Bitcoin whale activity is picking up. Over the past week, wallet 15oD9P withdrew 1,635 BTC (about $110.7 million) from Bybit, Binance, and OKX. Separately, a newly created wallet, bc1q3a, withdrew 450 BTC (about $30.08 million) from FalconX around six hours ago.

Bitmine (@BitMNR), tied to Tom Lee (@fundstrat), has increased its Ethereum ($ETH) exposure, according to Lookonchain. The firm staked an additional 167,578 ETH, valued at about $340 million, lifting total staked holdings to more than 3.31 million ETH. The position is now worth roughly $6.72 billion.

Bitcoin treasury company Nakamoto Inc reported that it sold roughly 284 BTC during March for proceeds of about $20 million, according to its latest 10-K filing. The average sale price was approximately $70,422 per bitcoin. The company said the transactions reflect active treasury management. It also noted a sharp expansion of its bitcoin position last year, having acquired 5,342 BTC across 2025. Total cost basis was $631.39 million, with an average purchase price of about $118,171 per bitcoin.

The Office of the Comptroller of the Currency (OCC) has released a 376-page Notice of Proposed Rulemaking (NPRM) that lays out the first comprehensive federal implementation framework for the GENIUS Act. The proposal, published Feb. 25, 2026 (Federal Register Docket No. 91 FR 10202), includes 211 questions for comment. Public comments are due May 1, 2026, and the OCC is targeting July 18, 2026 for finalization (Source: OCC NPRM, 202602). At the center of the draft is a strict licensing and operating regime for permitted stablecoin issuers (PPSIs). Key provisions include: a continuously maintained 1:1 reserve requirement at fair value, a maximum redemption period of two business days, and an explicit prohibition on using reserve assets for collateralization, recollateralization, or any other form of reuse. The proposal offers no exceptions and no transition period for this ban. For large issuers with more than $25 billion in circulating supply, the OCC would also require deposit insurance coverage equal to 0.5% of reserve size, capped at $500 million, spread across multiple institutions (Source: Sullivan & Cromwell, 202603). If annualized redemptions hit a 10% trigger, the redemption window could automatically extend to seven calendar days, with notice to the OCC required within 24 hours (Source: Jones Day, 202603). Capital is another gating factor. The OCC estimates a practical minimum viable capital level for new applicants of $6.05 million to $25 million (Source: Covington & Burling, 202602). Bank subsidiaries and affiliates are positioned to benefit from a structural cost advantage because they can leverage existing compliance infrastructure rather than building it from scratch. A further complication is compliance uncertainty: the OCC draft explicitly defers AML/CFT requirements until the U.S. Department of the Treasury issues separate rules. As a result, licensed issuers could begin operating without a fully established AML framework, bearing the regulatory risk themselves. Market narratives around a "yield ban" may be oversimplified. While some coverage has framed the proposal as a blanket prohibition on stablecoin yields, policy analysis has pointed to the use of rebuttable presumptions, which may still leave room for platform-level rewards and merchant discount programs (Source: CoinDesk, 202603). The draft also embeds a resolution-style "waterfall" mechanism that could materially reshape issuer viability. If an issuer fails the capital test at the end of any quarter, it would be barred from issuing new stablecoins. Failing to meet standards for two consecutive quarters would trigger mandatory liquidation, and the issuer would be prohibited from charging redemption fees during that process. The mechanism is widely viewed as particularly challenging for crypto-native issuers with thinner capital buffers. A key open question is Tether's pathway to U.S. access. Under the draft, registering as an offshore PPSI (FPSI) would require a U.S. Treasury determination that the issuer's home-country regulator is "comparable," and it would require submission to U.S. federal court jurisdiction. No such determination has been made, and the political uncertainty remains high (Source: Sullivan & Cromwell, 202603). Timing matters. Once the OCC final rule takes effect, a 120-day compliance countdown would begin. If the final rule is issued before July 2026, the GENIUS Act could take effect as early as November 2026, accelerating the timetable for market restructuring. Risk note: The final rule could diverge meaningfully from the draft. One of the most consequential unresolved issues among the 211 requests for comment is the choice of liquidity standard design (Option A: principle-based vs. Option B: quantitative and mandatory), which could materially soften or tighten capital expectations. The analysis is also limited by the absence of finalized Treasury AML/CFT rules, and the assessment reflects information available as of March 2026. A major counter-signal would be a positive Treasury "comparable" determination for Tether's home jurisdiction, which could significantly alter competitive dynamics. This article is for informational and analytical purposes only and does not constitute investment advice.

