hodler

28 Feb 2026, 11:05

MARA Revenue Dips 6% in Q4 as Production Slows and Asset Values Tumble

Marathon Holdings’ Q4 2025 revenue declined 6% to $202.3 million, primarily due to a 14% drop in the average price of bitcoin mined. The company also reported a $1.7 billion loss in the quarter. Annual Revenue Growth In its latest financial report, Marathon Holdings revealed a 6% decrease in revenues, totaling $202.3 million for the

28 Feb 2026, 10:30

Can BTC, ETH, and SOL Liquidity Work Together? LiquidChain (LIQUID) Crypto Presale Focuses on Staking and Settlement

Bitcoin, Ethereum, and Solana are three of the largest ecosystems in digital assets. Bitcoin anchors the market with deep liquidity and security. Ethereum supports most decentralized applications and DeFi protocols. Solana offers high-speed execution and low transaction costs for active trading environments. Individually, each network dominates its niche. Collectively, however, they operate in parallel. Liquidity remains segmented. Applications are often deployed separately across chains. Capital moves, but rarely without added steps, wrapped assets, or bridging mechanisms. This raises a structural question: can liquidity across BTC, ETH, and SOL operate within a coordinated system rather than remain siloed? LiquidChain (LIQUID) introduces its Layer 3 framework as a potential answer, with its crypto presale structured around staking incentives and cross-chain settlement infrastructure. How LiquidChain Coordinates Liquidity and Execution LiquidChain is a Layer 3 settlement environment that sits above major blockchains. However, rather than competing directly with Bitcoin, Ethereum, or Solana, it attempts to connect them through unified liquidity pools and synchronized execution. At the center of the model are shared liquidity structures. Instead of maintaining separate reserves across multiple ecosystems, assets from BTC, ETH, and SOL environments can be represented within a coordinated framework. The objective is to reduce duplicated liquidity and improve capital efficiency across decentralized markets. Execution is handled through a high-performance virtual machine built for multi-chain operations. This is designed to process interactions involving multiple ecosystems in real time. By coordinating execution within a single layer, the protocol aims to streamline settlement processes that would otherwise require traditional bridging. Security considerations are addressed through cross-chain proofs and messaging mechanisms. Bitcoin UTXOs, Ethereum account states, and Solana program states can be verified through cryptographic validation systems integrated into the Layer 3 design. The goal is to minimize additional trust assumptions while maintaining compatibility with the underlying chains. The framework positions LiquidChain as a settlement coordinator rather than a replacement network. Bitcoin continues serving as a store-of-value backbone. Ethereum retains its smart contract depth. Solana maintains throughput advantages. LiquidChain attempts to aggregate liquidity and align execution across them. $LIQUID Tokenomics, Staking, and Crypto Presale Structure The $LIQUID token underpins participation in this coordinated system. Its ongoing crypto presale marks the initial distribution phase ahead of full network deployment. Over $560,000 has been raised already. The total supply is set at 11,800,000,100 $LIQUID. Allocation includes 35% dedicated to development, supporting continued improvements to the Layer 3 infrastructure. LiquidLabs receives 32.5%, focused on ecosystem expansion and strategic initiatives. AquaVault accounts for 15% allocated toward business development and community activation. Rewards represent 10% of the supply, designated for staking incentives and ecosystem participation programs. Growth and listings account for 7.5%, intended to support exchange expansion efforts. Staking forms a central component of the token’s early utility. Participants can lock $LIQUID to receive reward emissions distributed proportionally across the staking pool. As more tokens are staked, rewards are shared among a larger base, which gradually reduces annual percentage yields over time. This reward structure is designed to encourage early buyers without fixing unsustainable returns. Early participants receive a larger proportional share of emissions when the staking pool is smaller. As adoption increases and more tokens enter staking, yields normalize based on total participation. The crypto presale therefore represents more than token distribution. It serves as a mechanism to bootstrap liquidity alignment, incentivize early adoption, and fund continued protocol development. A Framework for Cross-Chain Coordination Bitcoin, Ethereum, and Solana each command big capital and developer ecosystems. Yet fragmentation remains one of decentralized finance’s most persistent structural constraints. LiquidChain’s thesis centers on coordination rather than competition. By introducing a Layer 3 settlement environment supported by unified liquidity pools and dynamic staking incentives, the protocol seeks to create a shared execution framework across major chains. Success will ultimately depend on technical implementation, developer integration, and broader ecosystem participation. Infrastructure projects require sustained adoption to validate their models. Still, the core premise addresses a visible inefficiency: siloed liquidity across dominant ecosystems. Through its crypto presale, staking model, and layered settlement design, LiquidChain positions itself around the idea that cross-chain capital coordination may become a defining theme in the next phase of decentralized finance. Explore LiquidChain and its ongoing crypto presale: Presale: https://liquidchain.com/ Social: https://x.com/getliquidchain Whitepaper: https://liquidchain.com/whitepaper The post Can BTC, ETH, and SOL Liquidity Work Together? LiquidChain (LIQUID) Crypto Presale Focuses on Staking and Settlement appeared first on Cryptonews .

