Track market-moving stories before they get noisy

Chinese smartphone maker ‌Xiaomi (1810.HK) on Monday unveiled a new version of its in-house Xring handset processor, betting that deeper control over key components will help strengthen its supply chain and reduce reliance on external chip suppliers.

The introduction of Xring O3 comes a year after Xiaomi launched its first proprietary smartphone processor, the Xring O1, ​marking the latest step in the world's third-largest smartphone vendor's push to join rivals such as Apple (AAPL.O), Samsung Electronics (005930.KS) and ​Huawei in developing its own chips.

TSMC (2330.TW) will manufacture the new chip using its 3-nanometre production technology, two ⁠people familiar with the matter said.

One of the sources said the chip is expected to power Xiaomi's upcoming flagship folding phone, with ​a shipment target of 200,000 to 300,000 units.

Xiaomi's expansion into foldable phones, a more expensive segment of the market, could challenge leading ​domestic player Huawei.

Huawei shipped 1.6 million foldable phones in China in the second quarter, giving it a 68% market share, followed by Honor with 13.7% and Oppo with 8.5%, according to research firm Smart Analytics Global.

The people declined to be identified because the plans are not public. Xiaomi and TSMC did ​not immediately respond to requests for comment on the chip's manufacturer, production technology or shipment targets.

Smartphone processors, or system-on-chips (SoC), integrate computing, graphics, ​AI processing and imaging functions into a single component.

DEVICE MAKERS PUSH FOR IN-HOUSE CHIPS
Xiaomi's chip push reflects a broader industry trend as device makers seek ‌to differentiate ⁠products and lessen dependence on suppliers such as Qualcomm (QCOM.O) and MediaTek (2454.TW) amid intensifying competition in premium smartphones.

Xiaomi said during an earnings call last week that cumulative shipments of devices powered by the Xring O1, including smartphones, tablets and watches, had surpassed 1 million units since its launch.

Xiaomi has sold about 150,000 smartphones based on the Xring O1 chip since its May 2025 launch, according to the sources.

MEMORY ​COSTS PUSHING UP PRICES
Smartphone makers are ​contending with a global downturn ⁠in device sales, as memory and component costs push up prices and squeeze demand.

Xiaomi sold 65 million handsets in the first half of 2026 at an average price of 1,329 yuan ($197.74), compared with ​84 million units sold at an average price of 1,141 yuan in the same period of ​2025 and 83 ⁠million units at 1,123 yuan in 2024, according to data from Visible Alpha by S&P Global.

Global smartphone shipments are expected to decline 14% in 2026, according to research firm International Data Corporation.

Xiaomi said on Monday it had also contracted TSMC to manufacture two other Xring chips: the ⁠Xring O100, ​a 6-nm neural processing unit that will support Xiaomi's large language model, MiMo, on ​consumer electronic devices, and the Xring D100, a 3-nm chip for autonomous driving.

According to Xiaomi, the O3 has already entered mass production, while the O100 and D100 ​have completed development and are slated for deployment next year.

Solana (SOL) edges lower to $94 on Monday, following a 27% rebound last week to a two-month high. SOL-focused Exchange Traded Funds (ETFs) recorded four consecutive days of inflows last week, totaling $28.34 million, suggesting renewed institutional buying. Validator voting on multiple proposals started on Monday, which includes doubling the disinflation rate to 30%. 

Solana validator voting beginsSolana validators have begun voting on SGP 1, SGP 2, and SGP 3 proposals, with voting ending Thursday. The SGP 1 protocol addresses the ratification of the Solana Constitution to govern Solana’s network-level decision-making. SGP 2 proposes reducing the inflation rate by increasing the disinflation rate to -15% to -30%. Finally, SGP 3 plans to introduce a fixed base inclusion fee paid to the block leader and a resource fee with a requested transaction cost, which will be burned. 

Taken together, the proposals will restructure on-chain decision-making and reduce pressure on available supply. 

Solana ETFs regain strengthSolana regains institutional demand. SoSoValue data shows the SOL-focused ETFs recorded $28.34 million in inflows last week, the highest over the last two months. Typically, renewed buying from institutional investors implies a bullish trend reversal. 

SOL ETFs data. Source: SosovalueTechnical outlook: Will SOL price cross above $100?Solana trades around $94 on Monday, holding a bullish near-term bias above both the 50-day Exponential Moving Average (EMA) at $79.04 and the 200-day EMA at $92.67. SOL price remains capped below the $100 psychological level and the May 11 high of $98.41.

For a sustained recovery, SOL should confirm a decisive close above the $100 mark. This could open the path toward the 127.2% Fibonacci extension level of the $98.41 to $60.13 downswing at $112.52.

Momentum stays strong, with the Relative Strength Index (RSI) hovering in overbought territory near 79 on the daily chart. At the same time, the Moving Average Convergence Divergence (MACD) shows a steady upward trend with a positive histogram, suggesting that buying pressure remains dominant.

SOL/USDT daily price chart.On the downside, initial support is seen at the 200-day EMA around $92.67, followed by the 78.6% retracement near $88.56. Deeper declines would expose the 50-day EMA at $79.04 and the 50% retracement at $76.92.

(The technical analysis of this story was written with the help of an AI tool. Know more.)

Nvidia (NVDA.O) is in talks ​to invest in Perplexity as part of an ‌equity funding round that would value the AI startup at more than $30 billion, The Information reported on Sunday, citing people with ​knowledge of the discussion.

The funding round would increase ​Perplexity's valuation by more than 50% from its ⁠previous financing a year ago, according to the ​report.

Perplexity's annualized revenue has risen to more than $750 million from ​less than $250 million at the start of the year, the report said. Part of the revenue growth has been driven by Perplexity ​Computer, a cloud-based AI agent used by professionals ​to automate computer-based tasks, the report added, citing people familiar with the ‌matter.

Perplexity ⁠declined to comment on the Information report, while Nvidia did not immediately respond to a request for comment.

The Information reported in September last year that Perplexity had ​finalized a $20 billion ​valuation.

Earlier this ⁠year, Perplexity signed a $750 million agreement with Microsoft (MSFT.O) to use its Azure cloud service, ​according to a Bloomberg News report.

Perplexity is planning ​to ⁠go public in 2028 regardless of how the market receives the listings of Anthropic and OpenAI, CEO Aravind Srinivas ⁠told CNBC ​in an interview in June.

The ​startup's high-profile backers, along with Nvidia, include Amazon founder Jeff Bezos and Japan's SoftBank ​Group (9984.T).

Phantom is pulling the plug on Sui. The multichain wallet announced on August 24 that it will stop supporting the Sui network on September 24, giving users exactly one month to move their assets before the integration goes dark.

The decision came through a joint agreement between Phantom and the Sui team, which makes it a mutual parting rather than a unilateral cut.

What users need to do before the deadline Nobody is losing their funds. Phantom has been clear that users retain full control over their private keys and assets throughout the transition, and the wallet will not be touching custody arrangements.

The practical options are straightforward. Users can transfer their Sui assets to another Sui-compatible wallet, with Suiet and the official Sui Wallet listed as alternatives. Or they can swap their Sui holdings for assets that Phantom continues to support, and Phantom is waiving fees on those swaps before the September 24 cutoff.

Phantom is also rolling out in-app notifications and a step-by-step migration guide to walk affected users through the process.

A short-lived integration gets the axe Phantom first announced Sui support in December 2024 and officially launched the integration on January 29, 2025. When Phantom added Sui, the move was part of a broader multichain expansion push. The wallet was growing beyond its Solana roots, adding support for Ethereum and Bitcoin alongside newer networks.

The integration gave Phantom users access to Sui ecosystem assets, in-wallet swaps, and portfolio management across Sui without leaving the app.

Phantom dropped support for the Monad network on August 26, just two days after the Sui announcement, suggesting a deliberate thinning of the network roster rather than a one-off call.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.

고객센터

공지사항혜택·이벤트업비트소식거래 이용 안내수수료 안내입출금 이용 안내입출금 현황시장경보 현황Open API 안내정책 및 거래지원 문의1:1 문의하기문의내역이용자 가이드카카오톡 문의(24시간)증명서 발급공시 안내공지사항

거래

NEW

라이터(LIT) KRW 마켓 디지털 자산 추가

안녕하세요. 가장 신뢰받는 디지털 자산 거래소 업비트입니다.

업비트에 아래와 같은 디지털 자산이 KRW 마켓에 추가됩니다.

 
   
   
   
   
 
 
   
   
   
   
 
디지털 자산마켓네트워크거래지원 개시 시점

라이터(LIT)KRWEthereum8월 24일 13시 예정

디지털 자산 입금 전 네트워크를 반드시 확인 바랍니다.
입출금 서비스 개시 이후, 일정 수준의 유동성을 확보하지 못하는 경우 거래지원 개시 시점이 연기될 수 있습니다.

※ 거래 제한 안내
업비트 BTC 마켓에서 제공하는 시세를 기준으로 거래 제한 가격이 결정됩니다.

매수 제한 : 거래 지원 이후 약 5분간 매수 주문이 제한됩니다.
최저 매도가 제한 : 거래 지원 이후 약 5분간 전일 종가 대비 – 10% 이하 가격의 매도 주문이 제한됩니다.
주문 타입 제한 : 거래 지원 이후 약 2시간 동안 지정가 주문을 제외한 모든 타입 및 조건의 주문이 제한됩니다.
전일 종가는 아래 표를 통해 확인 바랍니다.

 
   
     
     
   
 
 
   
     
     
   
 디지털 자산BTC 마켓 전일 종가LIT0.00004500 BTC (4,803 KRW)

※ 입금 유의사항
반영 불가한 입금은 반환 절차에 오랜 시간이 소요될 수 있으니 유의사항을 꼼꼼하게 확인하세요

업비트에서 거래지원하는 LIT의 컨트랙트 주소는 0x232ce3bd40fcd6f80f3d55a522d03f25df784ee2입니다. LIT 입출금 진행 시 컨트랙트 주소를 확인 바랍니다.
트래블룰 이행을 위해, 입출금 가능 가상자산사업자 리스트에 포함 되어있지 않은 거래소를 통해 업비트에 자산이 입금될 경우
'본인 소유 확인'이 완료된 개인지갑 주소로만 입출금이 가능하며, 연동된 개인지갑을 통한 입금 건이더라도 해당 자산의 네트워크에 따라 입금 반환 처리를 해야 할 수 있습니다.
출처가 불분명한 고액의 디지털 자산 입금 시, 자금 출처에 대한 소명이 요청될 수 있습니다. (이용약관 제17조 제8항)

※ 추가 디지털 자산

가상자산 LIT는 이더리움 기반으로 발행된 유틸리티 토큰으로, 탈중앙화 영구선물 거래소 라이터(Lighter) 프로토콜에서 스테이킹과 유동성 참여 구조에 활용됩니다. 라이터는 주문서 기반 퍼페추얼 거래를 제공하는 인프라로, 자산은 L1에 예치되고 거래 실행은 L2 환경에서 처리되는 구조를 갖추고 있습니다. 거래량, 오픈 인터레스트 등 주요 지표는 온체인 데이터와 외부 플랫폼을 통해 확인할 수 있으며, 주문 매칭, 포지션 관리, 청산 및 펀딩 레이트 기능을 포함한 파생상품 거래 기능을 제공합니다. LIT 토큰은 스테이킹을 통해 LLP(Lighter Liquidity Pool) 참여 한도를 부여하는 방식으로 사용됩니다. 또한 일정 수량 이상 스테이킹 시 출금 및 전송 수수료 면제와 거래 수수료 혜택이 제공됩니다.

