Filling the SVB and Silvergate void
In the space of three days in March 2023, the American technology and crypto sectors lost most of their banking system. was seized on 10 March 2023 after depositors tried to pull about $42bn in a single day, roughly a quarter of its deposits. Signature Bank was closed by New York regulators in the same panic, having suffered $18bn of withdrawals on 10 March. Silvergate, the crypto sector's clearing bank, had already begun a voluntary liquidation after its deposits fell by around $8.1bn in roughly three months following the collapse of FTX; it relinquished its bank charter in July 2024.
What followed for crypto firms was the period the industry calls debanking: a tighter US banking environment in which companies struggled to find reliable banking partners at all, with some moving operations offshore or to smaller institutions. The costs were operational rather than dramatic. It is hard to run a business when nobody will hold your operating account.
Erebor is the answer to that, and it is explicit about it. Founded in 2025 by Palmer Luckey, the founder of Anduril and Oculus, and Joe Lonsdale, co-founder of Palantir and founder of the venture firm 8VC, it is positioned as a successor to fill the financing void left by SVB for frontier tech. Luckey has said so directly on the record, in an interview titled, with no ambiguity whatsoever, "Why I Started My Own Bank". The bank is headquartered in Columbus, Ohio, and its clients are drawn from the technology and defence firms of California.
One half of the founding story does not check out. The sector-wide debanking of crypto companies is well documented; we have not seen equivalent evidence of a comparable banking freeze-out for defence-tech firms. That part of the pitch is asserted rather than demonstrated.
The Tolkien naming metaphor
Erebor is Tolkien's Lonely Mountain: the dwarven treasure-hoard of The Hobbit, where the dragon Smaug sits on the gold until somebody comes to reclaim it. The name is a translation of "Lonely Mountain" in Tolkien's invented linguistics rather than decorative fantasy noise.
The founders have form. Palantíri are the seeing-stones of The Lord of the Rings, associated with vision at a distance and knowledge. Andúril is Aragorn's reforged sword, associated with restoration and righteous force. The pattern is to pick a Middle-earth name that matches the company's image of itself, and on that basis Erebor is the most candid of the three: a fortress vault holding a mountain of concentrated wealth, guarded, in one place, deep inside a rock.
Whether anyone in the naming meeting recalled that the defining feature of Erebor in the story is that the hoard attracts something enormous and then everything burns is not recorded.
Regulatory approval and rapid growth
Erebor moved through the American chartering process at a speed that de novo bank applicants do not normally enjoy. It filed its national bank charter application with the Office of the Comptroller of the Currency on 11 June 2025. The OCC granted preliminary conditional approval on 15 October 2025. FDIC deposit insurance approval followed in December 2025, and the final national bank charter arrived on 6 February 2026, less than four months after the conditional one. The bank opened in February 2026; one report gives 8 February, others leave the date vague.
The money arrived on a similar schedule. A $350m round in December 2025, led by Lux Capital with Founders Fund, 8VC and Haun Ventures participating, valued the bank at $4.35bn. Andreessen Horowitz is also among the backers, as is Peter Thiel. Private-market trackers cannot agree on the total: one estimates $635m raised, another $575m, which tells you how much of this is visible from outside. As of the August 2026 reporting, Erebor is in talks to raise a further $1.5bn at a pre-money valuation of $8bn, with some reports putting the figure at $9.5bn. Deposits had reached $4.6bn on that same August 2026 reporting, and the bank added hundreds of customers in the second quarter.
Leadership is reported less cleanly than the funding. Owen Rapaport and Jacob Hirshman appear as chief executives in one company profile; other named founders include Trevor Capozza and Aaron Pelz. Luckey and Lonsdale are founders and backers rather than operators.
Why it matters to this crash
Erebor sits precisely on the seam we track under : a federally chartered, deposit-insured national bank whose product design assumes blockchain settlement. It plans to use blockchain rails so clients can their money at any time, and it is discussed in the digital-asset press as a bank built for stablecoins. What it will actually custody, and which tokenised deposit products it will offer, has not been described publicly in any detail we can source. That gap is the point: this is a chartered bank whose most consequential design decision is not yet documented anywhere a depositor could read it.
It is not happening alone. On 20 August we wrote about Franklin Templeton putting its tokenised money-market fund inside conventional ETFs and mutual funds, as a holding or as collateral, as early as the fourth quarter — the first US regulatory clearance of its kind, by the firm's own account. Sandy Kaul's stated rationale was to "manage more precisely, capture more of the yield, better and more tightly manage how much cash liquidity they have to hold." That is a real efficiency. It is also how a settlement technology stops being a novelty and becomes a dependency: instruments designed to be movable at any hour turn out to be instruments that can leave at any hour, and the people holding them did not necessarily ask for them.
What would make this dangerous
The single most useful number here is $42bn in one day. That is what depositors tried to pull out of SVB on 9 March 2023, and they did it through mobile banking apps, Twitter and group chats — a run conducted by push notification, the fastest large-bank run in American history. Even that run was slowed by the physical infrastructure of banking: wire cut-offs, operating hours, the mechanics of the payment system.
Erebor is taking the same depositor profile — founders, investors and their portfolio companies, homogeneous, densely networked, many of them close to its own backers, reading the same group chats — and removing what remained of the friction. A bank whose clients can move balances at three in the morning on a Sunday has no last brake. We made this argument on 20 August and nothing since has changed it.
Four observable things to watch, in rough order of how much they should worry you. First, the asset side, which is what actually killed SVB and which nobody outside the firm can currently see: mismatch plus concentration is fatal, while concentration alone is survivable if the assets are Treasury bills. There is no public Erebor balance sheet to judge, and the bank has not been through a cycle. Second, deposit concentration: $4.6bn of deposits drawn from one boom-and-bust sector is a different animal from $4.6bn drawn from everywhere, and lenders serving a single industry have a bad history. Third, the settlement architecture actually going live — the moment tokenised balances can be moved outside banking hours without human intervention is the moment the SVB timeline compresses from a day to an hour. Fourth, any funding round that fails to close, or closes below the $8bn to $9.5bn being discussed, in a depositor base that reads the funding press as closely as it reads its own balances.
None of that is a prediction. It is a list of things that would be visible before the interesting part.