27 Feb 2026, 23:00

XRP Daily Liquidity Is Pointing To A Rally To $4, Analyst Explains What’s Going On

XRP’s liquidity structure on higher timeframes is in a situation where the path of least resistance could extend to the $4 level. The remark came from crypto analyst Bird in response to hourly and daily liquidity heatmaps shared by Cryptoinsightuk, which show a clear contrast between short-term and higher-timeframe liquidity positioning. At the time of writing, XRP is trading around $1.45, still below the large liquidity clusters visible above the current price. According to Bird, that imbalance may not stay unresolved for long. Hourly Liquidity Cleared, Short-Term Volatility Reduced XRP’s liquidity heatmap on the hourly candlestick chart shows that much of the nearby liquidity below the current price has already been swept. The visible clusters around the $1.30-$1.50 range have all been cleared, meaning that the short-term stop hunts and liquidation pools have largely been cleared out. Related Reading: Analyst Predicts Bitcoin Price Surge To $500,000 As Ribbon Fractal Emerges According to Bird, this trend shows that hourly XRP liquidity is basically gone. This means there is less immediate incentive for XRP to stay around current levels on lower timeframes. When short-term liquidity dries up like this, the outlook is that the price will gravitate to areas where larger pools are untouched. Since the nearby liquidity has already been taken, the next logical target is now where there are larger concentrations of resting orders. As noted by the analyst, these resting orders are stacked all the way up past $4. XRP Hourly Liquidity. Source: @Cryptoinsightuk on X Daily Liquidity Stacked Above $4 Liquidity on the daily heatmap appears layered and dense above the current price, stretching through multiple resistance bands and extending above the $4 price level. The upper regions show heavy trading activity and visible liquidity clusters between $2.50 and $4.00, which is a reflection of a thick concentration of stop orders and resting interest. Related Reading: Bitcoin Final Sell-Off Coming? Analyst Says It’s Time To ‘Buckle Up’ In liquidity-based trading theory, price action is often drawn to areas where there are large position orders, especially when those zones have not yet been tapped. Bird described this higher-timeframe liquidity as stacked all the way up past $4, with the notion that the higher-timeframe liquidity is sitting there like a magnet. XRP Daily Liquidity. Source: @Cryptoinsightuk on X Bird also referenced a five-month breakdown in Bitcoin dominance. At the time of writing, the Bitcoin dominance is at 57.9%, down from 58.2% last week. This means Bitcoin has been steadily losing dominance. A decline in dominance is always due to capital rotation into altcoins. If that trend continues, XRP could easily become one of the best beneficiaries, particularly given its visible higher-timeframe liquidity targets. The analyst also noted that sentiment has not yet reached extreme lows. XRP, in particular, has maintained a relatively positive positioning among investors compared to other cryptocurrencies like Bitcoin and Ethereum. That combination of declining dominance and neutral-to-cautious sentiment can create conditions for XRP’s projected rally above $4. Featured image created with Dall.E, chart from Tradingview.com

