds, commercial paper, to real estate debt, can exist as a token on the same ledger, a treasurer can build a continuously optimized portfolio of yield-bearing positions and rebalance in real-time based on cash flow needs. The operating account becomes a waystation, not a resting place. Capital passes through on its way to productive deployment.

The thing is, this infrastructure is already live. BlackRock’s BUIDL fund holds $2.5 billion in tokenized Treasury assets. Franklin Templeton launched a blockchain native institutional money market share class in January 2026. Ondo Finance has $1.4 billion in tokenized Treasuries with 24/7 subscriptions and redemptions across multiple chains. BNY and Goldman Sachs announced a joint initiative in late 2025 to build tokenized money market fund infrastructure with BlackRock, Fidelity, and Federated Hermes. And these are just the first-generation products. The range of tokenized yield instruments available to a corporate treasurer will only expand from here.

Run the math on a single commercial client. A $100 million corporate treasury balance earns roughly $250,000 per year at a typical demand deposit rate. That same capital, deployed across a risk appropriate mix of tokenized instruments, could earn $3 to 5 million. No CFO can justify that gap to a board. And no relationship banker, however skilled, can talk a sophisticated treasury operation out of that kind of annual drag indefinitely.

The New York Fed documented “the rise in deposit flightiness” in 2025, finding that corporate deposits have become significantly more rate-sensitive. This is the early signal of a structural shift. When every dollar can be productively deployed at all times, deposits stop being stable funding sources and become transient working capital flows, they become brief stops between yield-generating positions.

And here’s the irreversibility problem: once corporate treasurers build the infrastructure to continuously optimize across tokenized instruments, they never go back to parking cash in demand deposits. The concept of idle balances simply ceases to make sense. This is a one way door.

This doesn’t mean deposits disappear overnight. But it does mean the cost of retaining them rises structurally, permanently, and I suspect faster than most banks have modeled. The deposit franchise has always been the most valuable asset a bank owns. Its value was built on the assumption that moving money was hard enough that customers would tolerate low yields for the convenience of keeping it in one place. When that assumption breaks, what’s left?

Blind Spot #3: Onchain Collateral Is a 2026 Problem, Not a Next Decade One

It feels like most bank roadmap places onchain collateral in the final phase. “Allow digital assets to back credit facilities and loans” is positioned as the aspirational endgame, something to consider after the working group has met for three years.

The market has a different timeline.

JPMorgan’s Tokenized Collateral Network is live, processing transactions with tokenized money market fund shares as collateral across multiple buy side firms. Broadridge has incorporated JPM Coin for atomic settlement of both cash and securities legs of repo transactions. DTCC launched a platform for tokenized real-time collateral management in April 2025. Figure Technologies received the first AAA-rated blockchain securitization from S&P and Moody’s and plans to tokenize $5 billion in real estate. The CFTC recommended tokenized money market funds as eligible collateral. The Federal Reserve clarified in early 2026 that eligible tokenized securities receive the same capital treatment as their non-tokenized counterparts.

A Nasdaq survey found that 52% of global financial institutions expect to actively manage live tokenized collateral by end of 2026.

The Futures Industry Association published a landmark report on “collateral velocity” -- the concept that when collateral is tokenized, the same asset can be posted, moved, released, and reposted across counterparties in real-time rather than over multi-day settlement cycles. A DTCC pilot demonstrated that 50,000 collateral transactions costing approximately $75,000 to process traditionally could be executed for just over $1 on blockchain.

For CRE-heavy lenders, and many regional banks are exactly that, this feels existential. When commercial real estate can be tokenized and posted as smart contract managed collateral that any lender can verify in real-time, the relationship based lending model becomes a commodity. Credit approval collapses from days to hours. Collateral monitoring becomes continuous rather than periodic. Borrowers gain the ability to shop collateral across lenders instantly.

The banks that dominate CRE lending today have built that position on deep local knowledge, borrower relationships, and appraisal expertise. Those are real advantages. But when collateral itself becomes programmable, transparent, and portable, the question shifts from “who knows this borrower best?” to “who can price this collateral fastest?” That’s a fundamentally different competitive game, and the timeline for playing it is probably not the next decade.

What Gets This Right

The thread connecting all three blind spots is a framing problem. If banks are asking “how do we participate in tokenization?” That’s a reasonable question, but it’s the wrong first question. The right first question is “what happens to our existing economics when settlement is instant, yield is transparent, and collateral is programmable?”

The answer, in every case, is that the frictions banks currently monetize, settlement timing, deposit inertia, and collateral opacity, get compressed or eliminated. That’s not a reason to panic. It’s a reason to do the math.

Some banks have started to get this right. I’d bet JPMorgan didn’t build Kinexys because blockchain seemed interesting. They built it because they modeled a world in which settlement float disappears and decided they’d rather own the new infrastructure than be displaced by it. That’s the strategic posture that matters, not enthusiasm about the technology, but clarity about what happens to your business if you stand still.

For most regional and midsize banks, the honest answer is uncomfortable. Their deposit franchises are structurally more vulnerable to yield competition than maybe they want to acknowledge. Their settlement dependent revenue is more exposed than their annual strategy documents reflect. Their collateral management capabilities are further behind the market than their roadmaps suggest.

None of this means these banks can’t navigate the transition. Many of them have genuine advantages, regulatory relationships, local market knowledge, client trust built over decades, that will matter enormously in a tokenized world. But those advantages only compound if banks understand what’s actually changing. And what’s actually changing is not really about the technology. It’s about the economics.

The banks that come through this well won’t just be the ones that adopted tokenization first. They’ll be the ones that understood, early, that their existing business model was built on frictions that tokenization removes and rebuilt their strategy around what remains when those frictions are gone.