홈페이지 : 라이터 공식 홈페이지

X(구 트위터) : 라이터 공식 X

백서 : 라이터 공식 백서

✽ PC에서 해당 디지털 자산이 조회되지 않을 경우, 새로고침 (F5) 후 확인 부탁드립니다.

✽ 업비트 App에서 해당 디지털 자산이 조회되지 않을 경우, 더보기 > 화면설정 > 코인정보 다시 받기를 클릭 후 확인 부탁드립니다.

투자 위험 안내

• 디지털 자산 투자는 투기적 수요 및 국내외 규제환경 변화 등에 따라 급격한 시세 변동에 노출될 수 있습니다. 본 디지털 자산의 투자 판단의 책임은 본인에게 있으며, 발생 가능한 손실도 투자자 본인에게 귀속됩니다. 프로젝트 홈페이지 및 공시 자료 등을 면밀히 참고하시어 디지털 자산의 특성을 충분히 인지하시고 신중하게 거래해 주시기를 당부해 드립니다.

• 디지털 자산 거래의 특성상 국내외 거래소간 시세 차이가 지속적으로 발생하고 있습니다. 반드시 디지털 자산 투자 전 글로벌 거래소와의 시세 차이에 유의하시기 바랍니다.

• 업비트는 안정된 거래 환경 조성을 위하여 최소 주문 금액 제한 등 다양한 조치를 취하고 있으나, 과열된 투자 환경에 따라 일부 회원님에게는 주문 안정화 메시지가 노출 될 수 있습니다. 이는 먼저 주문한 회원의 주문을 처리하고, 안정적인 서비스를 위한 불가피한 조치이므로 이 점 유의하시기 바랍니다.

※ 가상자산은 고위험 상품으로 투자금의 전부 또는 일부 손실을 초래할 수 있습니다.

※ 가상자산의 투자 판단 및 그에 따른 원금 손실의 책임은 투자자 본인에게 있습니다.

※ 두나무 주식회사 준법감시인 심사필 제25-0156호 (25.09.22~27.09.21)

공유

PERTH, Australia, Aug. 23, 2026 (GLOBE NEWSWIRE) -- Alkane Resources Limited (ASX: ALK; TSX: ALK; OTCQX: ALKRY) (‘Alkane’ or ‘the Company’) is pleased to announce the latest exploration results highlighting the discovery of a new high-grade zone of the Cuffley Lode, and drilling results from the Sub-KC domain at depth below the Augusta mine and the Costerfield property in Victoria, Australia

Program Summary

23 new holes have been drilled in an unmined area between the historical Cuffley north and south high-grade grade panels, and 17 additional holes have targeted the Sub KC domain at depth below the depositA new grade pod with areas of very high grade gold and antimony was identified in the Cuffley infill area, in a sparsely drilled zone previously thought to contain low grade due to the influence of a crosscutting faultAdditional high grade infill intercepts have made in the Sub KC system, and the first target testing holes looking for repetitions of the structural setup have been drilledThe Cuffley pod is readily accessible from existing infrastructure and is being incorporated into the mine schedule Assay Highlights

From Cuffley 580.9g/t gold and 24% antimony over 0.61m (ETW 0.54m) in AD270168.9g/t gold and 33.5% antimony over 0.9m (ETW 0.78m) in AD26560.1g/t gold and 15.2% antimony over 1.26m (ETW 1.17m) in AD292124g/t gold and 48.6% antimony over 0.23m (ETW 0.22m) in AD275 From Sub KC28.2g/t gold and 0% antimony over 0.99m (ETW 0.82m) in CSK04473.2g/t gold and 18.8% antimony over 0.19m (ETW 0.18m) in CSK043192g/t gold and 0% antimony over 0.17m (ETW 0.07m) in CSK04316.1g/t gold and 12.7% antimony over 0.37m (ETW 0.34m) in CSK048 Alkane Managing Director & CEO, Nic Earner, said:

“This discovery of unmined high-grade material directly adjacent one of Costerfield’s top-shelf historical orebodies showcases the importance of Alkane’s directive of revisiting and challenging old models and preconceptions surrounding mineralisation to extract value. We will continue seeking this new mineralisation alongside generating new targets within our leases.”

Costerfield Gold-Antimony Field

Alkane Resources Ltd 100%

The Costerfield gold-antimony deposit was discovered in 1861, antimony having been already identified in the district as early as 1853 as prospectors attracted to the McIvor (Heathcote) alluvial gold rush began to explore the surrounding hills for the primary deposits. Several lodes along a 3km corridor were rapidly opened up, the bulk of historical production coming from leases at the northern end of the field; the Costerfield (Main), Bombay and Minerva mines. Production from these mines primarily took place in two phases, between 1861-1883 and 1903-1924, and a short-lived attempt at redeveloping the mine occurred between 1933-1939.

Modern mining has been continuous since 2006, when Australian Gold Development commenced underground operations at Augusta, at the southern end of the field. AGD’s Costerfield operation was purchased by Mandalay Resources in 2010, and extraction of the vertically continuous vein system has progressively moved north. Firstly from the initial Augusta series of lodes, to Cuffley and N Lode in 2014 and the Brunswick in 2018. Costerfield’s current locus of mining is beneath the Costerfield, Minerva and Bombay group of mines, where Mandalay’s high-grade Youle and Shepherd lodes were accessed in 2019.

Figure 1. Regional map of the Costerfield Project in GDA2020 grid showing Alkane tenements and the main corridors of mineralisation identified, highlighting the location of the Cuffley and Sub KC deposits.

Deposit Geology

The Cuffley and Sub KC deposits are found within the Central Corridor of deposits at Costerfield. This corridor approximately traces the apex of the Costerfield Dome, a structural high which consists of Silurian marine siltstones with turbiditic intervals becoming common towards the base of the known sequence. The Cuffley Lode occupies a vertical shear, running along a N-S field-scale anticline (Cuffley Anticline) next to the Augusta Deposit. The lode sits on the gently dipping western edge of its 200m-wide hinge zone, N Lode occupies the corresponding eastern axial zone with a steep east-dipping limb. Mineralisation at Cuffley is of the “classic” Costerfield style, consisting of quartz-carbonate veining grading to massive stibnite, gold being found in both quartz and stibnite. Updip, the Cuffley lode is truncated by the Mamushi/Flat Fault set, which offsets the mineralisation above eastward by approximately 40m, where it is known as the historical Alison deposit. The Alison mineralisation is itself bounded updip by the major west-dipping Adder Fault thrust. The footwall of the Cuffley system is delineated by the similarly west-dipping King Cobra Fault thrust, which breaches and offsets the Cuffley anticline.

Down-dip from Cuffley, westward along the King Cobra Fault plane, mineralisation resumes with the Sub-KC deposit. Sub-KC occupies the steeply dipping east limb of the continuation of the Cuffley Anticline. If the approximately 300m of offset along the fault plane is restored, reconstructing the anticline and stratigraphy, the Sub-KC deposit correlates very well with the downward continuation of N Lode. The axis of the anticline hosts a strong gougey shear termed the Lyre Fault, which merges into the King Cobra Fault forming the hangingwall of the main Sub-KC domain between them. The Lyre Fault exhibits some clear post-mineralisation movement, evidenced by clasts of mineralised vein material in fault gouge recovered in earlier Sub-KC drilling. Most mineralisation associated with Sub KC sits in the immediate footwall of the Lyre Fault and reduces in tenor at distance from the fault plane. Strike control is not fully understood at this stage of drilling due to limited drilling orientations, but appear to be related to north-east trending splays from the Lyre Fault somewhat similar in nature to the East Fault at Cuffley. The most important of these are the Bird and Bustard Faults, represented on the Sub-KC long section below.

The veins of the Sub-KC deposit generally fall into three categories:

Early, laminated bedding parallel quartz veins (Figure 7, CSK043), reactivated and dilated during the mineralisation, the later quartz-stibnite vein generation is often discreet and confined to one side of the vein with the lamination adhering to the other contact. Gold is commonly found in the laminations of the older quartz generation.Single-generation quartz veins in subvertical orientations, linking between the laminated, bedded veins (Figure 7, CSK044). These structures most likely developed under extensional stress contemporaneous with the time of mineralisation.East-dipping veins found on the western limb of the Cuffley Anticline, crosscutting the bedding at a high angle. These veins appear to be exploiting an earlier axial, spaced fabric including jointing and minor faulting developed at the time of anticline formation. The existence of these veins indicate that while the Lyre Fault has some post-mineralisation movement, it is not necessarily a hard boundary to mineralisation in this domain. All veins have demonstrated the capability to host coarse, high-grade gold and varying amounts of massive stibnite. In general, antimony grades are of lesser importance within the Sub-KC deposit relative to other mined deposits at Costerfield, perhaps due to comparatively limited vein volumes which does not appear to influence gold prospectivity. “Rusty” gold after aurostibite is commonly observed, along with occasional veins containing native antimony metal (Figure 8) typically with small amounts of pyrrhotite. This assemblage indicates minor activity of a lower sulphur, relatively reduced ore fluid phase.

Drilling Results - Cuffley

23 growth and infill holes have been completed, totalling 3,255m of diamond drill core. This activity resulted in 18 new intercepts on the mineralised structure, the remaining five fault blanking as the bounds of the new domain was explored. Four drillholes intercepted mineralisation grading over 10g/t gold equivalent over mining width:

580.9g/t gold and 24% antimony over 0.61m (ETW 0.54m) in AD270 Including 0.25m @ 1360g/t gold & 19.7% antimony (Figure 6) 168.9g/t gold and 33.5% antimony over 0.9m (ETW 0.78m) in AD26560.1g/t gold and 15.2% antimony over 1.26m (ETW 1.17m) in AD292124g/t gold and 48.6% antimony over 0.23m (ETW 0.22m) in AD275 An additional four holes graded over 2g/t gold equivalent over mining width.