27 Feb 2026, 18:00

BitFuFu: Undervalued After A Full Operational Reset

Summary BitFuFu has staged an operational and financial recovery since Q1 2025, with revenue doubling YoY and cloud mining demand accelerating. FUFU's fleet efficiency improved to ~17.5 J/Th, and unencumbered BTC holdings now cover over 20% of its $436 million market cap. I estimate Q4 revenue at $120–135 million, likely beating consensus, despite lower hashrate and Bitcoin prices; full-year revenue should reach $555–570 million. I reiterate a Buy rating on FUFU, citing undervaluation, institutional accumulation, and a forward P/S below 1.0x, but highlight BITMAIN dependency and hashrate volatility as key risks. I initiated coverage of BitFuFu ( FUFU ) in June 2024, when the company was relatively unknown. ButFuFu was a newly listed Bitcoin ( BTC-USD ) miner fresh off a $1.5 billion SPAC merger, but revenue was growing, and the stock was trading at a discount to all the publicly traded mining peers on an EV/sales basis. When I last covered BitFuFu in June last year on the release of the Q1 2025 results, FUFU was reeling from an unimpressive Q1 performance that saw revenue plunge 46% and a net loss of $16.9 million. Management blamed much of the financial decline on the Bitcoin halving that had happened a year before, and in that follow-up piece, I argued that management’s halving excuses were largely a deflection. The real issue, in my view, was a 28% collapse in hashrate due to expired procurement contracts and disruptions from supplier fleet relocations. Cloud mining registered users had nearly doubled YoY to over 607k users by Q1 2025. Demand was accelerating, but the infrastructure to serve the accelerated demand had temporarily lagged. I maintained that if BitFuFu could restock its shelves with newer-gen miners, a turnaround was imminent. And I maintained a Buy rating on that basis. BitFuFu's Operational Recovery Since Q1 2025: By The Numbers Some of the improvements that have happened on the ground since Q1 2025 include restoration of hashrate, which went up from 20.6 EH/s as of the end of Q1 to 36.2 EH/s by the end of Q2 2025. The mining hashrate as of the last monthly operational report for January 2026 has dropped to 29.6 EH/s, which I'll address in the risk section. But on a YoY basis, the hashrate remains up 46.5%. In addition to the hashrate recovery, BitFuFu's fleet efficiency has also improved. This is an underappreciated element of the recovery story. When I reported on BitFuFu's in May last year, average fleet efficiency stood at 19.1 J/Th. As of the latest monthly update (January 2026), that figure has improved to ~17.5 J/Th, driven by the deployment of next-generation Antminer S21 units. For investors who might not be familiar with mining terms and how the J/Th numbers are interpreted, lower joules per terahash (J/Th) means more Bitcoin produced per unit of electricity consumed and is a direct improvement in cost structure per BTC mined. In addition to the operational recovery, BitFuFu's financial trajectory post-Q1 has also been strong. As of Q1 last year, revenue was $78 million, and the company recorded a net loss of $16.9 million. The last 10-Q report , which was for Q3 2025, shows revenue improved to $180.7 million with $11.6 million in net income. Q3's $11.6 million net income includes a $3.1 million unrealized BTC gain. I'm flagging this not to diminish the bottom line results, but to show that the thesis does not need inflated numbers to hold. The core operational story (revenue doubling, cloud mining demand accelerating, hashrate recovering) stands on its own. On a YoY basis, revenue doubled. Adjusted EBITDA came in at $22.1 million. Q3 EPS beat the Street consensus of $0.03 by 100%, at $0.06. Cloud mining remains the dominant revenue segment for BitFuFu. In Q3 2025, cloud mining revenue reached $123 million, which was a 78.4% increase YoY, and it accounted for 68% of total revenue. Registered cloud users grew to over 641k, a 40.8% increase YoY. The metric I told readers to watch when I covered FUFU in Q1, Net Dollar Retention [NDR], came in at 118.8% in Q3. That means existing cloud mining customers have been spending more over time. BitFuFu held 1,796 BTC on its balance sheet as of