Tokenized Stocks Need Two Transfer Agents and a Registrar. That's the Real Bottleneck.
Everyone's arguing about throughput while the actual constraint on tokenized equities sits in a shareholder register run by a company most crypto people have never heard of. Galaxy's detour through two transfer agents shows exactly how much plumbing it takes to put a stock onchain, and why the winners won't be the fastest chain.
Tokenized stocks don't have a throughput problem. they've a back-office problem, and the back office is winning.
Look at what it actually took for Galaxy to get an issuer wired into a token. Not one transfer agent. Two. Equiniti runs the shareholder register underneath the whole thing, and that register had to agree to move before a single token could exist. Two hops through the plumbing just to connect one issuer to one new system.
That's the story. Not block space. Not finality. A registrar.
The Wrong Bottleneck
Ask a crypto engineer what's holding back tokenized equities and you'll get an answer about latency, or fees, or data availability. All of it's stale.
Ethereum's Pectra upgrade on May 7, 2025 pushed the blob target from 3 to 6 and the ceiling from 6 to 9, and the L2s have been swimming in cheap data space ever since. Solana clears thousands of transactions a second without breaking a sweat. Transaction costs on the major rollups sit in fractions of a cent. Throughput is table stakes now. It stopped being the interesting question around the time Dencun landed in March 2024.
So here's the thing nobody wants to say out loud. A tokenized share isn't a share. It's a claim on a share, and the share itself lives in a register maintained by a transfer agent who has no particular interest in your chain's roadmap.
That register is the authoritative record of who owns what. Transfer agents are the only parties who can add a line to it. DTCC has been settling US equities at T+1 since May 28, 2024, and it moves something in the neighborhood of $3 quadrillion a year through that machinery. None of that's going to be replaced by a sequencer. The real bottleneck is that a token can't be created until a human at a registrar updates a database, and every additional hop in that chain is a new reconciliation surface where things break.
Nobody cares about infrastructure until it breaks. This is the part that breaks.
Two Agents, One Register
Why two transfer agents? Because the issuer's existing agent doesn't do crypto, and the crypto-native side doesn't have a seat on the register. So you insert a second agent in the middle, and the token sits on top of that. It works. It's also three parties deep before you get to the actual asset.
Each hop costs something. Corporate actions have to propagate twice. Proxy voting has to be mapped through two systems that were built in different decades. Dividends get declared on a share that's held in custody and mirrored to a token that's held in a wallet, and someone has to guarantee those two records never drift. That's the tradeoff, and it's not a technical one you can optimize away with better tooling.
And this is exactly the layer Bullish has been recruiting Wall Street people to attack. That's telling. The money isn't going into making tokens faster. It's going into making the connection to the register credible enough that an issuer will sign off on it.
Which makes sense, if you think about what an issuer actually wants. They want distribution and they want liquidity. They don't want a second class of shareholder they can't audit, and they definitely don't want to explain to their counsel why a token contract is now part of their cap table. The transfer agent is the gate that keeps that from happening. Whoever owns the integration owns the market.
And regulators are moving, slowly. The SEC's posture shifted noticeably in 2025, with staff guidance and a friendlier commission making tokenized securities look less like a litigation risk. That helps. It doesn't remove the registrar.
What the Bears Get Right
Let's talk blob economics for a second, because the counterargument is stronger than the bulls admit.
The bear case goes like this. Tokenized equities are a demand problem dressed up as an infrastructure problem. Almost all the volume so far sits offshore, sold to European and Asian retail through apps that treat the token as a wrapper with thin secondary liquidity. Backed's xStocks launched in June 2025 with a few dozen names. Robinhood rolled out tokenized US stocks to EU users the same month. Real numbers, yes. Real depth, not really.
If the volume never shows up, none of the plumbing matters, and the two-transfer-agent route becomes an expensive science project. And the bears have a point about incentives. Why would a large issuer hand a slice of its register to a crypto intermediary when the upside is a few million dollars of offshore retail flow and the downside is regulatory attention plus a shareholder base they can't cleanly service?
That's the honest objection. It's not that the tech can't work. It's that the addressable market might stay small enough that the integration cost never pays back.
But here's where I think that argument breaks down. The cost of connecting to a register isn't paid once per issuer. It gets paid once per transfer agent, and then it gets reused. Equiniti doesn't run one company's register. It runs a lot of them. Once the pipe exists and the reconciliation is audited and the legal template is set, the marginal cost of connecting issuer number two, three, and four collapses.
The Verdict
I'll pick a side. The register integration is the moat, and it's going to consolidate hard.
Within a couple of years I'd expect the transfer agent layer to look like a handful of players with crypto rails bolted on, and the tokenization platforms that matter will be the ones with signed agreements rather than the ones with the best docs. That's a boring prediction. It's also the one the evidence supports. Galaxy routed through two agents to get one issuer live. That's a real cost paid up front for a reusable asset.
The losers here are the wrapper tokens with no register-level linkage. If your token isn't backed by a share that a named transfer agent will acknowledge, you're not tokenizing equity. You're selling a derivative with extra steps, and the moment enforcement gets serious about that distinction, the market for it evaporates.
So who wins? The registrars, quietly, and the platforms that got to them first. Not the chains. The chains already did their job.
The scaling roadmap just got more interesting, and it has nothing to do with block space.
Explore More
Key Terms Explained
Coinbase's Layer 2 blockchain built on the OP Stack (Optimism's technology).
A bundle of transactions that gets permanently added to the blockchain.
Who holds and controls your crypto assets.
The guarantee that transaction data is published and accessible to anyone who needs to verify it.