Oil markets have been jolted since the Strait of Hormuz shut on March 2, disrupting roughly 17.8 million barrels per day of global flows. In March, Brent crude jumped nearly 60% and WTI climbed about 53%. For Brent futures, it was the biggest monthly gain since the contract began trading in 1988, surpassing the 46% surge seen during the 1990 Gulf War. A spike of this magnitude would typically lift inflation expectations and push government bond yields higher. Over most of the past 20 years, oil prices and the 10-year U.S. Treasury yield have generally moved together. This March broke the pattern. For the first three weeks, the relationship held: WTI advanced from $67 to $100 and the 10-year yield rose from 4.15% to 4.44%. The break came between March 27 and March 30. Oil kept climbing, while the 10-year yield fell from 4.44% to 3.92% in three sessions—a 52-basis-point drop—slipping below the closely watched 4% level. The move had the hallmarks of a classic flight to safety, with bonds indicating that growth risks were starting to dominate inflation risks. Oxford Economics summarized the shift: "Growth risks are beginning to outweigh inflation risks." In plain terms, investors are not dismissing inflation—they are increasingly worried about recession. Such a decoupling is rare, and past episodes have often ended badly. Over the last half-century, oil has surged more than 35% over short windows five times. The 1973 embargo was followed by a 4.7% drop in U.S. GDP. In 1979, the Iranian Revolution pushed global GDP about 3 percentage points below trend. The 1990 Gulf War coincided with a short U.S. recession. In 2008, crude hit $147; the financial crisis was the primary driver, but the oil shock added momentum to the downturn. The outlier was 2022: Russia’s war with Ukraine lifted oil prices without triggering a recession, but produced the strongest inflation in 40 years. The March 2026 move exceeded all prior cases. Federal Reserve economist James Hamilton has argued there is no automatic, one-to-one link between oil shocks and recessions, yet "the greater the net increase in oil prices, the more pronounced the suppression on consumption and investment." Reflecting that risk, Goldman Sachs lifted its estimate of a U.S. recession to 30%, while EY-Parthenon put it at 40%. Markets also repriced policy expectations at unusual speed. Early in March, the CME FedWatch tool showed expectations for three rate cuts this year and a 70% probability of a cut in June. As oil rose, sentiment swung sharply. On March 26, the U.S. import price index jumped 1.3%, and incoming Fed Chair Kevin Warsh suggested the neutral rate may be higher. That day, the implied probability of a rate hike within the year jumped to 52%, and the 10-year yield reached 4.35%. FinancialContent dubbed it "The Great Hawkish Pivot." Four days later, the story flipped. On March 30, consumer confidence plunged, manufacturing unexpectedly contracted, and the 10-year yield sank to 3.92%. FinancialContent reported that bets on a dovish Fed pivot in May rose to 65%. Goldman Sachs said the market had misread the direction of rate moves. On the same day, Jerome Powell told undergraduates at Harvard that the Fed "hasn't yet reached the point where it must decide whether to look through the shocks from the war," while stressing that "anchored inflation expectations are critical." Axios said markets took his remarks to mean the Fed was neither eager to hike to fight inflation nor rushing to cut to support growth, waiting instead to see whether the supply shock proves temporary or persistent. Bond investors, by contrast, moved decisively. If history is a guide, Citigroup strategist McCormick framed the risk bluntly: stagflation ahead—bad for bonds, bad for stocks. The 1973–1982 stagflation era offers a clear scorecard in real terms: gold delivered +9.2% annualized, the S&P GSCI commodity index gained 586% over the decade, and real estate returned +4.5%. The S&P 500 produced a 2% real annualized return, and long-term Treasuries returned 3%. NYU Stern historical data show long-term Treasuries lost 8.6% in a single year in 1979. The traditional 60/40 portfolio was hit hard, while real assets were the only consistent inflation hedges. On the oil outlook, Société Générale forecasts Brent averaging $125 in April, with a "credible peak" of $150. Goldman Sachs is more conservative at $115 for April, assuming the Strait of Hormuz returns to normal within six weeks and prices fall to $80 by year-end. The bond market has already made its call: between inflation and recession, it is betting on recession. Source: 律动