Drilling Results – Sub KC

Four target-testing holes were drilled for 3705.32m, and ten parent / two wedge growth and infill holes were completed for 8168.28m.

Several ore-grade intercepts were made within the system, including:

28.2g/t gold and 0% antimony over 0.99m (ETW 0.82m) in CSK04473.2g/t gold and 18.8% antimony over 0.19m (ETW 0.18m) in CSK043192g/t gold and 0% antimony over 0.17m (ETW 0.07m) in CSK04316.1g/t gold and 12.7% antimony over 0.37m (ETW 0.34m) in CSK048 The target testing holes explored the footwall of the Lyre Fault along strike north and south of the main Sub-KC block, testing for repetitions in zones modelled to have favourable structural setup. The two holes drilled approximately 400m north of the main block found the Lyre Fault plane to have stepped eastward in position, faulting out much of the inferred favourable east-dipping fold limb. These two holes intercepted moderate grades in the Adder Fault, located just into the wall from the collar point.

Conversely, the holes drilled approximately 300m south of the Sub-KC zone found an intact anticlinal position, without the presence of the expected Lyre Fault, which appears to be located further westward at this point. Both holes encountered veins with anomalous gold however no ore-grade intercepts were made.

Figure 2. Long Section of the Cuffley System with major vein target envelopes displayed, recent drill traces and > 6g/t AuEq new intercepts labelled. New significant intercepts not associated with a named structure are represented as triangular icons. Older significant drill intercepts are displayed as smaller, unlabelled icons. Previous mining on the Cuffley Lode with face assays and depleted areas area are also shown.

Figure 3. Long Section of the Sub-KC System with major vein target envelopes displayed, recent drill traces and > 6g/t AuEq new intercepts labelled. Older significant drill intercepts are displayed as smaller, unlabelled icons. New significant intercepts not associated with a named structure are represented as triangular icons.

Figure 4. Plan Section of the Cuffley and Sub KC deposits with vein best fit traces displayed, recent drill traces and > 6g/t AuEq new intercepts labelled. Older significant drill intercepts are displayed as smaller, unlabelled icons. New significant intercepts not associated with a named structure are represented as triangular icons.

Figure 5. Cross section looking north at mine northing 4900N showing the Cuffley and Sub-KC systems (veins represented schematically), and > 6g/t AuEq new intercepts labelled. Older significant drill intercepts are displayed as smaller, unlabelled icons. New significant intercepts not associated with a named structure are represented as triangular icons.

Figure 6. Core tray photo of the high-grade Cuffley intercept in AD270. Note the very high gold interval of 1360g/t Au, and consistent high-grade antimony. Please refer to Appendix 1 for the relevant assay results relating to Figure 6.

Figure 7. Core tray photographs of the Sub-KC 405 Lode intercepts from drillholes CSK043 (bottom) and CSK044 (top). Intervals with grade above detection levels are labelled. Note the bedding-parallel nature of the major veins, due to reactivation of early laminated quartz structures acting as host structure, and the additional occurrence of high grade visible gold in very narrow veinlets (CSK044, 537m). Please refer to Appendix 1 for the relevant assay results relating to Figure 7.

Figure 8. Vein containing a significant volume of native antimony metal (metallic white) intercepted in CSK045W1 (538.15m). The vein also contained stibnite and pyrrhotite, and pyrite wallrock alteration can be seen in the image. Please refer to Appendix 1 for the relevant assay results relating to Figure 8.

Future Plans

The new grade pod at Cuffley has been integrated into the Costerfield mine plan and is scheduled to be mined. Considerable scope remains for further growth within the Sub KC domain with the structural information gained from the north and south target testing holes, however drilling will likely need to be undertaken from surface, or future development with improved intersection angles with the target areas.

This document has been authorised for release to the market by Nic Earner, Managing Director.

ABOUT ALKANE ‐ www.alkres.com ‐ ASX:ALK | TSX: ALK | OTCQX: ALKRY

Alkane (ASX:ALK; TSX:ALK; OTCQX:ALKRY) is an Australia-based gold and antimony producer with a portfolio of three operating mines across Australia and Sweden. The Company has a strong balance sheet and is positioned for further growth.

Alkane’s wholly owned producing assets are the Tomingley open pit and underground gold mine southwest of Dubbo in Central West New South Wales, the Costerfield gold and antimony underground mining operation northeast of Heathcote in Central Victoria, and the Björkdal underground gold mine northwest of Skellefteå in Sweden (approximately 750 km north of Stockholm). Ongoing near-mine regional exploration continues to grow resources at all three operations.

Alkane also owns the very large gold-copper porphyry Boda-Kaiser Project in Central West New South Wales and has outlined an economic development pathway in a Scoping Study. The Company has ongoing exploration within the surrounding Northern Molong Porphyry Project and is confident of further enhancing eastern Australia’s reputation as a significant gold, copper and antimony production region.

Competent Persons Statement

Certain information in this announcement relating to Exploration Results has been previously released to the ASX (refer to ASX announcement dated 14 July 2026 titled ‘Alkane Extends High Grade Gold Trend at Brunswick South’). Alkane confirms that it is not aware of any new information or data that materially affects the information included in those market announcements and that all material assumptions and technical parameters underpinning the estimates and Exploration Results in those announcements continue to apply and have not materially changed.

The information in this report that relates to the Costerfield Exploration Results is based on, and fairly represents, information compiled and verified by Mr Chris Davis. Mr Davis is a Chartered Professional (Geology) of the Australasian Institute of Mining and Metallurgy (MAusIMM CP(Geo)), and a Member of the Australian Institute of Geoscientists (MAIG).

Mr Davis has sufficient experience that is relevant to the style of mineralisation and type of deposit under consideration and to the activity being undertaken to qualify as a Competent Person as defined in the 2012 Edition of the “Australian Code for Reporting of Exploration Results, Mineral Resources, and Ore Reserves” (JORC Code).

For the purposes of National Instrument 43-101 – Standards of Disclosure for Mineral Projects (‘NI 43-101’), the scientific and technical information contained in this announcement relating to the Costerfield Exploration Results has been prepared under the supervision of, and approved by, Mr Chris Davis, who is a “qualified person” as defined in NI 43-101. Mr Davis is employed by Alkane as Chief Geologist and, as an employee of Alkane, is not considered independent of Alkane within the meaning of NI 43-101.

Mr Davis consents to the inclusion in this report of the matters based on his information in the form and context in which they appear.

Cautionary Note Regarding Forward-Looking Information and Statements

This announcement contains certain forward-looking information and forward-looking statements within the meaning of applicable securities legislation and may include future-oriented financial information or financial outlook information (collectively Forward-Looking Information). Actual results and outcomes may vary materially from the amounts set out in any Forward-Looking Information. As well, Forward-Looking Information may relate to: future outlook and anticipated events; expectations regarding exploration potential; production capabilities and future financial or operating performance, including AISC, investment returns, margins and share price performance; production and cost guidance and the timing thereof; issuing updated resources and reserves estimate and the timing thereof; the potential of Alkane to meet industry targets, public profile and expectations; and future plans, projections, objectives, estimates and forecasts and the timing related thereto.

Forward-Looking Information is generally identified by the use of words like "will", "create", "enhance", "improve", "potential", "expect", "upside", "growth" and similar expressions and phrases or statements that certain actions, events or results "may", "could", or "should", or the negative connotation of such terms, are intended to identify Forward-Looking Information.

Although Alkane believes that the expectations reflected in the Forward-Looking Information are reasonable, undue reliance should not be placed on Forward-Looking Information since no assurance can be provided that such expectations will prove to be correct. Forward-Looking Information is based on information available at the time those statements are made and/or good faith belief of the officers and directors of Alkane as of that time with respect to future events and are subject to risks and uncertainties that could cause actual results to differ materially from those expressed in or suggested by the Forward-Looking Information. Forward-Looking Information involves numerous risks and uncertainties. Such factors include, without limitation: risks relating to changes in the gold and antimony price.

Forward-Looking Information is designed to help readers understand Alkane’s views as of that time with respect to future events and speak only as of the date they are made. Except as required by applicable law, Alkane assumes no obligation to update or to publicly announce the results of any change to any forward-looking statement contained or incorporated by reference herein to reflect actual results, future events or developments, changes in assumptions or changes in other factors affecting the Forward-looking Information. If Alkane updates any one or more forward-looking statements, no inference should be drawn that the company will make additional updates with respect to those or other Forward-looking Information. All Forward-Looking Information contained in this announcement is expressly qualified in its entirety by this cautionary statement.

Disclaimer

Alkane has prepared this announcement based on information available to it. No representation or warranty, express or implied, is made as to the fairness, accuracy, completeness or correctness of the information, opinions or conclusions contained in this announcement. To the maximum extent permitted by law, none of Alkane, its directors, officers, employees, associates, advisers and agents, nor any other person accepts any liability, including, without limitation, any liability arising from fault or negligence on the part of any of them or any other person, for any loss arising from the use of this announcement or its contents or otherwise arising in connection with it.

This announcement is not an offer, invitation, solicitation, or other recommendation with respect to the subscription for, purchase or sale of any security, and neither this announcement nor anything in it shall form the basis of any contract or commitment whatsoever.