the latest monthly update, though 252 BTC are pledged as collateral for loans and miner procurement payables, leaving ~1,544 BTC unencumbered. Which would be valued around $105 million at Bitcoin's current spot price of $69,000. This is a meaningful holding for a company with a market cap around $436 million. That holding already covers over 20% of FUFU’s current valuation. BitFuFu Q4 Expectations Q4 results will be released in March, and there is a pattern I've noticed while the stock has been down. Institutions that already hold FUFU have been adding to their positions. Fintel Mirae Asset Global ETFs tripled its stake from around 160k FUFU shares in mid-2025 to over 543,838 shares presently, as disclosed in their 13F-HR form filed just about a week ago. Another investment firm, Harvest Portfolios, increased its position by over 150% in Q3; Vontobel Holding initiated a new stake in Q3, and Invesco and Vident Advisory both initiated positions in Q2 and have recently increased their stake. As of today, FUFU has 20 institutional holders, who are long only. Though FUFU’s institutional holding is still a small amount (institutional buyers hold ~1.18 million shares against a ~17.8 million tradable float), it is still a good signal that the institutional holders are long only. Going into the Q4 earnings, the monthly operational updates in October , November , and December give enough to build a credible Q4 revenue and operational metrics estimate ahead of the earnings, so let me work through it. Hashrate averaged around 27.7 EH/s across Q4 (30.5 EH/s in October, 26.4 EH/s in November, and 26.1 EH/s in December). That is a 23% decline from Q3's average, which was 36.7 EH/s. BTC production across the three months came in at 672 BTC in total (253 in October, 231 in November, and 188 in December). That is around 43% lower than the BTC produced in Q3. The sequential decline isn't just limited to hashrate decline. Bitcoin's spot price fell sharply through the quarter too, from above $100K in early October to closing near $90K in December, with a dip around $84K in November. The average realized BTC price across Q4 was ~$80,000, compared to ~$97,000 in Q3. The implication of those two variables (lower hashrate and lower average BTC price) will have a meaningful impact on Q4 top line numbers. BitFuFu self-mined 30, 41, and 37 BTC, respectively, in October, November, and December. At BTC's average realized price of ~$80,000, self-mining contributed around $8.6M in revenue in Q4. That is down sharply from Q3, where self-mining revenue was $20.1 million. The cloud mining segment, which I expect to hold better given its contract structure and the 648,221 registered users as of November’s report, will likely soften the negative impact of the BTC price drop but may not offset it entirely. Cloud mining contract prices are correlated to Bitcoin spot price movement, but they don't move 1:1 with BTC price. Contracts priced earlier in the cycle provide some buffer. Self-mining revenue, on the other hand, is directly exposed. Thus, my Q4 revenue estimate lands in the $120 million to 135 million range. Which is well below Q3's record $180.7 million, and it will look bad in the headline comparison. But ordinarily I don't think it should dampen sentiment much. For context, $120 million to $135 million in a quarter where Bitcoin spent most of its time between $84K and $95K, when BitFuFu's hashrate was in active transition from legacy S19 machines to next-gen S21 units, and where management was simultaneously deleveraging the balance sheet by slashing pledged BTC from 620 to 274, is a solid operational quarter, in my view. BitFuFu Q4 consensus estimate (Seeking Alpha) Consensus estimates revenue for Q4 to be around $105 million, but based on the calculations I've outlined above, I believe FUFU will beat consensus on the top line. In November, management sold 205 BTC at an average price of $107,000, gaining $21.9 million in proceeds realized at close to the quarterly high for Bitcoin. That capital funded the pledged BTC reduction, which has strengthened the unencumbered treasury heading into 2026. It is the kind of financial management that does not show up in the top line but matters a lot for balance sheet quality