BitGo absorbs NYDIG's derivatives desk to sell institutions the whole trade lifecycle

The seller's arithmetic is the more useful half of this. NYDIG is keeping the vertically integrated power generation, the bitcoin mining and the data centre development ^[5], and it sits inside Stone Ridge Holdings Group, whose subsidiary operates assets tied to roughly 3% of US natural gas production ^[14], which means the desk it just handed over was competing for internal capital against megawatts and lost. Thirty people ^[3] is a small enough headcount that the transaction reads less like a sale than like a decision not to fund a business line any more.

On the buying side the numbers are strange in an instructive way. BitGo booked $4.329bn of revenue in the quarter to June 30, up 79.6% year on year ^[11], against 5,833 customers ^[12], which is about $742,000 of revenue per customer per quarter ^[23]; a custody fee schedule does not produce that, so most of the line is trading flow booked gross, and buying a derivatives and financing desk is an attempt to put margin behind volume that already runs through the pipes. This is probably wrong, but the $19m net loss is the tell: strip out the $18.8m unrealised mark on BitGo's own digital assets ^[13] and the operating business landed within roughly $200,000 of breakeven ^[22]. From there you buy revenue that carries a spread.

The demand behind it is real and smaller than the language around it. CME's July average daily volume hit a record 27 million contracts across all asset classes ^[16], of which crypto averaged 237,000 a day and about $10.3bn of notional ^[17], or 0.88% of the contract count ^[24]. Galaxy Research put crypto-collateralised lending at $56.16bn at the end of the second quarter, down 16.78% and declining gradually rather than collapsing ^[18], which implies roughly $67.5bn three months earlier and about $11.3bn of book that simply left ^[25]; BitGo's own $65.2bn of Assets on Platform ^[15] is 1.16 times the entire lending market it has just taken exposure to ^[26]. The Fireblocks survey that will get quoted at every conference this autumn (638 decision-makers, 88% with budget committed or planned for 2026, 53% of those who had sized it at $1m or more ^[19]) works out, taking those percentages at face value, to something like 300 institutions and $300m of floor spend ^[28]. Surveys measure budget lines; custodians get paid on flow.

There are at least two other readings. One is that consolidation is genuine, that institutions do want custody, trading, financing and settlement from a single counterparty as Mike Belshe says they do ^[6], and that a federally chartered trust bank inside a listed company ^[7] is the shape that wins. The other is that best execution and counterparty conflict discipline keep serious allocators separating their venue from their custodian, in which case BitGo has bought a client book that leaks as relationships get repriced. The falsification test is unglamorous: the acquired relationships should show up as a step change in customer count and Assets on Platform, and if the next disclosure has customers still compounding at something close to 26.2% ^[12] with no visible trading uplift, this was a hiring round with a press release. (One source gestures at an April 23, 2026 Bank for International Settlements Financial Stability Institute paper on the risks of exactly this integration, but the passage supplied to us breaks off mid-sentence ^[21], so I will not tell you what it warns.)