Teucrium's rollout of what it describes as the first XRP ETF in the U.S. has sparked a rapid wave of inflows. The fund attracted $500 million in its first 12 weeks, an eye-catching result for an ETF—especially for a crypto-linked derivative that does not hold physical XRP. "The XRP community is an army," said Sal Gilbertie, Teucrium's founder and CEO. "They're willing to go to battle. They really are." Spot XRP ETFs are also reshaping investor behaviour, Gilbertie said. In his view, long-term holders have largely shifted toward spot products, effectively parking their XRP and waiting, while the leveraged ETF has become the preferred vehicle for short-term trading. "The buy-and-holders have migrated towards spot XRP ETFs, rightfully so," he said. "The levered product users are very aggressive, mostly day traders." Teucrium has an inverse XRP ETF idea on the shelf, but Gilbertie said there is no urgency to bring it to market. "Right now everybody's bullish. We'll give investors what they want and need." On the fundamentals, Gilbertie said his optimism centers on Ripple's focus. He pointed to the firm's decade-plus push to make cross-border payments faster and cheaper, and contrasted the XRP Ledger's 3-to-5-second settlement times with the traditional finance T+1 framework. "We need instantaneous settlements. That's going to be needed in the modern economy. It is needed right now." He also highlighted Ripple's acquisition pace and licensing efforts as signs it is building a broader, integrated financial ecosystem beyond a single token. "I like their work ethic. I like the fact that they stay the course." Gilbertie is also tracking the CLARITY Act, legislation that could formally define crypto assets and reduce long-standing regulatory uncertainty. On the likelihood of passage, he struck a pragmatic tone: "Neither side being happy means there probably is a deal. That's the mark of a really good deal on a complicated subject."

The closure of the Strait of Hormuz on March 2 has disrupted roughly 17.8 million barrels a day of global oil flows, sending crude sharply higher. In March, Brent jumped nearly 60% and WTI gained about 53%. For Brent futures, it was the biggest monthly surge since the contract began trading in 1988, topping the 46% spike during the 1990 Gulf War. Normally, a shock of that size would lift inflation expectations and push bond yields higher. For much of the past two decades, oil and the 10-year U.S. Treasury yield have broadly moved together. This month, they broke apart. For the first three weeks of March, the old relationship held: WTI climbed from $67 to $100 and the 10-year yield rose from 4.15% to 4.44%. The reversal came between March 27 and 30. Oil kept surging, but the 10-year yield slid from 4.44% to 3.92%, a 52-basis-point drop in three sessions that took yields below the psychologically important 4% mark. The move looks like a classic flight to safety: the bond market is signaling that growth risks are overtaking inflation risks. Oxford Economics summed it up: "Growth risks are beginning to outweigh inflation risks." Investors are not dismissing inflation; they are more worried about recession. Such decouplings are rare, and the historical record is uncomfortable. Over the past half century, oil has surged more than 35% in the short term on five occasions. The 1973 embargo was followed by a 4.7% drop in U.S. GDP. In 1979, the Iranian Revolution pushed global GDP 3 percentage points below trend. The 1990 Gulf War preceded a brief U.S. recession. In 2008, oil peaked at $147; the financial crisis was the main cause of the downturn, but the oil shock helped deepen it. The exception was 2022, when the Russia-Ukraine war drove an oil spike that did not trigger a recession but produced the worst inflation in 40 years. The March 2026 run-up exceeded all of those episodes. Federal Reserve economist James Hamilton has argued there is no mechanical link between oil shocks and recessions, but "the larger the net increase in oil prices, the more significant the suppression on consumption and