APPENDIX 1 – Tabulated Drilling Results
Significant intercepts from the Cuffley and Sub KC drilling programs at Costerfield

Drill Hole IDFrom (m)To (m)Interval (m)Estimated
True Width (m)Au (g/t)Sb (%)Gold-equiv.
grade diluted
to 1.8 m (g/t)Interpreted
VeinAD26578.3179.210.900.78168.933.5107.4CuffleyAD26697.5098.701.200.886.72.46.1CuffleyAD27075.7076.310.610.54580.924.0192.0CuffleyIncluding76.0676.310.250.201360.019.7  AD27280.5980.960.370.290.00.00.0CuffleyAD27389.5389.690.160.134.21.20.5CuffleyAD27472.4272.560.140.130.82.40.5CuffleyAD27576.1176.340.230.22124.048.628.8CuffleyAD27680.7880.950.170.140.50.00.0CuffleyAD27780.6581.030.380.3317.80.93.6CuffleyAD27873.8174.460.650.616.90.93.1CuffleyAD28382.5283.020.500.410.10.10.1CuffleyAD28681.7682.080.320.260.00.00.0CuffleyAD290105.50105.620.120.0738.519.13.5CuffleyAD29272.7474.001.261.1760.115.262.9CuffleyAD29395.8196.070.260.180.00.00.0CuffleyCSK043503.40503.660.260.230.10.20.1Sub KC 401CSK044496.86497.420.560.510.11.20.9Sub KC 401CSK048522.80523.170.370.3416.112.78.8Sub KC 401CSK032461.24461.420.180.150.00.00.0Sub KC 402CSK042533.22533.440.220.200.10.00.0Sub KC 402CSK043511.23511.420.190.1873.218.812.0Sub KC 402CSK044504.42505.270.850.841.60.10.9Sub KC 402CSK046525.64525.800.160.150.30.00.0Sub KC 402CSK048531.03531.230.200.1910.626.27.6Sub KC 402CSK042497.87498.000.130.120.00.00.0Sub KC 405CSK043491.10491.210.110.103.20.00.2Sub KC 405CSK044484.73485.030.300.301.90.00.3Sub KC 405CSK048508.43508.560.130.1222.911.33.4Sub KC 405CSK043547.70549.091.391.150.70.30.8Sub KC 410CSK044536.70537.690.990.8228.20.012.9Sub KC 410CSK045549.38549.630.250.208.60.31.0Sub KC 410CSK045W1543.82547.083.262.701.00.72.7Sub KC 410CSK043712.30712.470.170.07192.00.06.9Sub KC 420CSK045670.90671.901.000.400.20.00.1Sub KC 420CSK035942.41942.540.130.100.00.00.0Sub KC 425CSK040667.34668.030.690.400.10.00.0Sub KC 425CSK040W1634.05636.001.951.243.20.02.2Sub KC 425CSK042624.48624.710.230.1252.30.03.6Sub KC 425AD269106.20108.382.180.782.81.02.3Cuffley Assoc.AD269111.51111.810.300.121.811.42.0Cuffley Assoc.AD27777.9278.030.110.095.914.72.1Cuffley Assoc.AD27779.2379.670.440.362.31.91.4Cuffley Assoc.AD27978.9279.130.210.2030.321.78.9Cuffley Assoc.AD27980.6281.060.440.2099.10.111.0Cuffley Assoc.AD27985.6686.460.800.594.40.21.6Cuffley Assoc.AD281119.20119.540.340.2540.63.66.8Cuffley Assoc.AD291170.25170.700.450.1015.95.71.7Cuffley Assoc.AD29279.6579.790.140.1122.10.01.3Cuffley Assoc.AD294165.19165.300.110.087.08.01.1Cuffley Assoc.AD294177.26177.580.320.174.719.94.9Cuffley Assoc.CSK0364.584.760.180.152.038.47.9Adder Ft Assoc.CSK040725.30726.000.700.577.30.02.3Sub KC Assoc.CSK043487.70487.830.130.116.728.64.7Sub KC Assoc.CSK043518.47518.600.130.112.37.31.2Sub KC Assoc.CSK044533.36533.560.200.2089.90.09.9Sub KC Assoc.CSK044573.44574.320.880.857.20.03.4Sub KC Assoc.CSK045W1538.15538.480.330.3225.45.77.1Sub KC Assoc.CSK045W1549.90550.250.350.317.30.01.3Sub KC Assoc.CSK045W1554.05554.630.580.553.50.01.1Sub KC Assoc.CSK048537.98538.200.220.214.32.11.1Sub KC Assoc.  Notes
 1.The AuEq (gold equivalent) grade is calculated using the following formula:
                              AuEq g per t = Au g per t + Sb%   x
 Sb price per 10kg × Sb processing recovery     Au price per g × Au processing recovery   
Prices and recoveries used: Au $/oz = 2,500 (Au US$/gram = 80.39); Sb $/t = 19,000 (Sb US$/10kg = 190); Au Recovery = 91% and; Sb Recovery = 92%. The Au recovery assumption and Sb recovery assumption is based on established processing and sales in respect of Costerfield. It is the Company’s opinion that all elements included in the metal equivalent calculation have a reasonable potential to be recovered and sold. 2.The estimated true width of composites that are not interpreted to be connected to a major vein (identified as “Other” in the above table) have been calculated using a generic, conservative intercept angle (alpha angle) of 45 degrees. 3.Composites that are not interpreted to be connected to a major vein and are below 1 g/t AuEq when diluted to 1.8m are not considered significant and are not recorded here. 4.Gold-equivalent grades for intervals with estimated true width >1.8m are not diluted.
      Drill hole collar details from the Cuffley and Sub KC drilling at Costerfield covered in this release:

Hole IDNorthingEastingElevationDepthAzimuthDipDate CompletedAD265480615237968140.10314.1-0.32/01/2026AD266480615238968129.00323.1-12.65/01/2026AD269480615239968142.80331.2-19.36/02/2026AD270480115237969121.73308.114.913/01/2026AD272480215237969118.24321.49.116/01/2026AD273480615238968137.16321.7-6.512/02/2026AD274479915237969104.30274.917.917/01/2026AD275480015237968121.79301.8-2.018/01/2026AD27648061523896895.20310.7-9.025/01/2026AD277480015238969156.09305.224.021/01/2026AD278480115237969145.91291.221.023/01/2026AD27948051523897198.02295.034.226/01/2026AD280480615238971151.98314.835.229/01/2026AD281480615239967161.44327.3-28.110/02/2026AD283480615238968164.60316.9-10.01/02/2026AD284480615238968149.50325.3-19.84/02/2026AD285480615238969160.77315.515.325/02/2026AD286480615238970157.77317.723.623/02/2026AD290480615239968186.50334.9-14.115/02/2026AD291480615239968182.90338.0-22.020/02/2026AD292480615237969128.02307.14.823/04/2026AD293480615239969122.30331.76.926/04/2026AD294480615238970179.10327.418.929/04/2026CSK032503515301913899.50274.9-71.527/02/2024CSK033503515301913602.62291.2-67.019/03/2024CSK0344489154631030839.40263.5-72.131/01/2025CSK034W14489154631030940.00263.5-72.18/02/2025CSK0354489154631030995.57287.9-79.013/03/2025CSK036571014939945872.50133.8-33.810/03/2025CSK037571014939944881.35133.0-49.417/05/2025CSK040474415474865732.57290.0-55.69/06/2025CSK040W1474415474865656.40290.0-55.621/06/2025CSK042475815461865667.12302.4-50.122/06/2025CSK043475715461865934.90296.3-46.031/07/2025CSK044474415474865668.01280.3-48.021/09/2025CSK045475615461865760.50285.2-52.629/08/2025CSK045W1475615461865600.00285.2-52.67/09/2025CSK046475715461865659.14297.6-48.126/09/2025CSK048474515474866752.50298.9-44.020/10/2025CSK049474515474865760.22302.2-53.08/11/2025 Notes:

Coordinate System: Costerfield Local Mine Grid Appendix 2 - JORC Code, 2012 Edition – Table 1

Section 1 Sampling Techniques and Data

CriteriaJORC Code explanationCommentarySampling techniques Nature and quality of sampling (e.g. cut channels, random chips, or specific specialised industry standard measurement tools appropriate to the minerals under investigation, such as down hole gamma sondes, or handheld XRF instruments, etc.). These examples should not be taken as limiting the broad meaning of sampling.Include reference to measures taken to ensure sample representivity and the appropriate calibration of any measurement tools or systems used.Aspects of the determination of mineralisation that are Material to the Public Report.In cases where ‘industry standard’ work has been done this would be relatively simple (e.g. ‘reverse circulation drilling was used to obtain 1 m samples from which 3 kg was pulverised to produce a 30 g charge for fire assay’). In other cases more explanation may be required, such as where there is coarse gold that has inherent sampling problems. Unusual commodities or mineralisation types (e.g. submarine nodules) may warrant disclosure of detailed information. Sampling of Au and Sb mineralisation is from diamond drill core (HQ2 and NQ2).Due to the discrete mineralisation of the deposit, not all diamond drill core was required to be sampled. Sample intervals were determined and marked on the core by Alkane geologists using the following general rules:

All stibnite-bearing veins are sampled.Intersections of polyphase breccias, stockwork veins, laminated quartz veins or massive quartz veins were routinely sampled.A waste sample is taken either side of the mineralized vein (30–100 cm).Siltstone is sampled where disseminated arsenopyrite is prevalent.Fault gouge zones were sampled at the discretion of the geologist. Diamond core sampling intervals were standardised wherever possible and ranged from 5 cm to 1 m in length. Diamond drill core samples have been cut in half using the orientation line or cut line, with a consistent side of the cut core selected for assay to ensure unbiased sampling. The methodology was validated by the Costerfield QA/QC protocols. No sampling instruments required calibration. Assays were completed by On Site in Bendigo, which is independent of Alkane and holds current ISO/IEC 17025 accreditation. The general methods were as follows:

Gold grades were determined by either fire assay (25 g charge) with an AAS finish, screen fire assay or Chrysos photon assay technology.Antimony concentrations were determined using an aqua regia based acid digest with an AAS finish. Drilling techniques Drill type (e.g. core, reverse circulation, open-hole hammer, rotary air blast, auger, Bangka, sonic, etc) and details (e.g. core diameter, triple or standard tube, depth of diamond tails, face-sampling bit or other type, whether core is oriented and if so, by what method, etc.). Deepcore Drilling is the drilling contractor utilised for the whole of this project within the reporting period. All diamond drilling was completed from underground and was completed using LM90 drill rigs utilising HQ2 and NQ2 diameters. Core orientation is performed each run, typically using an AXIS Champ Ori kit.Drill sample recovery Method of recording and assessing core and chip sample recoveries and results assessed.Measures taken to maximise sample recovery and ensure representative nature of the samples.Whether a relationship exists between sample recovery and grade and whether sample bias may have occurred due to preferential loss/gain of fine/coarse material. Diamond drilling was routinely checked for core loss during both drilling and sampling. Core loss blocks were added by drillers and then checked by geologists or field technicians when the core was measured, and depth marks made. If problems were encountered with recovery and core block depths, the drill shift supervisor was advised and depth marking stopped until the issue was rectified.No relationship between grade and sample recovery has been established. Ore zones with poor recovery are redrilled until a representative sample is achieved.

