Baton Systems

It is a Tuesday morning at a tier-one bank. A trader books an FX deal worth nine figures. Twenty years ago, this kicked off two days of phone calls, faxes and SWIFT messages. Today, on Baton's rails, it settles before the coffee gets cold.

Baton Systems is not a bank. It is not a custodian, a clearing house, or a stablecoin. It is, in the unromantic phrasing of its own engineers, "the plumbing." That plumbing now carries roughly thirty billion dollars in value every day, between counterparties that used to need three days and a prayer to settle the same trade.

The company sits in an unremarkable office park in Fremont, California - which is precisely the kind of place serious infrastructure tends to live. There are no neon signs out front. No one is dressed for a TED talk. Inside, a few dozen engineers and ex-bank operations leads are mostly arguing about ISDA's Common Domain Model and whether a tri-party collateral movement should be modelled as one event or two.

The problem they saw

To understand Baton, you first have to accept an awkward fact about modern finance. The trading floor is sci-fi - microsecond latency, machine learning, dark fibre between exchanges. The back office, where the trade actually settles, is closer to the 1970s. Cash sits trapped in nostro accounts. Margin calls move on email. FX settlement, the bedrock of cross-border banking, still hides a small but real risk of one side paying and the other defaulting before delivering. The industry has a polite name for it: Herstatt risk. The polite name is older than most traders.

For most of the last decade, the proposed cure was the same answer to almost every question in 2017: a blockchain. Big banks announced consortia. Consortia produced press releases. The press releases produced more consortia. Very little money actually moved.

Baton's founders, perhaps wisely, took a different read on the problem. The blockers, they decided, were not cryptographic. They were operational. Banks could not rip out forty years of accumulated COBOL, Murex, Calypso, and an unspeakable amount of Excel just because somebody had a whitepaper. Whatever fixed post-trade had to talk to the legacy stack, not replace it.

The founders' bet

Arjun Jayaram is not the founder you'd expect to fix bank settlement. He had been running anti-spam at Twitter. Before that, the founding CTO of Compass Labs. Before that, Yahoo. Capital markets veterans tend to view such a CV with suspicion - a healthy reflex, given how many Silicon Valley founders have wandered into finance, declared the entire industry broken, and wandered back out twelve months later.

Jayaram stayed. He spent the first two years of Baton - briefly known as Ubixi - not building product, but sitting with operations teams inside banks asking, repeatedly, an unfashionable question: what would you actually use? The answer was rarely "a new asset class." It was almost always "real-time visibility, real-time control, and please do not make me rewire my settlement engine."

So Baton built a shared permissioned ledger. Not a public blockchain. Not a token. A workflow layer that connects to existing bank systems, mirrors their state in something close to real-time, and lets multiple counterparties agree on a single version of the truth without anyone giving up control of their own books.

The product

The Core Platform is the ledger and the rails. On top of it sit three things that any post-trade operations lead can name in their sleep.

What makes the product land in a procurement meeting is not the technology stack. It is the integration story. Banks do not have to migrate. The Core platform sits next to whatever ancient settlement engine they already trust, listens, mirrors, and offers the counterparty a synchronised view. The legacy stack stays. The latency disappears.

A short, slightly unfashionable history

The proof, in numbers

Bank infrastructure is sold on receipts, not vibes. Baton's receipts are unusually clean for a Series A company.

The customer list is the louder data point. J.P. Morgan. Citi. HSBC. Wells Fargo. These are not institutions that pilot something because it has a clever name. They run multi-year procurement gauntlets and demand SOC-2-and-then-some before they let a startup touch a single nostro balance. Baton survived them.

The mission

If you read Baton's own materials, the mission sentence is exactly what you'd expect: "Post-trade. Redefined." Read past the marketing and what they actually mean is narrower and more useful. They want settlement to look like the rest of the internet - on demand, real time, observable, atomic. They want collateral to move at the speed of the margin call rather than the speed of the email chain. They want capital to stop sitting in transit accounts earning nothing.

None of this is glamorous. None of it will trend. It is the kind of work that, when done well, becomes invisible. The trader books the trade. The trade settles. Nothing breaks. Nothing makes the news. That, in the post-trade world, is the highest possible compliment.

Why it matters tomorrow

Three forces are converging on Baton's lane. First, regulators are moving FX and securities settlement to T+1 - a move that, in the absence of better plumbing, just compresses the same painful workflow into half the time. Second, central banks are experimenting with tokenised cash. Third, every tier-one bank now has a "real-time intraday liquidity" line item somewhere in its risk committee minutes. All three problems have the same answer: a settlement layer that can hold state across counterparties, agree on it, and act on it without waiting for someone to press a button at 5pm London time.

That is, conveniently, the thing Baton already built.

The competitive picture is interesting. The incumbents - DTCC, CLS, Broadridge - are big, trusted and structurally slow. The newer DLT networks (Fnality, Partior) overlap with Baton's worldview but operate as networks, not as platforms that sit inside an individual bank. Baton has chosen to do both: run the rails between banks and live inside each bank's operations stack. It is a harder thing to sell. It is also a harder thing to dislodge.

Back to that Tuesday morning

Same trader. Same nine-figure FX deal. The phone does not ring. No one chases anyone for a payment confirmation. By the time the trader has answered their next Slack message, the trade has netted, settled PvP across a shared ledger, and the cash has landed in two different banks' books at almost the same instant.

There is no fanfare. There is, importantly, no Herstatt risk. The back office stops calling the front office to chase breaks. The risk committee looks at intraday liquidity on a dashboard rather than reconstructing it from yesterday's data.

This is what Baton Systems is for. Not a revolution. Not a manifesto. Just settlement that finally behaves like every other piece of software the rest of the world has taken for granted since 2009.

Quiet, useful and trillions deep. The plumbing is finally worth talking about.