investment." Goldman Sachs has lifted its estimated probability of a U.S. recession to 30%, while EY-Parthenon puts it at 40%. The market's rate narrative has also flipped with unusual speed. In early March, the CME FedWatch tool showed expectations for three cuts this year, including a 70% chance of a June cut. As oil climbed, the tone shifted: on March 26, the U.S. import price index rose 1.3%, and incoming Fed Chair Kevin Warsh suggested the neutral rate could be higher. That day, the implied probability of a rate hike within the year jumped to 52% and the 10-year yield hit 4.35%. FinancialContent dubbed it "The Great Hawkish Pivot." Four days later the story turned again. On March 30, consumer confidence plunged, manufacturing unexpectedly contracted, and the 10-year yield fell to 3.92%. FinancialContent said expectations for a dovish pivot by the Fed in May rose to 65%. Goldman Sachs said markets had misread the direction of rate hikes. The same day, Jerome Powell told Harvard undergraduates the Fed "hasn't reached the point where it must decide whether to look through the impacts of the war," while stressing that "anchored inflation expectations are critical." Axios reported markets took the remarks to mean the Fed is reluctant to hike aggressively to fight inflation and not eager to cut to support growth, preferring to see whether the supply shock proves temporary or persistent. Bond investors appear less patient. Citi strategist McCormick framed the risk bluntly: stagflation ahead—bad for bonds, bad for stocks. The 1973–1982 stagflation era offers a stark scorecard: gold delivered a real annualized return of +9.2%, the S&P GSCI commodity index rose 586% over the decade, and real estate returned +4.5%. The S&P 500 posted a real annualized return of 2%, while long-term Treasuries returned 3%. NYU Stern historical data shows long-term Treasuries suffered a single-year loss of 8.6% in 1979. A traditional 60/40 portfolio (60% stocks, 40% bonds) is vulnerable in that backdrop, with real assets the main hedge against inflation. On the oil outlook, Société Générale sees Brent averaging $125 in April, with a "credible peak" of $150. Goldman Sachs is more restrained at $115 on average in April, assuming the Strait of Hormuz returns to normal within six weeks and prices slide to $80 by year-end. For now, the bond market has delivered its verdict: between inflation and recession, it is pricing the recession risk.

Crypto firm Bitmine has added 167,578 Ethereum (ETH) to its staking position, a deposit valued at about $342.4 million, according to on-chain analytics provider Onchain Lens in a post dated April 10, 2025. The transfer was made from a Bitmine-controlled wallet to Ethereum's official staking deposit contract. After the latest allocation, Bitmine's total staked balance stands at 3,310,221 ETH, worth roughly $6.7 billion. The scale of the position indicates sizeable liquid ETH reserves and a willingness to commit capital over the medium to long term in exchange for staking rewards, while also reducing the amount of ETH immediately available for trading. The move also expands Bitmine's validator footprint. Based on standard validator sizing, the new stake corresponds to about 5,250 validator nodes, bringing Bitmine's implied total to roughly 103,400 nodes. Onchain Lens estimates the firm's stake at around 2.8% of total staked ETH, using an estimated ecosystem total of approximately 118 million ETH. A larger validator set generally increases Ethereum's security by raising the economic cost of attacks and improving network resilience. As aggregate staking grows, staking yields can face gradual compression because rewards are distributed across a larger pool of participants. Bitmine's expansion lands amid a broader trend of institutional participation in staking, as asset managers and corporates seek yield while supporting the networks they hold. Market watchers also track large staking deposits as a signal of longer-horizon positioning, though analysts caution that single events are not reliable short-term price catalysts and that macro and regulatory factors remain key drivers. Disclaimer: This content is for informational purposes only and does not constitute investment advice. Conduct your own research and consult a qualified professional before making any investment decisions.

