The public sees the spark; I track the fuel lines. On May 20, 2024, Bit Digital (NASDAQ: BTBT) drew down a $50 million loan from Galaxy Digital, pledging 49,000 LsETH—74% of its staked ether position. The company framed this as a strategic move to finance WhiteFiber, an AI infrastructure subsidiary. But the ledger doesn't lie, and it reveals a structure where the margin call clock is set to 9 hours. That is not a safety net. It is a fuse.
Context: The Corporate Pivot with a Layer-2 Twist Bit Digital, a New York-based crypto miner, has been undergoing a transformation. In Q2 2024, it reported a non-cash impairment of $46 million on its LsETH holdings, staking income of $0.9 million (down from $2.3 million in Q1), and a plan to pivot toward AI computing via WhiteFiber. The company converted 73,235 ETH into 66,192 LsETH through Stader Labs, a liquid staking derivative protocol. Then it collateralized 49,000 LsETH with Galaxy Digital to secure a $50 million term loan at 5.45% annual interest. The remaining 17,192 LsETH were kept as a buffer.
This is not a new narrative. Several BTC miners—Core Scientific, Hut 8, Marathon Digital—have announced AI pivots. But Bit Digital's approach is distinct: using LSD tokens as collateral for traditional debt, rather than selling equity or issuing convertible notes. The structure is a cross between on-chain lending and corporate finance, but with a critical vulnerability: the margin call mechanism operates on a 24-hour standard window and a 9-hour emergency threshold. I have audited liquidation processes in DeFi protocols for years, and 9 hours is laughably short for a publicly traded company. Aave gives you minutes. Corporate treasuries operate on days. This is a mismatch.
Core: A Systematic Teardown of the Leverage Structure Let me dissect the three layers of this structure: the collateral, the loan, and the downstream investment.
Layer 1: The Collateral (LsETH) LsETH is a liquid staking derivative issued by Stader Labs. It represents staked ETH plus staking rewards, but it trades at a discount to ETH due to liquidity and redemption mechanics. As of Q2 2024, the 49,000 LsETH had a book value of $105.6 million after impairment. The impairment itself is revealing: Bit Digital uses a cost-minus-impairment model for LsETH, while holding ETH at fair value. This accounting asymmetry means the company cannot recognize appreciation on LsETH, but must record write-downs when the market drops. In Q2, they wrote down $46 million, implying the original cost basis was around $151.6 million. The impairment is non-cash, but it reflects a real economic loss—and a warning sign for future volatility.
From my forensic work on DeFi liquidations, I know that LSD assets exhibit a liquidity discount during stress periods. In May 2022, when the broader market crashed, stETH (Lido's derivative) traded at a 5% discount to ETH. LsETH, with lower liquidity, could see a 10-15% discount. Bit Digital's loan agreement likely uses a price feed that reflects the market value of LsETH, not ETH. If the discount widens, the implied loan-to-value ratio (LTV) rises faster than if they had used native ETH.
Layer 2: The Loan (Galaxy Digital) The loan carries a 5.45% interest rate, which is cheap by traditional standards but expensive relative to staking yields. Bit Digital's staking income in Q2 was $0.9 million, which annualizes to $3.6 million. The loan interest is $2.725 million annually. That leaves a slim margin of $875,000—before accounting for any operational costs or WhiteFiber's funding needs. The coverage ratio is 1.32x, barely above breakeven. If staking rewards decline further (as they have been, with the ETH staking yield dropping to ~3% annualized), the coverage evaporates.
The loan terms include a standard 24-hour margin call window and an emergency 9-hour window. The exact LTV triggers are not disclosed. However, based on typical institutional lending practices, I estimate the initial LTV was between 35% and 42%, assuming the LsETH was valued at $1.2-1.4 billion at drawdown. The liquidation threshold is likely around 70-75% LTV. That means the LsETH price would need to drop by roughly 50% from the drawdown price to trigger a margin call. That seems safe, but the buffer of 17,192 LsETH (worth $27.6 million) is only 55% of the loan amount. If ETH falls 30%, the buffer would be largely consumed, and the company would be close to the margin call line.
But the real risk is time. The 9-hour emergency window is triggered when the LTV crosses a certain threshold. I have seen corporate treasuries fail to respond to 24-hour windows. A 9-hour window during a weekend or holiday is a structural trap. The company would need to either wire additional cash or transfer more LsETH to Galaxy. Wire transfers are not instantaneous; crypto transfers are, but the company's operational procedures may not be. In the 2020 DeFi crash, several protocols failed because their liquidation bots were too slow. Here, the counterparty is a human-run institution, not a smart contract. Galaxy Digital has discretion on whether to liquidate partially or fully. That introduces counterparty risk: if Galaxy faces its own liquidity issues, it might liquidate aggressively.
Layer 3: The Downstream Investment (WhiteFiber) The $50 million flows into WhiteFiber, an AI infrastructure company in which Bit Digital holds a majority stake. The structure is a delayed draw term loan facility, initially $100 million, expandable to $150 million. Bit Digital is essentially lending to its own subsidiary. The interest rate on this internal loan is not disclosed. There is a structural risk: if WhiteFiber fails to generate sufficient returns, Bit Digital is left with a loan that is underwater. The loan is secured by a pledge of Enovum NC-1 Topco shares and a parent guarantee from White Fiber Operating Partnership. This is a multi-layered leverage chain, where each layer can amplify defaults.
Contrarian: What the Bulls Got Right To be fair, the bulls have a point. This structure allows Bit Digital to retain exposure to Ethereum's upside while financing an AI pivot. Selling 49,000 ETH would have been a taxable event and would have removed the company from the crypto narrative. By using LSD collateral, they keep the ETH on the balance sheet (sort of) and borrow at a low rate. The staking income, though small, at least offsets some interest. The WhiteFiber investment, if successful, could provide a new revenue stream separate from mining. The share buyback signal from CEO Sam Tabar indicates management believes the stock is undervalued—a classic contrarian signal.
Moreover, the margin call terms are not unusual for institutional loans. Galaxy Digital is a sophisticated lender; they have an incentive to work with borrowers rather than liquidate. The 9-hour window is likely only for extreme scenarios, and the company has a buffer of 17,192 LsETH. The impairment is non-cash and may reverse if the market recovers. The accounting asymmetry is a reporting artifact, not a fundamental flaw.
Takeaway: The Clock Is Ticking Bit Digital's structure is a bellwether for corporate crypto leverage. If ETH holds above $3,000 and WhiteFiber starts generating revenue, the model may prove viable. But if ETH drops 20% and the LsETH discount widens, the 9-hour window becomes a trap. The market is not pricing this tail risk adequately. The stock's valuation still reflects the sum of ETH holdings plus mining operations, with little discount for the leverage. I have seen this before: in 2022, similar structures unraveled when the music stopped.
Investors should monitor three things: the ETH price, the LsETH discount to ETH, and any filings about WhiteFiber's revenue. The ledger doesn't forget. And the clock is ticking.