Nomura Is Shopping Laser Digital, and Four Straight European Losses Explain Why

Four consecutive quarters of red ink in Europe will change a bank's mind about almost anything. Even the crypto business it launched with a press release, a Zurich office, and a five-year plan.

Nomura is exploring a sale of Laser Digital, its digital assets subsidiary, according to Bloomberg. That's the same unit the Japanese bank stood up in September 2022 as a three-pronged bet on trading, venture investing, and asset management. Jez Mohideen ran it as chief executive. Steven Ashley, a former head of wholesale, chaired it. The pitch was institutional: regulated entities, real balance sheet, grown-up risk controls.

Three years on, the parent is testing the market. Nobody at Nomura has said the word "exit" out loud, and a review isn't a sale. But reviews have a way of becoming sales when the division next door is bleeding.

The Unit It Built

Laser Digital wasn't a token gesture. It picked up a Dubai virtual asset license, ran a Bitcoin adoption fund aimed at institutions, launched digital asset funds for professional investors, and sat inside Nomura's wider crypto footprint alongside Komainu, the custody joint venture it shares with Ledger and CoinShares.

That's a real franchise on paper. Crypto trading, venture checks, and asset management under a Swiss regulatory wrapper is exactly the structure every bank said it wanted in 2022.

Here's the problem. Institutional crypto revenue turned out to be narrow and brutally competitive. Market making compressed to a handful of firms. Options flow concentrated in the same five or six desks. Custody became a fee war. Every quarter, the addressable revenue shrank relative to the number of well-funded players chasing it.

And a bank subsidiary carries bank costs. Compliance headcount. Multiple licensed entities. Audit. Legal. Prime brokerage relationships. Treasury. The fixed cost base of a mid-size institutional crypto operation runs into the tens of millions a year before a single trade prints, and that bill lands whether or not the flow shows up.

Why the Math Broke

The obvious read is that Nomura is souring on crypto. That's lazy. The more useful read is that Nomura's European operations forced the decision, and crypto was simply the easiest thing to place in the strategic review pile.

Four straight quarterly losses in Europe is a cost structure problem, not an asset class problem. Cutting a subsidiary that sits inside the wholesale division is fast, visible, and signals discipline to shareholders without touching the core Japanese franchise. This is how bank boards work. They don't kill the experiment because the experiment failed. They kill it because something else needs to look tidy.

So what does that mean for the rest of the market?

The skew tells a different story than the headline. Institutional options flow in Asia and Europe is still dominated by a short list of crypto-native desks with crypto-native cost bases. Those firms pay traders in tokens and equity, not in Tokyo salary bands and pension obligations. When a bank-owned desk bids for the same flow, it's effectively betting on relationships and credit lines beating speed and pricing. That bet hasn't paid.

Under neutral conditions, a licensed, well-capitalized crypto options desk should be printing revenue in a market with this much realized volatility. It isn't. That's a structural problem, and it won't fix itself when the next bull market arrives.

Here's a question worth sitting with. If a bank with Nomura's balance sheet can't make a Swiss-licensed crypto desk work after three years, what's the pitch for the next one?

The uncomfortable answer is that there isn't one. Not for the trading piece, anyway.

Who Buys It

This is where it gets interesting, because there's no comparable transaction to price against. Crypto M&A in the licensed institutional space is thin, and the handful of deals that closed did so at discounts to invested capital. Buyers aren't paying for the book. They're paying for the licenses, the counterparty list, and the compliance machinery that would take two years and a fortune to rebuild.

So the realistic bidder list looks like this. Crypto-native trading firms that want a Swiss entity and an institutional Rolodex. Middle East capital with an appetite for regulated infrastructure. Possibly private equity, though PE struggles to underwrite a business whose revenue depends on volatility.

And there's a fourth option, which the market is underpricing. Nobody buys it, the trading book gets wound down, and the licensed entities get sold or shelved separately. That's not a sale. That's an orderly dismantling with a press release stapled to it.

Professional traders are pricing in something closer to that outcome than a clean strategic buyer. Not because Laser Digital is broken, but because the pool of buyers who can absorb a bank-grade cost base and turn it profitable is tiny.

My honest take: the banks got the build-versus-buy trade backwards. Komainu is the tell. Nomura's custody play works because it shares cost and risk with partners, and it sells a service every trading firm needs. Laser Digital sat entirely inside the parent, carrying all the downside and none of the shared risk. One structure survives a bad year. The other doesn't.

If Nomura wants crypto exposure in 2026, the smarter version is minority stakes in plumbing. Custody. Settlement. Prime services. Fee-based, non-directional, and indifferent to where spot trades next month. That's a far better use of bank capital than running a directional book against twenty-five-year-olds in Dubai.

This is how the smart money is positioned, whether or not it says so publicly.

Watch the licenses, not the press release. If the Swiss entity changes hands and the trading operation gets wound down quietly over two quarters, that's the sale of a regulatory wrapper, not a business. And it tells you precisely what the market believes a bank-built crypto franchise is worth.