Logging Whether core and chip samples have been geologically and geotechnically logged to a level of detail to support appropriate Mineral Resource estimation, mining studies and metallurgical studies.Whether logging is qualitative or quantitative in nature. Core (or costean, channel, etc.) photography.The total length and percentage of the relevant intersections logged. All drill core was geologically logged as full core for the relevant rock quality designation, lithology, structural data, and sample intervals.Data capture was digital into the AcQuire software using validated codes.

All drill core was photographed wet with high resolution photographs stored on the site’s server, which is routinely backed-up.

Sub-sampling techniques and sample preparation If core, whether cut or sawn and whether quarter, half or all core taken.If non-core, whether riffled, tube sampled, rotary split, etc., and whether sampled wet or dry.For all sample types, the nature, quality and appropriateness of the sample preparation technique.Quality control procedures adopted for all sub-sampling stages to maximise representivity of samples.Measures taken to ensure that the sampling is representative of the in situ material collected, including for instance results for field duplicate/second-half sampling.Whether sample sizes are appropriate to the grain size of the material being sampled. Diamond core sampling intervals were standardised wherever possible and ranged from 5 cm to 1 m in length. Diamond drill core samples have been halved for sampling (whole core sampled if representative halving was not possible) guided by the orientation line or a cut line, with a consistent side of the cut core selected for assay to ensure unbiased sampling. The following sample preparation activities were undertaken by Alkane staff for both diamond drill core and underground channel samples:

Sample information and characteristics were measured, logged, recorded in the acQuire database and assigned a unique sample ID.Sample material was placed into a calico bag previously marked with the unique sample ID.Calico bags were loaded into plastic bags such that the plastic bags weighed less than 10 kg.An assay submission sheet was generated and placed into the plastic bag.Plastic bags containing samples were sealed with a metal or plastic tie and transported to On Site in Bendigo via private courier or Alkane staff. The following sample preparation activities were undertaken by On Site staff:
Samples were received and checked for labelling, missing samples, etc. against the submission sheet.If the sample batch matched the submission sheet, sample metadata were entered into On Site’s LIMS. In the event that discrepancies were noted, Mandalay Resources was contacted by On Site to resolve the discrepancy prior to further work commencing. Records of all discrepancies and corrective actions taken are recorded by the Mandalay Resources database administrator.A job number was assigned, and worksheets and sample bags were prepared.Samples were placed in an oven and dried overnight at 106°C.Samples were weighed and recorded.The entire dried sample was crushed using a Rocklabs Smart BOYD Crusher RSD Combo with a jaw closed side setting of 2 mm.If the dried sample weight was less than 3 kg, the entire sample was retained for pulverisation. If the dried sample weight was greater than 3 kg, the sample was spilt to 3 kg using the rotary splitter that is incorporated in the BOYD crusher.Rejects from splits greater than 3 kg were retained as coarse rejects in labelled calico bags and returned to Mandalay Resources.The 3 kg sample was then pulverised in an Essa LM5 Pulverising Mill to 90% passing 75 µm. For fire assay and base metal samples:
The 3 kg pulverised samples were then subsampled to take a master ~200 g pulp split for assay by a manual scooping procedure across the full width and depth of the mill bowl and loaded sequentially into labelled pulp packets. For photon assay:
The ~3 kg pulverised samples were then subsampled to fill a ~280 g photon assay jar by a manual scooping procedure across the full width and depth of the mill bowl. For all methods:
For every 21 primary samples, a sample was randomly selected by LIMS and a duplicate 200 g split for fire assay or second jar for photon assay was submitted for analysis using the same analytical procedure as the primary sample.The remaining pulp was returned to its sample bag and then returned to Mandalay Resources for retention following the completion of assay. A quarterly check-assay program is in place to monitor the representative nature of sampling and assay methodology.Quality of assay data and laboratory tests The nature, quality and appropriateness of the assaying and laboratory procedures used and whether the technique is considered partial or total.For geophysical tools, spectrometers, handheld XRF instruments, etc., the parameters used in determining the analysis including instrument make and model, reading times, calibrations factors applied and their derivation, etc.Nature of quality control procedures adopted (e.g. standards, blanks, duplicates, external laboratory checks) and whether acceptable levels of accuracy (i.e. lack of bias) and precision have been established. The assaying protocols used at Costerfield have been developed to ensure expected levels of accuracy and precision are met for the style of mineralisation tested and utilised in the MRE. Samples were assayed for gold, antimony, arsenic, and iron using representative partial digest methodologies:

Gold grades were determined either by a 25g charge with lead flux fire assay and an AAS finish, or by Chrysos photon assay technology.Antimony, iron and arsenic concentrations were determined using an aqua regia based acid digest with an AAS finish. The quality control procedures utilised at Costerfield used CRMs prepared by commercial laboratories Geostats and OREAS.CRMs were either prepared using Costerfield material or were otherwise matrix matched to ensure a representative nature.

At least one CRM was submitted with every batch of diamond core samples and typically at a rate of 1 standard per 25 samples. Up to six CRMs covering the expected ranges of gold and antimony mineralisation were in rotation during routine sampling.

An assay result for a CRM was considered acceptable when the returned assay fell within three standard deviations of the CRM certification grade. Outside this range, the CRM assay was considered to have failed and all significant mineralised samples within the batch were re-assayed, where significant grades were defined as mineralised samples that may have a material-impact in future resource estimates. All actions or outcomes were recorded as comments in the QA/QC register.

Alkane submitted uncrushed samples of basalt as blank material sourced from Geostats into assay sample lots, at a rate of 1 in every 30 samples, to test for contamination during sample preparation.

The failure threshold for gold is 0.10 g/t, which was chosen since it represents ten times the detection limit of 0.01 g/t for AAS. The failure threshold for antimony is 0.05%, which was chosen for being five times the detection limit of 0.01% for AAS.

Pulp duplicates were collected routinely at a rate of 1:22 by On Site and submitted with the primary sample for analysis. Precision was in line for the expected a variance in both gold and antimony.

Umpire laboratory checks to three additional commercial assay laboratories are completed each year covering all new assays generated at the property.

Verification of sampling and assaying The verification of significant intersections by either independent or alternative company personnel.The use of twinned holes.Documentation of primary data, data entry procedures, data verification, data storage (physical and electronic) protocols.Discuss any adjustment to assay data. Sampling intervals and numbering were validated by geologists prior to cutting, with pre-numbered sampling bags systematically used by the field technicians to ensure the correct sample was submitted under each ID. Internal validation of significant intercepts was completed by the exploration and senior geologists. Photographs, logging, sample weights and assay results were checked to ensure manual errors were eliminated.

Key intercepts at Costerfield were also validated by the Resource Geologist and Competent Person during the interpretation and modelling or the Costerfield resource estimation.

Assay and sampling data was automatically uploaded into the Acquire database system and QA/QC validated at the point of upload. Any issues were entered into a QA/QC register and resolved before data acceptance.

Alkane staff conduct periodic visits to the On Site Laboratory in Bendigo and meet regularly with the Lab managers. In early 2023 a review was conducted by a third party (RSC Consulting Pty Ltd) to ensure the practices are appropriate. Nothing of major concern was found.

Twinned holes are typically only drilled intentionally to get full recovery of an ore zone when the initial hole has core loss. There are inadvertent twinned intercepts within the database, particularly when the collar position is close to the mineralisation. Twinned intercepts provide consistent correlation of structure and mineralisation character however due to the short range grade variability common structurally controlled gold systems, may not have the same mineralisation tenor. No adjustment has been made to the assay data.

Location of data points Accuracy and quality of surveys used to locate drill holes (collar and downhole surveys), trenches, mine workings and other locations used in Mineral Resource estimation.Specification of the grid system used.Quality and adequacy of topographic control. Drill hole collar locations have been determined by differential GPS or theodolite surveying methods, either by external surveyors or Alkane surveyors. A digital report is created and entered into the acQuire Database. Data entry accuracy is validated against a LiDAR topographic map and high-resolution satellite imagery.Downhole surveys are conducted using a digital Reflex EZ-TRAC tool, in both single-shot (30 m while drilling) and multi-shot mode (3 m spacing at end of hole) where required.

All downhole survey data is digitally uploaded to the Reflex EZ-TRAC and automatically imported into the acQuire database.

Data spacing and distribution Data spacing for reporting of Exploration Results.Whether the data spacing and distribution is sufficient to establish the degree of geological and grade continuity appropriate for the Mineral Resource and Ore Reserve estimation procedure(s) and classifications applied.Whether sample compositing has been applied. The data spacing at Costerfield is variable. Initial drilling on any particular lode is sporadic but generally approximates 100 × 100 m spacing. This approach is considered appropriate for establishing a geological and grade continuity acceptable for an Inferred Mineral Resource. Following initial drilling and prior to mining, each lode is drilled to a spacing of approximately 40 m × 40 m. This is reduced in areas of structural complexity. This approach is considered appropriate for establishing a geological and grade continuity acceptable for an Indicated Mineral Resource.Where veins or mineralisation zones were sub-sampled, a full-length composite of variable thickness was used in the MRE.

Orientation of data in relation to geological structure Whether the orientation of sampling achieves unbiased sampling of possible structures and the extent to which this is known, considering the deposit type.If the relationship between the drilling orientation and the orientation of key mineralised structures is considered to have introduced a sampling bias, this should be assessed and reported if material. Drill holes at Costerfield are designed to ensure an Alpha angle greater than 30°, indicating that the orientation of the drill holes (and therefore samples) is appropriate for the structure. The drilling orientation compared to that of key mineralised structures is not considered to have introduced any sampling bias as the structures are currently interpreted.

Sample security The measures taken to ensure sample security. All drill core was delivered to the Brunswick site, which is securely gated, with video surveillance, and time stamped swipe card access.Drill core logging and sampling was completed in this secure facility.

Sample bags containing sample material are placed in heavy duty plastic bags in which the sample submission sheet is also included. The plastic bags are sealed with a metal twisting wire or heavy-duty plastic cable ties.

The bags are taken to a storage area that is under constant surveillance.

A private courier collects samples daily and transports them directly to On Site in Bendigo, where they are accepted by laboratory personnel.

Sample pulps from On Site are returned to Alkane for storage. The pulps are stored undercover, wrapped in plastic.

Audits or reviews The results of any audits or reviews of sampling techniques and data. Internal reviews of the exploration process and procedures are completed by senior geologists.Routine monthly lab visits and reviews are conducted by site personnel and make up part of the QA/QC protocols.

RSC Consulting Pty Ltd reviewed the sampling and QA/QC procedures and practices in early 2023. There were no major outcomes related to sampling techniques and data.

Section 2 Reporting of Exploration Results
Criteria listed in the Section 1 also apply to this section.

