Bit Digital CEO discusses pivot to AI and Ethereum

Bit Digital pivoted in 2025 from bitcoin mining to staking and AI/HPC, selling its bitcoin holdings and acquiring roughly $500 million of Ethereum while generating about a 3% staking yield. The company spun out its high-performance computing business into WhiteFiber (NYSE: WYFI), retaining a 70% stake; WhiteFiber, which IPO’d in August, recently secured an $865 million contract with Nscale and has clients such as Cerebus. Management will not sell its WhiteFiber shares when the lock-up expires (Feb. 6) and plans to pursue M&A or operational initiatives in 2026 to generate EBITDA so markets can value Bit Digital on earnings multiples rather than NAV.

Analysis

Market structure: Bit Digital’s pivot crystallizes winners — WYFI (AI/HPC retrofit providers), hyperscalers contracting capacity, and EOS/ETH staking operators — while legacy bitcoin miners and high-opex greenfield data-center builders are losers. WYFI’s retrofit play (6-month build vs 24-month greenfield) gives it near-term pricing power on deployment and captive demand from clients like Nscale ($865m contract), which can re-rate WYFI within 6–18 months if execution stays on plan. The move also shifts energy demand composition (from ASIC mining spurts to steady datacenter load), leaving commodity/energy names exposed to different seasonality and capex profiles.

Risk assessment: Principal tail risks are (1) regulatory action against ETH staking or new staking taxes within 3–12 months that could cut the ~3% yield; (2) execution/contract risk at WYFI (client delay or default) that could reduce projected revenue by >30%; and (3) a >30% ETH drawdown that revalues BTBT’s NAV materially. Hidden dependency: BTBT’s market value is now highly correlated to WYFI shares (70% stake) plus ETH price; the Feb 6 lock-up expiry is a behavioral catalyst despite management’s stated intent not to sell.

Trade implications: Direct long in WYFI is a preferred exposure to AI infra — consider 3–4% portfolio position entered in two tranches, take profits if shares rise >35% or cut at -20%. BTBT is a levered NAV/option on ETH+WYFI; size at 1–2% only and hedge with 3-month 20% OTM puts or buy BTBT Jan 2027 call spreads to limit downside. Pair trade: go long WYFI / short MARA (or RIOT) equal dollar to express rotation from miners to AI infra; target spread capture 15–25% over 3–9 months.

Contrarian angles: Consensus underweights governance and liquidity friction — keeping the 70% WYFI stake illiquid can create a persistent NAV discount that caps BTBT’s rerating even if WYFI performs. Market may be underpricing WYFI’s backlog (Nscale + Cerebus) but overpricing BTBT’s ability to self-fund M&A quickly; historical parallels (miners that pivoted to services in 2018–2021) show execution matters — a single large contract miss can wipe 30–50% of upside. Monitor lock-up flows, WYFI quarterly bookings, and ETH staking rule changes as primary inflection points.

Key Decisions for Investors

• Establish a 3–4% long position in WYFI (WhiteFiber) in two tranches: 50% now and 50% after next WYFI quarter if revenue growth >15% QoQ; place a stop-loss at -20% from entry and take-profit at +35% within 6–12 months.
• Take a tactical 1–2% long in BTBT sized as a levered NAV play; hedge with 3-month BTBT 20% OTM puts (buy protective puts) or buy a Jan 2027 call spread to cap downside; exit if ETH falls >25% or management announces sale of WYFI stake.
• Implement a pair trade: long WYFI / short MARA (or RIOT) equal-dollar to play rotation from bitcoin mining to AI infra; target 15–25% spread capture over 3–9 months and unwind if WYFI misses bookings by >10% or miners rally >20% on BTC strength.
• Use options to express conviction on ETH: buy ETH Jan 2027 30% OTM calls (smaller size) or a call-calendar to leverage potential re-rating from staking/regulatory clarity; alternatively buy a 6–12 month ETH put spread (10–30% OTM) sized to cap downside risk on BTBT exposure.