Bit Digital Posts Revenue Decline as Ethereum Staking Yields Compress to Post-Merge Lows
Nasdaq-listed Bit Digital reported first-quarter 2026 revenue of $27.9 million on May 15, a 13.6% drop from Q4 2025, as falling Ethereum network yields and softer GPU cloud demand squeezed the company's two main income sources.
The results underscore a structural problem spreading across every institution that stakes ETH for yield: an Ethereum validator set that has grown past 1.1 million active participants is now generating rewards thin enough to hurt balance sheets. For DeFi developers and protocol treasuries in South Asia and Africa, that same pressure applies regardless of company size.
The staking yield problem
Ethereum staking revenue fell 29.4% quarter-on-quarter to $2.3 million, even though ETH staking carries Bit Digital's highest gross margin at 94.7%. The decline is not a company-specific failure. According to network data, roughly 35.86 million ETH is currently staked, representing about 28.9% of total supply. The average annualised consensus-layer yield has compressed from a post-Merge high of around 5.3% to a current range of 2.84% to 3.3%. Base rewards without MEV boosts are running as low as 1.87% in some estimates. Net staking flows turned negative in late November 2025, meaning more ETH is now leaving staking contracts than entering them.
In response, Bit Digital repositioned approximately 70,000 ETH from native staking into liquid-staked ETH (LsETH). As of April 30, 2026, the company still held 60,677 ETH natively staked, meaning the shift was partial rather than a full exit from direct validation. Liquid staking tokens, such as LsETH and Lido's stETH, allow holders to redeploy staked assets within DeFi applications rather than locking them in a validator indefinitely. Liquid staking protocols now control roughly 31% of all staked ETH and account for roughly 40% of DeFi's total value locked.
The shift gives Bit Digital more treasury flexibility and access to diversified yield sources, though it also signals that straightforward protocol staking alone may no longer cover operational costs at scale.
Balance sheet under pressure
Beyond the staking story, the numbers show stress at the asset level. The company holds 155,444 ETH with an average acquisition cost of roughly $3,045 per token. At the March 31 closing price of $2,104, that position carried an unrealized loss of approximately $941 per ETH, or around $144.6 million in total. Total digital assets on the balance sheet declined by $120.7 million quarter-on-quarter to $295.0 million. Cash and equivalents fell from $118.4 million at year-end 2025 to $79.5 million, and convertible notes outstanding stand at $334.2 million, a debt load more than four times the company's current cash position that adds a significant leverage overhang at a time of declining asset values.
The net loss for the quarter was $146.7 million, an improvement from the $185.3 million loss in Q4 2025, but adjusted EBITDA was negative $9.4 million.
GPU cloud revenue, the company's largest single segment at $16.8 million, declined 13.1% from the prior quarter. Colocation services were the one bright spot, growing 23.9% to $4.8 million. Digital asset mining revenue, now a small legacy line following Bit Digital's full exit from Bitcoin mining in mid-2025, fell 32.9% to $3.7 million. Since that pivot, the company formally describes itself as a Strategic Asset Company (SAC), a designation that reflects its repositioning away from pure mining and toward Ethereum infrastructure and AI compute.
CEO Sam Tabar described the dual-pillar strategy this way: "Ethereum infrastructure and AI compute infrastructure are not separate strategies, they are components of a single integrated platform aligned with the future of the digital financial system."
Bit Digital holds approximately 70% of WhiteFiber (Nasdaq: WYFI), an AI and high-performance computing data centre subsidiary that IPO'd in August 2025, raising approximately $183 million in gross proceeds. WhiteFiber reported Q1 revenue up 31% year-on-year to $21.9 million on the same day. However, the subsidiary posted a net loss of $12.0 million, a reversal from net income of $1.4 million in the same quarter a year earlier, as it funds capacity expansion across facilities in Montreal and North Carolina.
Why this matters in South Asia and Africa
The ETH yield compression story is not confined to Nasdaq-listed companies. Any protocol treasury, DAO, or DeFi application in Africa or South Asia that relies on staked ETH as a yield source is operating under the same network constraints. With DeFi lending products such as DAI now offering 7.2% to 8.7% APY compared to ETH staking at roughly 2.6% to 3.1%, capital is already rotating toward alternative yield strategies. Even the Ethereum Foundation swapped 5,000 ETH for 11.11 million DAI in early April 2026, receiving roughly $2,222 per ETH, a rate slightly above the March 31 closing price of $2,104, rather than staking those funds.
Africa represents one of the most consequential regional contexts for this shift. South Africa alone accounts for an estimated $35 to $40 billion in annual crypto volume, and stablecoins make up more than 45% of all African crypto activity. Analysts at Ripple and Zawya have described 2026 as a potential tipping point for African crypto commerce. For developers building DeFi applications across the continent, the composability of liquid staking tokens is directly relevant: unlike native staked ETH, LSTs can be deployed inside lending protocols, yield aggregators, and payment rails, making the institutional shift away from raw staking a technical opportunity as much as a capital allocation question.
In South Asia, India illustrates the stakes most clearly. The country is the region's largest crypto market by user count, but a 30% capital gains tax and a 1% tax deducted at source on transactions have constrained retail participation and sharpened developer interest in optimising yield wherever possible. For those already navigating high tax friction, the gap between a 3% staking yield and an 8% DeFi lending rate carries real weight. In markets including Bangladesh, Sri Lanka, and Nepal, where remittance flows drive significant interest in stablecoins and DeFi infrastructure, the same pressure toward higher-yield alternatives is already visible.
In Pakistan, the government's allocation of 2,000 megawatts of surplus power capacity to Bitcoin mining and AI data centres provides a potential regional model for the infrastructure pivot Bit Digital has undertaken. The Pakistan Finance Ministry described the country, in a 2025 policy statement, as a digital bridge between Asia, Europe, and the Middle East. The economics driving Bit Digital away from mining and toward AI compute are precisely the economics that Pakistan's energy surplus is positioned to attract.
Looking ahead
Bit Digital's trajectory will depend on whether WhiteFiber can fill its data centre capacity pipeline quickly enough to offset compressed staking yields. The subsidiary is targeting 76 megawatts of gross capacity by late 2026, with a 1,500 MW long-term pipeline under review. For ETH holders across the board, the more immediate question is whether yields, currently running between roughly 2.84% and 3.3% with some estimates as low as 1.87%, will hold at these levels or compress further as validator growth continues. Liquid staking protocols and higher-yield DeFi alternatives are already absorbing the institutional response. In this writer's assessment, retail and developer-scale users in emerging markets are likely to follow a similar capital rotation path, though the pace will depend heavily on local regulatory conditions and the availability of compliant yield products.