and future operational flexibility. The unencumbered BTC moved from roughly 1,342 BTC at the end of Q3 to ~1,506 BTC at the end of Q4. On profitability, the Q4 net income picture will depend heavily on whether management recognized any unrealized BTC fair value gains, a variable that has swung results in both Q2 and Q3. My base case is that Q4 operating income would be thin but positive, supported by the fleet efficiency improvement to ~18.1 J/TH as the S21 transition progressed and the continued cloud mining user base. On the other hand, a net loss in Q4 should not be ruled out given the Bitcoin price environment, but I'd expect it to be modest, not a repeat of Q1's $16.9 million loss, which was driven by the diminished capacity problem, which no longer exists. With full-year BTC production of 3,662 BTC confirmed, the full-year FY25 top line should show revenue in the $555 million to $570 million range, which will be ~20% growth over FY24's $463.3 million. Data by YCharts For a company trading at ~$2.60 with a $436 million market cap, that full-year revenue figure produces a forward P/S ratio below 1.0x, which is not a multiple that reflects a company with 648k+ cloud mining users, a 17.5 J/Th efficiency, and 1,500+ unencumbered BTC on its balance sheet. This is one of the lowest sales multiples among publicly traded Bitcoin miners. Valuation Rerating Case Sell side analysts have set price targets between $6 and $7 for FUFU. B. Riley has maintained a $6 target, and H.C. Wainwright maintained a $7.00 target. BitFuFu hasn't been an aggressive diluter, so it is safe to say at current share count a ~$7 FUFU share price would imply a market cap around $1.17 billion. Which is around 2.7x of where FUFU trades today. On a P/S basis, $7 implies around 2x P/S against full-year FY25 revenue in the $555 million to $570 million range. For a miner with a cloud segment posting 118.8% NDR with 648k+ registered users and a fleet running at ~17.5 J/TH efficiency, I believe a 2x P/S is not a stretch at all. Risks I've covered the BITMAIN dependency in every article I've written on BitFuFu, and I'll state it one more time. The relationship is BitFuFu's single greatest competitive advantage but also its most material structural risk at the same time. BitFuFu signed a two-year framework agreement with BITMAIN spanning 2024 to 2026, granting exclusive rights to purchase up to 80,000 next-generation S-series miners; that is the hardware access that makes the fleet efficiency story possible. But the flip side is equally real. BITMAIN controls the hosting infrastructure, the hardware supply chain, and the service arrangements that BitFuFu's cloud customers depend on. Any deterioration in that relationship in terms of pricing or capacity allocation flows directly into BitFuFu's financials in a way that the company cannot easily offset. Also, the January 2026 hashrate of 29.6 EH/s, down from the Q2 peak of 36.2 EH/s, is a sequential decline worth watching closely. Of that 29.6 EH/s, 25.9 EH/s comes from third-party suppliers and only 3.7 EH/s from BitFuFu’s self-owned facilities, which means capacity can evaporate quickly when supplier contracts expire or hosting arrangements shift, which was exactly what triggered the Q1 2025 collapse. The hashrate remains up 46% YoY, and January's rebound from December's 26.1 EH/s low is encouraging, but management will need to provide a credible explanation on why the hashrate has stepped down materially from its mid-2025 peak and a clear path back toward the 33 EH/s full-year guidance they already surpassed last May to ease investor concerns about a repeat of 2025's early stumble going into 2026. The risk that cuts across everything else is where Bitcoin sits right now. We are at nearly 50% below the October all-time high price of $126,000 for Bitcoin. We are roughly now around 140 days into a bear market, and with all of BitFuFu's cloud mining contracts being Bitcoin denominated, prolonged price weakness could dampen sentiment among cloud miners if the mining economics become unfavorable. Takeaway BitFuFu is a fundamentally sound business trading at a price that currently reflects none of it, as the valuation multiples show. Therefore, I'm reiterating a Buy on FUFU.