CoinShares has released its Q1 2026 Bitcoin Mining Report, warning that miner economics deteriorated sharply after a steep BTC pullback and continued hash rate growth pushed hash price to multi-year lows. Key points - Profit squeeze intensifies: Q4 2025 was the toughest quarter for miners since the April 2024 halving. Bitcoin fell from about $124,500 in early October to around $86,000 by end-December (about a 31% drawdown) as network hash rate stayed near record highs. Hash price dropped below $30/PH/s/day in early 2026, a five-year low, leaving roughly 15%–20% of older machines globally cash-flow negative. - AI/HPC pivot accelerates: Publicly listed miners have announced more than $70 billion in AI/HPC deals. Equity markets are paying for the AI story, with EV/NTM sales multiples around 12.3x for firms with HPC exposure versus 5.9x for pure-play miners. The sector is splitting into data-center infrastructure providers and mining-focused operators. - Short-term hash rate pullback: Global hash rate fell about 10% from its Q4 peak amid margin compression, winter curtailments, and regulatory inspections. CoinShares still expects a rebound, forecasting global hash rate at 1.8 ZH/s by end-2026 and 2 ZH/s by end-March 2027 (about one month later than its prior forecast). - Costs and leverage diverge: AI buildouts have inflated all-in cost metrics for hybrid miners as debt, SG&A, and depreciation are spread over a shrinking BTC production base. Some hybrids have taken on large debt loads (e.g., CIFR, WULF), while low-leverage miners such as CLSK and HIVE are showing tighter financial discipline and clearer advantages on pure mining costs. - Bottom line: The industry is undergoing structural reshaping. CoinShares argues that if BTC fails to reclaim $100,000 by 2026, high-cost operators are likely to capitulate faster, while ultra-low power cost miners and firms that successfully monetize AI infrastructure could dominate future capital flows. Q4 2025: the post-halving stress test CoinShares describes Q4 2025 as the most challenging quarter since the halving. With hash rate near all-time highs, hash price compressed to the lowest level in five years. The weighted-average cash cost to mine one BTC for publicly traded miners rose to about $79,995 in Q4 2025. The report highlights three themes: - Profitability under pressure: Hash price fell to about $36–38/PH/s/day in Q4, close to breakeven for many operators. Three consecutive difficulty reductions—the first such streak since July 2022—are interpreted as a sign of miner capitulation. Early in Q1 2026, hash price slid further to around $29/PH/s/day. - AI/HPC divergence: The gap is widening between pure miners and operators repositioning as AI data-center providers. WULF, CORZ, CIFR, and HUT are increasingly behaving like data-center companies that also mine. - Balance-sheet reset: Several miners have raised substantial debt to finance AI infrastructure. IREN holds $3.7 billion in convertible notes; WULF reports $5.7 billion in total debt; CIFR issued $1.7 billion in senior secured notes. AI competes with mining for rack space CoinShares notes that AI is increasingly displacing mining in data centers, potentially pushing Bitcoin mining toward more intermittent and cheaper power over time. Based on company disclosures, listed miners could generate as much as 70% of revenue from AI by year-end, versus about 30% today. The economics are driving the reallocation: hash price is near cyclical lows and fee revenue is exceptionally weak, while AI/HPC infrastructure offers higher and more stable returns. CoinShares also stresses the cost gap: mining infrastructure is roughly $700,000 to $1 million per MW, while AI infrastructure can run $8 million to $15 million per MW. Selected AI/HPC buildout updates cited in the report - CORZ (Core Scientific): ~350 MW energized, ~200 MW billed. CoreWeave contract expanded to $10.2 billion over 12 years, targeting full commissioning of 590 MW by early 2027. - WULF (TeraWulf): Lake Mariner has 39 MW of critical IT capacity online. Total contracted HPC revenue at $12.8 billion, with a five-site platform targeted at ~2.9 GW. - CIFR (Cipher Digital): Working with Fortress Credit Advisors on the 300 MW Barber Lake site; reached a multibillion-dollar Fluidstack agreement backed by Google; no revenue yet. - IREN: Expanded to more than 