CriteriaJORC Code explanationCommentaryMineral tenement and land tenure status Type, reference name/number, location and ownership including agreements or material issues with third parties such as joint ventures, partnerships, overriding royalties, native title interests, historical sites, wilderness or national park and environmental settings.The security of the tenure held at the time of reporting along with any known impediments to obtaining a licence to operate in the area. Alkane manages the Costerfield Operation and holds a 100% interest in licences MIN4644, MIN5567, EL5432, EL5519, EL6842, EL6847, EL8320 and RL007485 which comprise the Property. There are no advanced projects in the immediate vicinity of the Property, and there are no other Augusta-style gold-antimony operations in production within the Costerfield district. Exploration on adjacent tenements (EL5546, EL006504, EL006280, EL5490, EL006001, EL6951, EL7352, EL007348, EL007366, EL007382, EL007498, EL007499 and EL007481.

There are currently no known impediments to obtaining a licence to operate in the area. Alkane and its subsidiaries have been conducting both exploration activities and mining activities on the adjacent mining lease MIN4644 since 2006.

Exploration done by other parties Acknowledgment and appraisal of exploration by other parties. The Costerfield Property has been explored using modern methods since 1966. Previous exploration by Mandalay Resources (2009–2025), prior to its merger with Alkane, represents the most significant period of exploration having discovered Cuffley, Youle and Shephard lodes in that time. Exploration Results prior to this have either been validated by more result drilling or are not considered material to the project.Geology Deposit type, geological setting and style of mineralisation. Narrow vein, gold-antimony and gold-only lodes are the targeted deposit styles at the Costerfield Property. Economic lode material consists of either a ‘typical’ gold-bearing quartz and carbonate with massive stibnite, or gold-only quartz and carbonate veining as seen in the Shepherd system. The mineralised shoots are understood to be structurally controlled, typically by the intersection of the lodes with major cross-cutting, gouge filled fault structures and shears. Notable west to northwest dipping thrust faults typically bound the mineralisation packages at the Costerfield Property but can become significantly mineralised themselves along the fault planes. Shallower and dominantly west dipping thrust faults, typically at very low angles or even parallel to bedding with a laminated quartz component, link between the larger order thrust faults. The link faults can also offset the vertical lode structures up to 50 m in an east–west sense. This structural framework leads to the subvertical, north–south extensional veining seen in the Augusta, Brunswick, True Blue and Shepherd systems, along with the moderately west-dipping fault reactivated deposit at Youle.

Drill hole Information A summary of all information material to the understanding of the exploration results including a tabulation of the following information for all Material drill holes: easting and northing of the drill hole collarelevation or RL (Reduced Level – elevation above sea level in metres) of the drill hole collardip and azimuth of the holedownhole length and interception depthhole length. If the exclusion of this information is justified on the basis that the information is not Material and this exclusion does not detract from the understanding of the report, the Competent Person should clearly explain why this is the case. Refer to Appendix 1 for the summary of drill holes related to the Costerfield Property.Data aggregation methods In reporting Exploration Results, weighting averaging techniques, maximum and/or minimum grade truncations (e.g. cutting of high grades) and cut-off grades are usually Material and should be stated.Where aggregate intercepts incorporate short lengths of high grade results and longer lengths of low grade results, the procedure used for such aggregation should be stated and some typical examples of such aggregations should be shown in detail.The assumptions used for any reporting of metal equivalent values should be clearly stated. Reported Exploration Results are intercept length weighted with no truncation of minimum and/or maximum grade applied. Exploration Results have been reported to represent the discrete structural shear or vein as determined by the resource geologist and Competent Persons. There is no cut-off grade for the inclusion of drill intercept if it is on structure.

Aggregates are full-width of target structures/lodes and limited in true width to underground ore development widths of mining of 4.5 m and rely on structures being interpreted as parallel in orientation and representative in nature of the continuous vein.

Gold is the dominant element of value and exploration results are reported as gold equivalent (AuEq) where:

                               AuEq = Au (g/t) + 2.39 x Sb (%)

And the AuEq factor of 2.39 is calculated:

at a gold price of US$2,500/ozan antimony price of US$19,000/twith 2025 predicted metal recoveries of 91% Au and 92% Sb. Relationship between mineralisation widths and intercept lengths These relationships are particularly important in the reporting of Exploration Results.If the geometry of the mineralisation with respect to the drill hole angle is known, its nature should be reported.If it is not known and only the down hole lengths are reported, there should be a clear statement to this effect (e.g. ‘downhole length, true width not known’). Exploration Results that have been included in the resource are reported as drill widths and true widths as determined by the drill hole orientation relative to the vein. Those results not yet included in the resource have been reported as drill widths and estimated true widths.Diagrams Appropriate maps and sections (with scales) and tabulations of intercepts should be included for any significant discovery being reported These should include, but not be limited to a plan view of drill hole collar locations and appropriate sectional views. Appropriate cross sections, plan sections and long sections are included in the body of the report.Balanced reporting Where comprehensive reporting of all Exploration Results is not practicable, representative reporting of both low and high grades and/or widths should be practiced to avoid misleading reporting of Exploration Results. For veins that are interpreted though multiple drill holes all intercepts are tabulated in Appendix 1 and illustrated in the images within the body of the report. Any intercepts that are not interpreted at this stage, to be part of a wider structure are tabulated in Appendix 1 if the sampled grade is above 2g/t when diluted to 1.8m.Other substantive exploration data Other exploration data, if meaningful and material, should be reported including (but not limited to): geological observations; geophysical survey results; geochemical survey results; bulk samples – size and method of treatment; metallurgical test results; bulk density, groundwater, geotechnical and rock characteristics; potential deleterious or contaminating substances. Additional exploration data used to assist and validate interpretations at Costerfield include the use of surface geological mapping and a 2D seismic line. Bulk density work using the immersion methodology was completed in 2021 on similar lode and waste material at the Costerfield deposit.

A regression formula is used for the BD of lode material:

Augusta, Cuffley, Brunswick Lodes:

     BD= ((1.3951*Sb%)+(100-(1.3951*Sb%)))/(((1.3951*Sb%)/4.56)+((100-(1.3951*Sb%))/2.74))

where the host rock BD is 2.74 g/cm3

Youle/Shepherd/True Blue:

If (Sb%>1) BD=((1.3951 × Sb%)+(100-(1.3951 × Sb%)))/(((1.3951 × Sb%)/4.56)+((100-(1.3951 × Sb%))/2.69) )If (Sb%<1) BD= (0.05661 × Fe%) + 2.5259where:Empirical formula of stibnite: Sb2S3.Sb%: Antimony assay as a percentage by mass.Molecular weight of antimony (Sb): 121.757.Molecular weight of sulfur: (S): 32.066.1.3951 is a constant calculated by 339.712/243.514 where 339.712 is the molar mass of Sb2S3, and 243.514 is the molar mass of antimony contained in one mole of pure stibnite.BD of pure stibnite: 4.56.BD of unmineralised gangue: 2.69, representing a ratio of 1:3 siltstone to quartz.Fe%: Iron assay as a percentage by mass. The host rock BD of waste rock is 2.76 g/cm3.There are no material occurrences of deleterious elements.

Further work The nature and scale of planned further work (e.g. tests for lateral extensions or depth extensions or large-scale step-out drilling).Diagrams clearly highlighting the areas of possible extensions, including the main geological interpretations and future drilling areas, provided this information is not commercially sensitive. The Exploration Results reported in this document refer to areas of the Costerfield Property already in production as well as potential future production areas. Future exploration will be focused on advancing these areas through to an Indicated Resource, if drilling is successful. In addition, exploration will be conducted on the margin of currently operating areas to increase mine life where possible. CONTACT:  NIC EARNER, MANAGING DIRECTOR & CEO, ALKANE RESOURCES LTD, TEL +61 8 9227 5677
INVESTORS & MEDIA:  NATALIE CHAPMAN, CORPORATE COMMUNICATIONS MANAGER, TEL +61 418 642 556

Photos accompanying this announcement are available at:
https://www.globenewswire.com/NewsRoom/AttachmentNg/adb351a0-caf6-4777-bc9e-cb50336502b7
https://www.globenewswire.com/NewsRoom/AttachmentNg/1d81658e-65cf-4048-8608-4b95b4fedaaf
https://www.globenewswire.com/NewsRoom/AttachmentNg/085c1cf1-f271-4654-90f4-76ed3d2a06f4
https://www.globenewswire.com/NewsRoom/AttachmentNg/69045f50-831c-4367-bbf8-42141610be38
https://www.globenewswire.com/NewsRoom/AttachmentNg/a59f7d32-1bde-4a89-a114-1a8263eceea0
https://www.globenewswire.com/NewsRoom/AttachmentNg/7fcf21c8-cc10-4992-bcdc-5d26fa4757e3
https://www.globenewswire.com/NewsRoom/AttachmentNg/91fe17a8-2985-440c-9397-ecc378e4ff17
https://www.globenewswire.com/NewsRoom/AttachmentNg/9c94c948-755f-4ea5-83b5-f10a647effe7
https://www.globenewswire.com/NewsRoom/AttachmentNg/8c7f1cb1-e0aa-45c4-99a9-810efa4f926a

UPS (UPS -0.56%) spent 18 months deliberately shrinking its relationship with its biggest customer, and on July 28 it declared the job finished. In the earnings release, CEO Carol Tomé thanked employees for having "successfully completed our Amazon glide down and related network reconfiguration initiatives as designed."

The scale of what ended is massive. On the earnings call, Tomé said UPS had eliminated about 2 million pieces per day of what she called lower-quality Amazon (AMZN -0.57%) volume, removing roughly $4.5 billion of related expenses along the way.

For Amazon shareholders, the story runs the other way. Two million packages a day stopped moving through UPS trucks. Who's moving them now, and at what cost?

Image source: Amazon.

UPS got what it wantedThe carrier's results say a lot about the volume it gave up. UPS's second-quarter U.S. domestic revenue rose 6% year over year on a 9.3% increase in revenue per piece -- more money on fewer packages. The segment's non-GAAP (adjusted) operating margin expanded to 8%, up a full percentage point from a year earlier. And the company raised its full-year revenue outlook to about $91.2 billion.

And Tomé told analysts that excluding Amazon and the volume UPS intentionally handed to the market, its volume grew in the second quarter.

Put another way, the packages UPS shed were the ones diluting its profitability. Residential e-commerce delivery is expensive relative to what shippers pay, and UPS's margin went up as less of it flowed through the network.