27 Feb 2026, 17:02

Critics target Dorsey's 'managerial incompetence' as Block lays off staff in 'intelligence' pivot

Block has let go of 40% of its workforce, leaving just under 6,000 employees. Jack Dorsey, the company’s cofounder and CEO, has cited the use of intelligence tools in the company’s operations as a reason for the massive layoffs. Despite Jack Dorsey stating that the layoffs are to streamline operations, he has been accused of using AI as a cover for the cleanup of his hiring spree from 2019 through 2022. Is AI taking jobs, or did Jack Dorsey make a mistake? Jack Dorsey’s Block announced that it will cut over 4,000 employees, bringing its total staff to just under 6,000. Between 2019 and 2022, the company had a hiring spree that saw its employee headcount more than triple from 3,900 to 12,500. Despite the significant blow to the workforce, investors celebrated the news by giving Block a 25% bump in stock price. Dorsey, the company’s cofounder and CEO, addressed staff in a memo, bluntly stating that the company had become too big, too slow, and too fragmented. Rather than spread the reductions over years and destroy morale, he chose a single, large-scale cut. In Dorsey’s memo , he stated that Block was efficiently carrying out its operations with intelligence tools and “smaller and flatter teams” as the reason behind the layoffs, but he also admitted that during the pandemic, he incorrectly built two separate company structures for Square and Cash App and created a duplication of roles that resulted in massive complexity. A debate about whether or not artificial intelligence is to blame for the massive layoffs or if it’s just a convenient excuse for mistakes made by the company’s management has started online. By Dorsey’s admission, from 2019 to 2024, Block’s efficiency remained flat at approximately $500,000 in gross profit per person. With the new, leaner structure, Dorsey is now targeting over $2 million in gross profit per person, representing a 4x increase in efficiency, with internal intelligence tools like Goose. Dorsey went on to say that he believes most companies are late to realize what he has, but he predicts that the majority of tech firms will make similar structural changes within the next year. For those leaving Block, the company is offering 20 weeks of salary plus one extra week for every year of service. Employees will also receive equity vesting through the end of May, six months of health care, and $5,000 to assist with the transition. Dorsey stated he wanted the process to feel “awkward and human” rather than “efficient and cold.” He’s keeping communication channels like Slack open until Thursday evening so colleagues can say goodbye. How is Block’s financial health? Cryptopolitan recently reported that Block’s stock soared because it beat gross profit guidance and reported a high operating income of $485 million. Cash App has monthly transacting active users reaching 59 million. Primary banking stats grew 22% year-over-year to 9.3 million. Square’s gross payment volume (GPV) in international markets grew 24% year-over-year. It successfully scaled its lending products, such as Cash App Borrow, which helped drive a 33% surge in Cash App’s gross profit. Block recently shipped its first Proto Bitcoin mining units, which are modular ASIC miners designed to be more energy-efficient and longer-lasting than traditional hardware. Block has raised its guidance for the full year of 2026 to $12.20 billion in gross profit, up 18% from the previous year. Its adjusted operating income is projected to hit $3.20 billion, representing a 26% margin. Block also continues to return value to shareholders through buybacks, having purchased 11.9 million shares for $790 million in the recent quarter. Amazon, like Block, recently cut 16,000 corporate roles to integrate AI agents, and Meta eliminated over 1,000 positions in its AI division to slim down operations. So far in 2026, tech companies have announced over 30,000 job cuts worldwide. Get seen where it counts. Advertise in Cryptopolitan Research and reach crypto’s sharpest investors and builders.