10,900 NVIDIA GPUs; Childress Horizon 1–4 GPU expansion (up to 200 MW) underway; AI cloud revenue was $17.3 million in Q4. - HUT (Hut 8): Signed a 15-year, $7 billion lease for 245 MW with Fluidstack at River Bend (Louisiana); first data hall targeted for early 2027. CoinShares points to the failed CORZ–CoreWeave merger vote (rejected Oct. 30, 2025) as a reminder of the gap between infrastructure value and equity value. CORZ later restated financials tied to accounting for assets slated for decommissioning during the HPC transition. Early revenue mix is shifting but mining still dominates - CORZ AI/HPC data centers: 39% of Q4 revenue - WULF HPC: 27% - IREN AI cloud: 9% - HIVE HPC: 5% Global hash rate: milestone, then a pullback The Bitcoin network exceeded 1 ZH/s for the first time in late August 2025 and peaked near 1,160 EH/s in early October. In Q4, global hash rate fell about 10% to roughly 1,045 EH/s by end-December, dipped to 850 EH/s in early February, then recovered. CoinShares attributes the pullback to: - Older S19-era rigs slipping below breakeven as BTC fell; for example, the S19 XP breakeven power price dropped from about $0.12/kWh (Dec. 2024) to about $0.077/kWh (Dec. 2025). - Higher winter power costs and ERCOT curtailments increasing unprofitable hours from November to December. - Renewed enforcement in Xinjiang in December 2025, which restricted mining activity but did not permanently relocate capacity. CoinShares estimates the network still added roughly 300 EH/s across 2025. At the time of writing, total hash rate was around 1,020 EH/s. The report argues the decline is far less severe than the 2021 China ban and is driven more by cyclical and weather-related factors. Geographic distribution The U.S., China, and Russia together control about 68% of global hash rate. The U.S. share rose about 2 percentage points quarter-over-quarter. Paraguay, Ethiopia, and Oman moved into the top ten, supported by projects such as HIVE's 300 MW build in Paraguay and BTDR's 40 MW in Ethiopia. Hash price: new lows into 2026 Hash price peaked near $63/PH/s/day in July 2025 and slid through Q4. By November it was around $35–37/PH/s/day, then briefly rebounded to $38–40 at year-end before falling again to about $28–30/PH/s/day by early March 2026. Drivers cited: - Record difficulty, peaking at 155.97T after a 6.31% increase on Oct. 29 - BTC down roughly 31% from October's high - Extremely low fee income, consistently below 1% of the total block reward, averaging about 0.018 BTC per block CoinShares says that at a $30/PH/day hash price, any rig weaker than an S19 XP paying $0.06/kWh or more is losing money, implying 15%–20% of the active fleet is at risk. BTC sales accelerate among public miners The report describes widespread miner selling. Public miners' aggregate BTC holdings are down by more than 15,000 BTC from their peak. Examples: - Core Scientific sold about 1,900 BTC (around $175 million) in January and plans to liquidate nearly all remaining holdings by Q1 2026. - Bitdeer zeroed out its treasury in February. - Riot sold 1,818 BTC (about $162 million) in December 2025. CoinShares' scenario framing - For hash price to sustainably exceed $40/PH/day, CoinShares says BTC would need to rally to $100,000 by year-end and outpace ongoing hash rate growth. - If BTC remains below $80,000 and difficulty continues rising, hash price could keep falling, though equipment shutdowns could stabilize the network. - A retest of the $126,000 historical high could lift hash price to about $59/PH/s/day. Mining cost analysis: Q4 2025 highlights CoinShares emphasizes that AI/HPC development is distorting headline cost-per-BTC metrics for hybrid operators because buildout-related debt, SG&A, and depreciation are allocated across fewer mined coins. Notable observations cited: - Depreciation and amortization is the largest non-cash cost and varies widely by accounting policy. MARA's $136,000/BTC and CIFR's $88,000/BTC are flagged as outliers. - Stock-based compensation remains a major differentiator. HUT's $48,500/BTC and CORZ's $35,500/BTC stand out. - Interest expense is increasingly material for several miners. WULF ($145,000/BTC), CIFR ($56,000/BTC), and BTDR ($16,000/BTC) are highlighted, while HIVE ($320/BTC) and CLSK ($830/BTC) are described as very low leverage. Company snapshots (selected figures as