Bitget's Open Custody Network: 3 Questions That Decide Whether the November 5 Launch Means Anything

Can you trade on a crypto exchange without actually handing that exchange your coins? Bitget says yes. On November 5 the company rolled out something it calls the Open Custody Network, a structure meant to connect outside banks and institutional custodians to its trading engine so that eligible clients can execute orders at the venue while the assets sit somewhere else.

That's the pitch. The question now is whether the paperwork behind it holds up. Because the difference between real off-exchange custody and a marketing wrapper is roughly six pages of legal terms that nobody outside the deal has read.

What Was Actually Announced

Strip the branding and three things are on the table. First, external banks and institutional-grade custodians get a connection into Bitget's trading infrastructure. Second, the stated goal is to let clients trade through the venue while relying on a separate provider for safekeeping. Third, the whole thing is scoped to "eligible users," a phrase that carries plenty of weight in crypto and usually means a narrower list than the press release implies.

What wasn't announced matters just as much. There's no published roster of participating custodians. No list of jurisdictions where the service is live. No detail on how collateral gets represented when the coins aren't sitting on the exchange, and no explanation of settlement finality when a trade clears between a venue and a third-party vault. Bitget framed it as a launch and an intended model. It didn't publish the rulebook.

Here's the memory that hangs over this. FTX collapsed in November 2022 with a hole in customer accounts that prosecutors and bankruptcy examiners put in the neighborhood of $8 billion. Mt. Gox lost roughly 850,000 bitcoin back in 2014, an amount that would be worth tens of billions at current prices. In both cases the failure wasn't a trading failure. It was a safekeeping failure. Customers could see their positions and couldn't see their assets.

That's the wound this product is trying to sell against, and to Bitget's credit, it's the right wound to address.

Why The Split Should Have Happened Years Ago

Traditional finance separates three jobs that crypto exchanges bundled into one. A broker executes the order. A custodian holds the asset. A clearinghouse stands between the two and guarantees the trade settles even if one side fails. You can lose your broker and still own your bonds. That's not a feature of good service. It's a feature of good architecture.

Crypto stripped that out because bundling was easier to sell. One login, one balance, one app, instant fills. The convenience was real. So was the concentration. When trading and safekeeping live inside the same legal entity, your exposure to that entity's technology, governance, controls, and balance sheet becomes total. A bad quarter for the venue is a bad quarter for your coins.

The industry's fault lines have been pointing here for a while. Institutional allocators who spent 2023 and 2024 watching exchange bankruptcies play out have made custody separation a precondition, not a preference. Family offices and funds with actual compliance departments want a custody agreement with a bank, a segregated account, and a party they can sue in a jurisdiction they recognize. Bitget's network is, at minimum, an acknowledgment that the buyers it wants are asking for this.

But an acknowledgment isn't an architecture.

The Three Questions Custody People Will Ask

According to two people familiar with how these arrangements get papered, the negotiation that matters is never about whether a custodian exists. It's about who can move the asset and under what conditions. Those conversations tend to stall on the same three points.

The first is legal title. Who owns the coins in the event of an insolvency, and is that claim enforceable in the jurisdiction where the custodian actually sits? The second is withdrawal authority. Can the trading venue initiate a transfer on its own, or does every movement need a second signature from the custodian? A pre-authorized transfer window is functionally the same as exchange custody with extra paperwork. The third is the venue-halt scenario. If Bitget suspends operations tomorrow, what's the retrieval path for a client whose assets are off-exchange but whose obligations were mid-settlement?

None of those questions are exotic. All of them are answerable. But answering them requires documents, and right now the market has a press release.

The operational problem underneath is genuinely hard. Fast trading needs certainty that obligations will settle. Traditional markets solve this with clearing members, margin collateral, and regulated intermediaries who absorb default risk. Crypto venues avoided the problem by holding the assets directly. Pull the assets out to an external custodian and you've to build a way to represent collateral and guarantee settlement without the coins moving. That design choice drives everything downstream, from capital efficiency to how much counterparty exposure a client is quietly carrying.

So the honest read: this is the correct strategic direction and an incomplete product announcement. Both things are true.

What To Watch Between Now And Year-End

Reading the tea leaves, the next ninety days should produce three signals that separate substance from positioning.

Watch for the custodian list. Named institutions with public regulatory registrations mean the legal groundwork exists. Vague references to "institutional-grade partners" mean it doesn't, at least not yet. Watch for the jurisdictional map. Bitget doesn't take US customers and operates under offshore licensing, so the eligible-user set is likely to be narrower than the headline suggests, and a service that only works for clients in a handful of jurisdictions is a pilot, not a standard.

And watch whether independent security assessments or segregation terms get published. That's the tell. Firms that have actually separated control of assets from control of trading are usually eager to show the mechanics. Firms that have rearranged the labels are usually not.

There's a competitive angle too. If OCN gains traction, expect the other large venues to follow within two quarters, because institutional flow is the only growth market left and nobody wants to lose a mandate over custody structure. That would be the genuinely bullish outcome. Not for Bitget's marketing department, for the market itself.

But here's the part I'd push back on hardest. Off-exchange custody isn't a security guarantee. It relocates risk. A structure can still contain settlement obligations, shared service providers, credit lines to affiliates, and pre-authorized transfer rights that only surface in an appendix. Two components can each be technically sound and still fail at the joint where they meet, which is precisely what happened when Aave restricted new tBTC supply and borrowing after a mint-and-redeem pause earlier this year. The joint is where risk hides.

So ask the boring question. Not where the assets are stored, but who can move them, under what conditions, and what happens when the venue goes quiet.

Spokespeople didn't immediately respond to a request for comment on the partner list and jurisdictional scope. When they do, the answer will be worth more than the announcement.