Today's Change

(

-0.56

%) $

-0.57

Current Price

$

102.01

The biggest parcel carrier is the shipperMost of that volume appears to have gone to Amazon itself. According to logistics data firm ShipMatrix, Amazon's delivery arm handled an estimated 6.7 billion U.S. parcels in 2025. The U.S. Postal Service handled 6.6 billion, UPS came in at 4.4 billion, and FedEx delivered 3.6 billion. That made Amazon the country's largest parcel carrier by volume.

The growth rates were just as lopsided. Amazon's volumes rose nearly 10% in 2025, while UPS and the Postal Service each shrank 8.6%, ShipMatrix found. FedEx was the only one of the three traditional carriers whose volume grew.

To be fair, Amazon hasn't said precisely how much of the departed UPS volume it absorbed itself. The company still hands packages to the Postal Service and other carriers for portions of the last mile, so some of the load simply moved between carriers.

But Amazon's own delivery network is already expanding fast. The company is spending more than $4 billion to triple its rural delivery footprint by the end of this year, growing that network to over 200 delivery stations reaching more than 13,000 ZIP codes -- capacity it says will handle over a billion additional packages a year. The rural build is one slice of the capital expenditures Amazon keeps pouring into its delivery network.

The contrast with the carriers is sharp. FedEx and UPS charged remote-delivery surcharges of about $15.50 and $15.35 per package last year, and Amazon is building density in the kinds of places its rivals charge extra to visit.

A $27.9 billion quarterly shipping billThat capacity isn't free, and Amazon's income statement shows where the cost lands. The company's worldwide shipping costs hit $27.9 billion in the second quarter, up 19% from $23.4 billion a year earlier. Across the first six months of 2026, shipping costs rose 17% to $53.6 billion, so the bill is growing faster as the year goes on.

Today's Change

(

-0.57

%) $

-1.48

Current Price

$

258.63

Compare that to what the shipping supports. Amazon's online-store sales grew 15% year over year in the second quarter, so the delivery bill is outpacing the revenue it serves.

Why carry the load anyway? Because for Amazon, delivery is part of the product. The company's own filings credit its sales growth partly to its "fast shipping offers," and speed is easier to guarantee on a network Amazon controls than on one it rents.

That, I think, is the right way to read the 2 million daily packages. Volume that was a margin problem for a carrier is, for Amazon, the cost of owning its promise to customers.

The trade-off is right there in the numbers: shipping costs rising 19% against 15% online-store growth. UPS is done with the volume. The cost of carrying it sits on Amazon's own network now, and it likely will for a long time.

16h38 ▪ 4 min read ▪ by Evans S.

Summarize this article with:

Donald Trump’s memecoin has just gained more than 80% in a few hours. TRUMP crossed the $3 mark before settling around $2.94. The move caught traders positioned short off guard: over $30 million worth of positions were liquidated within 24 hours. A rumor about a new crypto project from the Trump family was enough to start the momentum.

In brief TRUMP jumped more than 80% and briefly exceeded $3. More than $30 million worth of positions were liquidated in 24 hours. A rumor suggests a new Trump asset on Robinhood, without official confirmation. The TRUMP crypto returns to $3 The memecoin comes back after several difficult months. In July, nearly one million TRUMP wallets already showed a cumulative loss of $3.81 billion. The current rebound drastically changes the atmosphere. TRUMP went above $3 during Asian trading hours. The token then retreated to around $2.94.

Volume follows. More than $1.79 billion were traded over 24 hours, the highest level in three months. Yet, Trump crypto remains far from its all-time high. Its drop still exceeds 90% compared to levels reached after launch.

This time, the spark came from social networks. Rumors mentioned the launch of a new asset tied to the Trump family on Robinhood Chain. No official confirmation was made. Traders, however, did not wait.

Sellers get trapped Several traders had opened short positions on TRUMP. The reasoning seemed simple: a new Trump family crypto could divert some capital from the existing memecoin.

The market went in the opposite direction. On Binance alone, about $8.59 million of short positions were liquidated. Total liquidations around TRUMP exceed $30 million in 24 hours.

The short squeeze then fueled the rise. When a leveraged crypto trader is liquidated on a short position, the platform must buy the asset to close their position. The higher the price rises, the more shorts might be forced to close as well.

TRUMP had already shown this kind of volatility. In June, the memecoin jumped over 25% in one day. This time, the move goes much further. Most short positions below $3 may have already disappeared. Above that, much less liquidity waits until about $3.50. New sellers will probably think twice.

$172 million remain open The crypto market has not fully deflated after the squeeze. Open interest on TRUMP reaches about $172 million, its highest level since April. New long and short positions continue to appear.

However, the Robinhood rumor remains the main issue. No new Trump token has been officially announced. There is also no guarantee that TRUMP holders would receive any special benefit if another asset were launched.

WLFI also benefited from the move. The World Liberty Financial token briefly reached $0.07 before falling back to around $0.06. The Trump name is still capable of moving a lot of money in crypto. It also attracts criticism. In early August, the TRUMP memecoin got involved in negotiations around the CLARITY Act, raising new questions about conflicts of interest of the US president. Today, the market mainly looks at price. Over 80% rise. $30 million liquidated. And still no confirmation of the project that triggered the move.

Maximize your Cointribune experience with our "Read to Earn" program! For every article you read, earn points and access exclusive rewards. Sign up now and start earning benefits.

Join the program

A

A

Lien copié

Evans S.

Fascinated by Bitcoin since 2017, Evariste has continuously researched the subject. While his initial interest was in trading, he now actively seeks to understand all advances centered on cryptocurrencies. As an editor, he strives to consistently deliver high-quality work that reflects the state of the sector as a whole.

DISCLAIMER

The views, thoughts, and opinions expressed in this article belong solely to the author, and should not be taken as investment advice. Do your own research before taking any investment decisions.

21h58 ▪ 6 min read ▪ by Fenelon L.

Summarize this article with:

Justin Sun claims to have won a first round against World Liberty Financial. On August 20, a federal judge in California reportedly refused to send all of his claims to private arbitration. This decision comes just days after the conditional green light granted by the OCC to World Liberty Trust Company, which is set to resume issuing the USD1 stablecoin.

In brief According to Justin Sun, Judge James Donato refused to submit all his claims to private arbitration, maintaining his individual claims before the federal court. On August 14, the OCC granted preliminary conditional approval to World Liberty Trust Company, which plans to take over the issuance of USD1 and management of its reserves from BitGo. The fate of claims filed by Blue Anthem and Black Anthem remains to be determined. The written order regarding the August 20 hearing was not yet publicly available at the time of writing. Justin Sun obtains the maintenance of part of the case before the court The standoff between Justin Sun and World Liberty Financial has been going on for several months. The founder of Tron, who bought $45 million worth of WLFI tokens, brought the case before a federal court in San Francisco in April.

This dispute between WLFI and Justin Sun notably concerns the freezing of his tokens and the control powers that World Liberty would have integrated into the WLFI contract.

On June 2, World Liberty Financial asked Judge James Donato to force the plaintiffs to go through arbitration and to suspend the judicial procedure. The hearing on this request took place on August 20.

At its outcome, Sun stated that his personal claims would remain publicly reviewed before the federal court. In other words, World Liberty did not get the transfer of the entire dispute to a private procedure.

This distinction matters. Indeed, arbitration generally allows settling a conflict outside of public hearings and with much more limited visibility on exchanged documents.

However, nothing has yet been decided on the merits. Blue Anthem Limited and Black Anthem Limited, two companies also parties to the complaint, have their own claims. Their treatment remains under discussion. Importantly, the written order from the judge was not yet publicly available at the time of writing this article.

World Liberty prepares in parallel its bank for USD1 The timeline adds another dimension to the case. On August 14, six days before the hearing, the Office of the Comptroller of the Currency granted a preliminary conditional approval to World Liberty Trust Company.

The future national trust bank is to take charge of issuing and redeeming USD1, as well as managing its reserves. It plans to take over these activities from BitGo, which currently serves this role.

But World Liberty Trust cannot start its operations yet. Final authorization depends on meeting several conditions imposed by the OCC.

The bank must notably have at least $20 million in Tier 1 capital. It must also maintain sufficient liquid assets to cover 180 days of operational expenses and notify the regulator before any significant changes to its business model.

The OCC also retains the possibility to modify, suspend, or withdraw its preliminary approval before the official opening of the institution.

Another element in the document deserves attention: World Liberty Trust will neither be able to issue, hold, nor trade WLFI tokens. However, the OCC specifies that World Liberty Financial and the future bank indirectly share some owners.

On paper, a clear boundary is thus drawn between the WLFI token and the banking activities related to USD1.

Previous token freezes fuel questions This separation comes as the control powers exercised over certain assets linked to World Liberty are already at the heart of several conflicts.

In June, World Liberty Financial froze some on-chain addresses associated with HTX as part of a sanctions compliance review. The crypto platform challenged this decision and suspended several pairs involving WLFI and USD1. It also announced the conversion of its users’ USD1 holdings into USDT.

A few months earlier, in September 2025, Justin Sun himself saw a significant portion of his WLFI tokens blocked after movements to exchange platforms. This episode notably led to the current dispute.

The legal battle is not limited to California. World Liberty also sued Justin Sun for defamation and market manipulation in a separate proceeding.

For now, none of these proceedings have concluded on the merits of the accusations.

The next steps should bring more clarity. On one side, Judge Donato’s written order will specify which claims will remain before the court and which might still be subject to arbitration. On the other, World Liberty Trust must meet the OCC’s requirements before obtaining its final authorization.

Two separate cases, but the same underlying question: how far does the control exercised within the World Liberty ecosystem extend, as USD1 is about to enter a federally regulated banking framework reinforced by the American legislation on stablecoins?

Maximize your Cointribune experience with our "Read to Earn" program! For every article you read, earn points and access exclusive rewards. Sign up now and start earning benefits.

Join the program

A

A

Lien copié

Fenelon L.

Passionné par le Bitcoin, j'aime explorer les méandres de la blockchain et des cryptos et je partage mes découvertes avec la communauté. Mon rêve est de vivre dans un monde où la vie privée et la liberté financière sont garanties pour tous, et je crois fermement que Bitcoin est l'outil qui peut rendre cela possible.

DISCLAIMER

The views, thoughts, and opinions expressed in this article belong solely to the author, and should not be taken as investment advice. Do your own research before taking any investment decisions.

Michael Saylor, executive chairman of Strategy, reiterated his perspective on Bitcoin as “digital energy,” emphasizing its role as a scarce, transferable form of economic value. Saylor described Bitcoin’s capacity to securely convert economic value into a new format that is accessible to individuals, corporations, institutions, and even governments.

