Consider what it means when the most heavily regulated company in American crypto reduces its headcount for the second time in three years. Anchorage Digital — the first federally chartered digital asset bank in the United States — has reportedly trimmed 17% of its staff. If the roughly 400 employees its chief executive, Nathan McCauley, cited in Congressional testimony in February 2025 still held, that is about 68 people. The company has not commented publicly. But the precedent is already on the record: in 2023, Anchorage cut 20%. Two reductions, three years apart, in the same direction. That is not a company trimming excess. That is a company discovering the excess grew back.
Anchorage is not a startup with a token and a dream. It holds an OCC charter, custodies institutional assets, and its chief executive has testified before the United States Congress — a position of rare regulatory legitimacy. Which makes it the least likely name on a list that now reads like a roll call: Gemini down 30%, Dune down 25%, Luno down 20%, Coinbase and Bitwise down 14%, Crypto.com down 12%. And then the harder entries, the ones that are not reductions but exits: BitMEX, the derivatives venue that once defined offshore leverage; Lisk, an aging sidechain project; Blast, the L2 that bought its way to prominence with yield and airdrops.
Challenger, Gray & Christmas reports that American technology companies announced 10,799 layoffs in September alone, up 77% month over month, and 165,925 year to date, up 54%. Crypto accounts for roughly 29% of that total. Hold that number. We will come back to it, because it does not match this industry's share of the technology economy, and mismatches like that are where the real information lives.
Here is the part most coverage skips. Layoffs are described, correctly, as a lagging indicator — a consequence of a downturn already priced in. That framing is comfortable and incomplete. A single reduction is a lagging indicator. A second reduction at the same firm is something else: evidence that the first restructuring failed to establish a sustainable cost baseline. When I spent 600 hours manually auditing Aave V2's early interest-rate logic in 2020, I found three errors that would have compounded silently — each individually defensible, collectively fatal. My subsequent manifesto argued that verification must extend to the social contract, not only the code. Cost structures behave the same way. The 2023 cut was supposed to be the correction. It wasn't.
The moat that wasn't. The industry has spent years telling itself a story about regulatory legitimacy as a competitive shield. Anchorage holds the scarcest license in the sector. It cut anyway. A charter is permission to operate; it is not a hedge against demand collapsing. Regulatory standing is necessary. It is nowhere near sufficient, and treating it as sufficient is how institutions confuse compliance with durability.
Debanking is the part no bull market repairs. In 2023, McCauley attributed part of that year's reduction to being cut off from banks. Read that again. A federally chartered digital asset bank, describing difficulty accessing the banking system. If that pressure recurs, the current reductions carry a policy component that price recovery cannot fix. Market cycles heal. Correspondent banking relationships, once severed, do not heal on their own.
The cull is the signal. Blast's shutdown matters less for its own users than for what it says about the L2 thesis. An entire generation of rollups competed on emissions and points; the ones without durable demand are now exiting. Lisk's closure removes another mid-tier chain. BitMEX's shutdown closes a chapter of offshore derivatives. This is not collapse. It is consolidation, and consolidation is short-term brutal, medium-term clarifying.
Attention is migrating. Dune's 25% reduction is paired with a stated pivot toward AI agents and institutional data. Read that as a directional indicator: capital and talent are moving from crypto-native data work toward AI-adjacent infrastructure.
The 29% figure deserves scrutiny. Crypto's share of American tech layoffs sits far above its share of the tech workforce. Two explanations compete. Either the industry is contracting harder than the broader sector, or the classification is loose and every fintech with a blockchain slide is being counted as crypto. Both readings are useful. The first says the downturn is real and disproportionate. The second says our data infrastructure is as sloppy as our due diligence.
One structural note. The entities in this story that file notices, brief employees, and appear before Congress are the ones with legal personality. Code is law, but ethics is soul — and a legal person is what makes ethics enforceable. The industry's parallel affection for structures with no legal status, where liability lands on individuals rather than entities, would not even generate a headline like this one.
The consensus reading is bearish, and I want to push against it — partly.
There is a genuine bull-market interpretation available here, and it deserves a hearing. Companies cutting 14% to 25% while pivoting toward AI are not necessarily retreating; they may be reallocating. Headcount reduction is also a capital decision, and in a market where prices have recovered, the cost of carrying a large human layer has risen relative to automating it. On that reading, these are repositioning cuts, not distress cuts. Dune fits it neatly. So does Coinbase, which has cycled through this motion before.
I find that argument partly persuasive and partly self-serving. It explains Dune. It does not explain Anchorage cutting twice.
The blind spot is measurement. We evaluate this industry by price, by TVL, by active addresses — all of which can be manufactured, rented, or inflated. Transparency is the oxygen of trust, and payroll is the one metric nobody can fake for long. When the custodians of institutional assets are reducing staff, the institutional adoption narrative is running ahead of its own evidence. Watch the payrolls, not the press releases.
Decentralization without stewardship is just abandonment. The question for the next twelve months is not whether crypto survives a hard cycle — it always has. The question is whether the layer that holds other people's assets can keep its promises while shrinking. If the answer is no, the charter was never the moat we believed it was.