cited) - MARA: Q4 production 2,011 BTC; total cost $153,040/BTC; cash cost $103,605/BTC; energized hashrate 53.2 EH/s; disclosed policy shift in its March 2, 2026 10-K authorizing sales from its full 53,822 BTC reserve; LTV on its $350 million Bitcoin-backed facility rose to about 87% as BTC neared $68,000; acquired a 64% stake in Exaion for $174.5 million (Feb. 2026) and announced a Starwood Capital partnership in AI/HPC. - IREN: Q4 production 1,664 BTC; total cost $140,441/BTC; cash cost $58,462/BTC; power cost $34,325/BTC; $3.7 billion in convertibles (2029–2033) with coupons 2.75%–3.50%; AI cloud revenue $17.3 million (9% of total) and up to 200 MW liquid-cooled GPU build at Childress Horizon 1–4. - CLSK: produced 1,821 BTC; total cost $118,932/BTC; cash cost $71,188/BTC; interest expense $830/BTC; fleet efficiency about 16 W/T; exploring supplier diversification away from Bitmain; fiscal year ends Sept. 30 (figures are FY2026 Q1). - RIOT: produced 1,324 BTC; total cost $170,366/BTC; cash cost $102,538/BTC; Q4 ERCOT credits $9.9 million; held 17,722 BTC at Dec. 31; developing 1 GW Corsicana site with 600 MW allocated for AI. - CORZ: produced 421 BTC; total cost $168,693/BTC; cash cost $110,282/BTC; hosting revenue $31.3 million (39% of total); highlighted restatement of 2024–2025 financials, auditor change to KPMG, and determination of ineffective internal controls. - WULF: produced 262 BTC; average cost $471,841/BTC; cash cost $384,517/BTC; Q4 HPC leasing revenue $9.7 million (27% of Q4 revenue); total debt $5.7 billion; contracted capacity 522 MW under $12.8 billion in long-term agreements. - CIFR: produced 591 BTC; total cost $231,980/BTC; cash cost $103,516/BTC; issued $1.733 billion in 7.125% senior secured notes (Nov. 2025); renamed Cipher Digital Inc. on Feb. 20, 2026; Barber Lake 300 MW site planned with Fortress and Fluidstack agreement supported by Google; no HPC revenue yet. - HUT: produced 719 BTC; total cost $160,402/BTC; cash cost $50,332/BTC; SBC $48,527/BTC driven by one-time CEO/CSO awards; holds 15,679 BTC (about $1.37 billion); signed 15-year $7 billion, 245 MW deal with Fluidstack in Louisiana. - BTDR: produced 1,673 BTC; total cost $118,188/BTC; cash cost $87,144/BTC; interest expense $16,306/BTC tied to about $1 billion convertibles and related-party borrowings; highlights SEALMINER A2 at 16.5 W/TH and upcoming A3 at 9.7 W/TH. - HIVE: produced 884 BTC; total cost $144,321/BTC; cash cost $75,274/BTC; interest expense $320/BTC; debt $13.8 million; expanded in Paraguay with 100 MW Valenzuela online and 300 MW ANDE PPAs; notes contingent VAT liability of about $79.2 million tied to Swedish Tax Agency assessment under appeal. Valuation and positioning CoinShares says the AI/HPC premium widened in Q4: operators with HPC contracts trade at EV/NTM sales of 12.3x versus 5.9x for pure miners. The BTC price decline also reduced the value of miners' treasury holdings. Short interest remains elevated, with MARA around 30% of shares outstanding at the time of writing. Looking into 2026 CoinShares flags six themes: 1) Hash price recovery depends on BTC price; around $70,000 and $30/PH/day leaves many mid-generation fleets at breakeven. 2) Next-gen hardware: Bitmain S23 and SEALMINER A3 (below 10 J/TH) expected to scale in H1 2026. 3) AI/HPC revenue inflection: investors will watch whether contracted revenues convert into billing and whether margins exceed 85%. 4) Leverage divergence may drive M&A; healthier balance sheets (e.g., HIVE, CLSK) could become acquirers, though CLSK also raised $1.15 billion of 0% convertibles for AI infrastructure. 5) Regulatory and geographic shifts: U.S. share rising; Xinjiang enforcement may push migration; Texas SB 6 (signed June 2025) requires large ERCOT-connected mining/data-center loads to support remote shutdown. 6) Consolidation: efficiency gaps (about 15 W/T leaders versus above about 25 W/T laggards) could make buying efficient capacity cheaper than retrofitting older sites. Methodology note (as described by CoinShares) All-in cost per BTC allocates electricity (net of curtailment credits), SG&A (excluding SBC), D&A, net interest, income tax, and SBC to self-mining using the ratio of self-mining revenue to total revenue. Cash cost excludes D&A. Asset impairments and non-operating items such as BTC fair-value changes and derivative marks are excluded.