Markets may be irrational about what things are worth, but they always strive for efficiency in how things get done. I’ve watched that instinct play out for decades, as every layer of friction and every unnecessary intermediary gets stripped away.
Decimalization in 2001 cut the minimum price increment from a teenie (six and one quarter cents) to a penny, and speed became the edge. Co-location and algorithms followed with firms placing servers inside exchange data centers, where software could make decisions in microseconds. Speed became a business model, and within a decade high-frequency firms accounted for more than half of US equity volume. Spreads narrowed, competition increased and the cost of trading plummeted as a result.
The same instinct for efficiency made the tokenization of assets inevitable from the moment Satoshi Nakamoto published the Bitcoin white paper on October 31, 2008. It has taken time, but it’s happening. Robinhood and Kraken already offer tokenized equities that trade around the clock, and ICE is building toward the same through its venture with OKX and a new NYSE venue that never closes. More than half of all tokenized-stock trading already happens while US markets are closed.
Think about what that means for dollar settlement.
A fund can sell tokenized shares on a Sunday and the shares settle in seconds, but the dollars from the sale can’t keep pace. Sent over traditional wires, the earliest would be Sunday night when Fedwire reopens, but more likely it would take till Monday morning for client access. So for now, ICE, OKX and others are turning to stablecoins, trading the delay of bank dollar settlement for the speed of a claim on a non-bank issuer.
It’s a trade-off that suits a trader. It doesn’t suit a credit committee at scale.
Netting got us this far
Because money couldn’t move continuously, markets learned to batch trades, net them down and settle later, and it has worked remarkably well. On an average day in 2024, DTCC’s National Securities Clearing Corporation (NSCC) netted more than $2 trillion of trades down to about $34 billion of payments, a 98% reduction in the cash that has to move.
The cost for this efficiency is time.
Today’s settlement cycle is T+1, but the rails carrying the money still close. Fedwire operates 22 hours on business days, closes on Saturday, most of Sunday and on holidays, and yet most banks support only business hours - a much narrower window. This means that a market trading continuously faces an average wait of more than 14 hours, extending to around 60 hours over a normal weekend and even more over a holiday weekend.
During that time, capital sits idle and every participant still carries exposure to the clearinghouse and to its counterparties. Most importantly though, a dollar that hasn't settled can't be redeployed. What a 24/7 market needs is settlement that is instant, final and gross, so each dollar is free to work again the moment a trade completes.
I call these high-frequency dollars.
The case for netting has always been that less money needs to move. That only matters when money is slow. If a dollar can settle many times a day, it can do the same job that used to take many more dollars sitting in a net. The advantage isn’t the speed of the transfer. It’s how much more work the same dollar can do.
Take a book that needs to turn its capital over every 12 hours. Settle instantly, and the same dollar can work 730 times a year. Leave it waiting for today’s average time of closer to 14 hours, and it can only work about 337 times. Instant settlement turns that dollar about 2.2 times more often. Over a weekend, when the wait stretches to 60 hours and the same book falls to about 120 turns. Instant settlement puts that dollar to work almost six times as often.
For a firm settling $1 billion a day, the year-round average wait time ties up just under $400 million on an average day. Redeployed, those funds provide significant additional earning potential when measured against what trading capital normally earns.
We know this works because we’ve watched it happen. When US equities moved from T+2 to T+1, NSCC’s clearing fund fell 29%, returning billions to member firms.
And trading is only the start. Trillions of dollars sit prefunded in correspondent accounts around the world, waiting for cross-border payments that move on banking hours. Free the dollar share of that, and the liquidity released would be measured in trillions.
Netting was how markets conserved slow money. With high-frequency dollars, velocity lets the same dollar do all the work without the wait and without the unsettled credit exposure.
Speed has come with a catch
Stablecoins proved the demand for faster dollars, even at 2 a.m., and even across borders. They’re instant, always on and open, and credit where it’s due: they’ve taken the dollar to places banks never reached.
Yet even with 24/7 equities trading live overseas and on its way in the US, there are only about $300 billion of them. Fedwire moves fifteen times that every day. So why is the float so small?
The trouble for stablecoins starts when you put them in front of a credit committee.
At $1 million, exposure to the issuer is a rounding error. At $1 billion intraday, it’s an exposure to a non-bank with no lender of last resort. That’s when the truth comes out: this isn’t bank money. It’s a claim on a non-bank issuer, not a deposit, without the legal standing of bank money. Accountants are still deciding whether stablecoins count as cash, and bank capital treatment depends on the stablecoin’s structure with those that fail tests facing the same punitive 1,250% risk weighting applied to unbacked cryptoassets.
Notice what ICE has done. The same company bringing tokenized NYSE stocks to OKX’s users, funded in stablecoins, chose something different for its own clearinghouses. Where margin calls must be met and risk is managed around the clock, ICE is working with BNY and Citi on tokenized bank deposits. Stablecoins are good enough for the trading front end. For the risk at the center, ICE wants bank money.
Tokenized deposits put bank money on-chain, but they bring two problems of their own.
The first is reach. Most tokenized deposits only move between a bank’s own customers. Interoperability is coming through bank-to-bank acceptance, and that will be progress. But there’s no always-on, tokenized central bank money for those tokens to settle against, so interoperability must run on credit. Yes, to the client, it will appear as if the money moves instantly, but behind the scenes, settlement still happens on the old rails with credit extended between the banks and fees to cover that credit delivered to the client as a cost.
The second is what backs the token. At a fractional-reserve bank, the answer is a loan book you have no view into. That has been manageable while money moves on business days, because withdrawals are slow and the Fed’s backstop keeps the same hours as the bank. Make those deposits move at 2 a.m. on a Sunday, and a run can start at any hour.
The digital dollar a credit committee would approve
If tokenized dollars represent “deposits” the question to ask is what risks come with them, and what backs them: loans, or cash and Treasuries.
Only a fully reserved, tokenized bank dollar meets all four tests. Because it’s backed by cash and short-term Treasuries rather than loans; it’s the closest thing to central bank money, and holding it doesn’t mean betting on a loan book.
This isn't theoretical.
At N3XT, we run our core banking on a blockchain. Clients already move dollars between parties on our private permissioned chain, instantly and around the clock. That dollar is a deposit at a fully reserved bank, backed 1:1 by cash and short-term Treasuries.
High frequency dollars benefit the many, not the few
High-frequency trading was a race among a small number of firms to be faster than each other. High-frequency dollars help every balance sheet that holds idle cash as a buffer against slow money and transfer windows that close.
- Trading firms - turn capital over many more times a year with the same balance sheet.
- Clearing houses - meet a Sunday margin call on Sunday, while also returning excess collateral to members faster.
- Multinational corporations - gain full control of working capital. No more need to fund a Monday payroll move on a Thursday; they can invest in growth.
- Global trade - stops parking dollars in transit and uses them to fund the next shipment.
I’ve seen firsthand how markets are transformed when friction comes out of the asset. But that was only half the job. The bigger change comes from taking friction out of the money.
Until now, the dollar leg of a trade has been safe but slow on bank rails. Stablecoins offered speed, but only by giving up the protections of bank money. High-frequency dollars end that trade-off, and the firms and treasurers who hold them will turn their capital faster than everyone else, without taking on new lending risk.
That’s the dollar we’ve built at N3XT, and the one we’re now bringing to capital markets.