27 Feb 2026, 15:39

Hut 8: The Energy Story Is Real, But The Earnings Aren't There Yet

Summary Hut 8 Corp. trades at a premium, pricing in digital infrastructure earnings not yet realized in financials. Compute revenue more than doubled in 2025, but the power segment—central to the infrastructure thesis—contracted, raising questions about diversification progress. Signed long-term contracts with Google and Anthropic de-risk the pipeline, yet material income from these deals won't appear until 2027. I maintain a Hold rating: near-term earnings remain volatile and Bitcoin-dependent, while premium valuation demands flawless execution and Bitcoin resilience. Introduction Hut 8 Corp. ( HUT ) is being valued as if its transformation into a digital infrastructure platform is already complete. It isn't. The company has made genuine progress. It has signed contracts, developed a credible energy pipeline, and achieved compute revenue that more than doubled in fiscal 2025. The stock, however, trades at roughly three times the forward EV/EBITDA multiple of its large-cap mining peers, implying that structural diversification has already shown up in earnings. The Revenue Picture Hut 8 posted 45% year-over-year revenue growth in fiscal 2025 , with total revenue reaching $235.1 million. That puts it in the middle of the large-cap mining peer group . According to Seeking Alpha figures, Riot Platforms, Inc. ( RIOT ) posted 103.6% over the same period. MARA Holdings, Inc. ( MARA ) came in at 53.5%. That would be unremarkable if forward profitability was above peers, but Hut 8 is expected to post approximately -14% forward EBITDA growth. Riot stands at 11.7%, and Marathon is at 15.6%. Mid-tier revenue growth and negative forward EBITDA growth at a premium multiple are the main issues here. Valued For Future Income The valuation gap between Hut 8 and its peers is substantial. The stock trades at 24.7x trailing sales versus 8.9x for Riot and 3.2x for Marathon, and at 31.5x forward EV/EBITDA, roughly three times both peers. Even on a forward EV/sales basis, Hut 8 sits at 16.5x compared with 10.5x for Riot and 6.3x for Marathon. Those multiples price in infrastructure earnings that have yet to contribute significantly to the income statement. With negative forward EBITDA growth, the burden of proof rests on execution rather than current financial performance. What The Income Statement Actually Shows The headline numbers for fiscal 2025 aren’t great, with a net loss of $248 million and an Adjusted EBITDA loss of $135.4 million, compared to net income of $331.4 million and Adjusted EBITDA of $555.7 million in 2024. However, the swing was driven largely by a $220 million unrealized mark-to-market loss on Bitcoin holdings, reversing a $509.3 million gain the prior year. Strip out the mark-to-market change, and the picture is considerably less bleak. Gross margin expanded from 47% to 54%. Q4 2025 revenues grew 179% year-over-year with gross margin expanding from 36% to 60%. The underlying operational trajectory is improving. The reported bottom line simply doesn't reflect it clearly. That is precisely the problem for anyone trying to support the valuation. As long as Hut 8 holds a significant Bitcoin treasury and mining remains the dominant earnings driver, reported EBITDA will continue to move with the Bitcoin price rather than with operational progress. Computing Up, Power Down The segment mix in 2025 is showing signs of transition. Compute revenue more than doubled to $202.3 million from $80.7 million, driven by ASIC colocation and the Highrise AI GPU cloud business. Compute now accounts for most of the total revenue. Power revenue, meanwhile, declined to $23.2 million from $56.6 million, reflecting the termination of the Ionic Digital managed services agreement. The compute growth is solid. But the power segment, which is central to the long-term thesis of owning and monetizing energy infrastructure, actually contracted in fiscal 2025. Management attributes this to the Ionic termination rather than structural weakness, and that explanation is plausible. But investors need to see power revenue grow, not shrink, for the infrastructure thesis to gain credibility. A company positioning itself as an energy infrastructure platform should, at some point, show growing energy revenue. The Contracts Are Signed To be fair to the bull case, recent news on the infrastructure side is genuinely encouraging. In Q4 2025, Hut 8 signed a 15-year lease for 245 megawatts of AI data center capacity at its River Bend campus, valued at $7 billion, with payments financially backstopped by Google. The company also signed a deal to build an AI data center in Louisiana for Anthropic, with Fluidstack operating the high-performance clusters, a deal management expects to generate roughly $454 million in net operating income annually over the base term. However, income from these deals is not expected to materialize until 2027. These are not vague pipeline announcements. They are signed agreements with creditworthy counterparties. The 8,500-megawatt development pipeline now has real anchors behind it. That said, signed contracts and recognized net operating income are two different things. The River Bend and Anthropic deals are meaningful de-risking events for the bull case, but they don't resolve the near-term earnings picture. What matters now is execution: whether the infrastructure earnings start flowing through the income statement on schedule and at the margins management has guided to. Until it does, the valuation is still running way ahead of the financials. The Sideways Bitcoin Problem If Bitcoin trades sideways from current levels for an extended period, Hut 8's negative forward EBITDA growth becomes even more difficult to reconcile with a 31.5x forward EV/EBITDA multiple. Post-halving, the operating leverage cuts both ways. Reduced block rewards mean margins are more sensitive to Bitcoin price movements than in prior cycles. The cost to mine one Bitcoin has effectively doubled in reward terms since April 2024. Current margins are holding up because the Bitcoin price has been elevated. Flat Bitcoin means the expansion that would justify the forward multiple simply doesn't arrive. Layer in the mark-to-market dynamics, and the picture gets worse. Flat Bitcoin means no unrealized gain tailwind on treasury holdings. A 20% Bitcoin decline would amplify the problem. Operating leverage works in reverse. EBITDA compresses faster than revenue. Treasury mark-to-market adds downside to already weak reported earnings. Risks And Catalysts The primary downside risk is Bitcoin price weakness combined with negative forward EBITDA growth at an elevated multiple. Multiple compression from current levels could be significant even without operational deterioration. Delays or cost overruns on the development pipeline would compound the problem. The most meaningful catalysts are concrete: River Bend and the Anthropic deal beginning to contribute real earnings, power segment recovery, and greater segment disclosure that isolates recurring infrastructure earnings from mining volatility. Cleaner reporting would itself support valuation stability. Right now, the two are too tangled together to give the market clarity. Conclusion Hut 8 has made genuine progress. The near-term challenge is that reported earnings remain volatile and Bitcoin-dependent, the power segment contracted in 2025, and the stock trades at a substantial premium across every key valuation metric. The infrastructure income is coming, but is not yet reflected at scale. Until then, the premium valuation requires investors to carry both Bitcoin cyclicality and development execution risk simultaneously, which may be asking a lot. I maintain a Hold rating.