Bitcoin holdings push treasury above cost basisOn August 16, Strategy reported holding 840,447 BTC, reflecting approximately 4% of Bitcoin’s total capped supply of 21 million coins. The company disclosed an average acquisition cost of $75,385 per Bitcoin, including all associated fees and expenses.

With Bitcoin trading around $77,175 on August 23, the estimated market value of Strategy’s Bitcoin holdings reached $64.86 billion. This positions the company’s Bitcoin treasury roughly $1.5 billion above its recorded purchase cost of $63.36 billion. The unrealized gain remains subject to rapid market changes, as Bitcoin’s price continues to fluctuate.

Saylor sees Bitcoin’s fundamental breakthrough in its ability to move economic value through a decentralized network, introducing digital scarcity that no central authority can alter. He pointed out that this quality allows holders to safeguard and transfer value securely across various economic actors.

Bitcoin’s most profound breakthrough is the ability to convert economic energy into digital form and bind it securely to a person, family, company, machine, or nation.

Despite recording significant paper gains, these valuations do not equate to distributable corporate profits. The company’s balance sheet must also account for debt obligations, taxes, operating costs, and preferred dividend commitments. A decline in Bitcoin’s price below the average purchase cost would result in unrealized losses, reinforcing the volatility that remains inherent in Strategy’s treasury model.

Strategy has structured its capital platform under the Digital Credit brand, offering products such as STRC, STRF, STRK, and STRD preferred shares. These financial instruments trade like traditional securities on exchanges including Nasdaq. However, investors in these preferred shares do not have direct claims on any particular Bitcoin within Strategy’s holdings.

The company’s STRC offering consists of variable-rate perpetual preferred stock, with each unit carrying a $100 stated value. The board manages dividend declarations monthly, and the market price can deviate from the stated amount based on trading activity.

Between August 10 and August 16, Strategy repurchased 1.39 million STRC shares for approximately $132.2 million using proceeds from its recent MSTR common-share sales. During the same period, the company brought in $333.7 million by selling around 3.46 million MSTR shares. Of these funds, $52.4 million went to STRC dividends, while the remaining $149.1 million was added to its dollar reserve, which stood at $4.80 billion as of August 16.

Strategy reported no Bitcoin purchases or sales during this timeframe. Chief Executive Phong Le has connected further BTC accumulation to STRC’s value aligning with its $100 stated amount, although no timeline has been given for additional Bitcoin purchases. Management decisions in coming weeks may involve new MSTR share issuances, further preferred share buybacks, or expansion of the company’s Bitcoin position.

Strategy’s next regulatory filings will clarify how these capital management moves align with both Saylor’s digital energy thesis and liquidity objectives. They will also detail any changes to the $4.80 billion reserve and whether Bitcoin acquisitions resume.

Mini dictionary: Strategy is a US-based technology company that has become widely known for its significant Bitcoin holdings, positioning itself as a corporate BTC investment pioneer.

DateBTC HoldingsAvg. Cost per BTCMarket Price per BTCAcquisition CostMarket ValueUnrealized GainAug. 16840,447$75,385$77,175$63.36 billion$64.86 billion$1.50 billion Bitcoin digital energy remains a metaphor highlighting both opportunity and risk, as actual returns depend on market volatility, operating costs, and corporate obligations.

Disclaimer: The information contained in this article does not constitute investment advice. Investors should be aware that cryptocurrencies carry high volatility and therefore risk, and should conduct their own research.

Xaman Wallet, the self-custodial app built by XRPL Labs and the dominant wallet for the XRP Ledger ecosystem, has removed its 0.09 XRP minimum fee on trades. The stated goal is straightforward: less friction at the entry point means more people can actually use the network’s liquidity, including traders moving smaller amounts who were previously priced out by the floor.

It is a notable reversal for a wallet that only introduced its service fee structure in late January 2025, when it rolled out a 0.8% trading fee on swaps and decentralized exchange activity, paired with that 0.09 XRP minimum. For context, the XRP Ledger’s own base transaction cost sits at roughly 0.00001 XRP. Xaman’s minimum was, by comparison, orders of magnitude higher than the network itself charges.

Why the minimum fee mattered more than it looked Removing the floor does not eliminate Xaman’s service fees entirely. The 0.8% trading fee on swaps and DEX activity remains in place. What changes is the lower bound: trades that previously couldn’t make economic sense below a certain size now have a clearer path to execution without being penalized simply for being small.

Context: Xaman’s evolving fee structure Xaman, formerly known as XUMM, rebranded as part of a broader effort to expand its identity beyond a simple transaction signing tool into a full-featured financial interface for XRPL. The January 2025 fee introduction was the wallet’s first significant shift toward a revenue model that didn’t rely on its Pro subscription tier.

That Pro subscription was itself discontinued for new customers on January 30, 2026, a decision that closed off one monetization path and reinforced the platform’s reliance on transaction-based fees. Free payment options within the wallet were preserved.

The June 2026 launch of Xaman Swap added one-tap swaps directly inside the wallet interface. Dropping the minimum fee is, in that light, a logical complement to the Swap feature: make it easy to trade, then make sure small trades aren’t punished for being small.

Disclosure: This article was edited by Editorial Team. For more information on how we create and review content, see our Editorial Policy.

An artificial intelligence (AI) security firm went public with a Ledger Ethereum app bug. Ledger says it had already fixed the flaw quietly, two weeks earlier.

Chief technology officer Charles Guillemet called the disclosure fear-mongering. The patch shipped on August 12 with a one-line note and no security bulletin.

What the Ledger Ethereum App Bug Actually DidLedger sells one core promise. The screen shows you what you are signing. That promise has a name. Ledger calls it clear signing, and it turns raw transaction code into plain words on the device screen.

TestMachine says it found a way around that. The firm builds an AI agent called Azimuth that hunts exploits in smart contracts. On its own EVMBench benchmark, Azimuth catches 86.3% of known bugs with roughly 2.7% false positives.

Found by Azimuth during an autonomous scan of the Ledger Ethereum app. Validated on Flex. Shared and verified with the team. Declining any bounty. Same shared APDU/UI code across Nano X, Nano S Plus, Stax, Apex.

The power of always on securityhttps://t.co/fH5mO97Kkp

— TestMachine (@testmachine_ai) August 22, 2026
Here is the flaw in plain terms. A malicious website could send the device a second command while you were still reading the first one.

The channel between browser and device is called the Application Protocol Data Unit, or APDU. It kept listening during the review. So it accepted the swap.

You would read a small transfer on screen. Then you would tap approve. And you would actually sign an unlimited token approval to a stranger.

That last part is why this matters. Chainalysis has traced roughly $1 billion in crypto stolen through approval phishing since May 2021. Those victims signed the approvals themselves.

TestMachine says it confirmed the bug on a Ledger Flex. Ledger has sold more than 7 million devices across 180 countries.

Ledger’s Donjon Team Says It Got There FirstGuillemet flips the timeline. Donjon is Ledger’s in-house hacking team. He says it caught the bug with its own AI tools and shipped the fix first.

The public changelog backs the date. Version 1.22.2 landed on Aug. 12. Its entire security note says “Security issues.”

Donjon has published 22 numbered security bulletins. None of them covers this bug. The latest, dated June 4, deals with a Monero key-recovery issue instead.

That silence is the gap TestMachine walked into. Ledger closed the hole, then never told owners what it had closed.

Guillemet’s sharper complaint is about manners. He says TestMachine contacted the bounty program only after the patch shipped. It never spoke with the bounty team.

“…Then they published a thread implying the problem is unsolved. It is not. That’s not security research. That’s manufacturing fear for attention,” Charles Guillemet, Ledger CTO remarked.

Follow us on X to get the latest news as it happens

TestMachine praised the speed of the fix and turned down the reward. Ledger pays bounties in Bitcoin, at an amount it sets case by case.

AI Found the Bug Twice, But Humans Still FoughtBoth sides used machine learning to reach the same defect. That is the part worth watching.

Ledger has made this argument before. Its executives have said for months that AI attackers threaten wallets more than weak hardware does.

Guillemet drew his line at discipline.

“AI-speed research only makes the ecosystem safer if the people doing it still follow basic security principles. Disclose responsibly. Verify before you publish. Don’t confuse noise with a finding.”

The fight itself is familiar. Security firms have gone loud after hacking a Trezor device, and CertiK researchers fought Kraken over disclosure terms in 2024.

So is the flaw. Back in January 2021, Donjon disclosed that this same Ethereum app failed to show transaction data for unsupported assets. Same app, same lesson. What you saw was not what you signed.

AI now surfaces these bugs in hours. Vendors and researchers still coordinate at human speed. That gap is where this argument lives.

For owners, the fix is dull. Open Ledger Live, update the Ethereum app, and check that it reads 1.22.2.

Tether’s two Uruguay mining sites cost an estimated $120 million. A dispute with state utility UTE centered on how much electricity the sites could draw. UTE disconnected both facilities in July 2025 after bills went unpaid. Tether has continued pursuing Bitcoin mining elsewhere in South America. Tether’s attempt to establish Uruguay as a launchpad for a larger South American Bitcoin mining business ended with two abandoned facilities after a contractual dispute with state-owned electricity provider UTE. The sites, estimated by a person with direct knowledge to have cost about $120 million, lost power in July 2025, according to documents and interviews reviewed by Reuters, turning what began as a renewable-energy mining expansion into a costly example of how electricity contracts can determine the economics of industrial-scale Bitcoin production.

A Power Contract Became the Project’s Central Problem When Tether announced its Uruguay expansion in 2023, the country appeared well suited to the company’s mining ambitions. Uruguay offered political stability, a reliable electrical grid and a power system with substantial renewable generation.

The two facilities were built in the department of Florida, with a former contractor estimating investment of roughly $60 million in each site. Uruguay was intended to serve as a testing ground before Tether expanded further into countries including Brazil, Paraguay and Argentina.

Operations initially generated revenue, according to former contractors interviewed by Reuters. The problem emerged as electricity requirements increased.

Tether and UTE interpreted a critical part of their electricity agreement differently. Tether understood the contracted power figure as a minimum allocation that could eventually be expanded. UTE regarded it as the maximum amount of electricity available to Tether’s local entity, Microfin.

That distinction became increasingly important as the mines scaled. Bitcoin mining facilities need a large and predictable electricity supply because their economics depend heavily on keeping specialized machines operating continuously. According to Reuters, insufficient supply sometimes left the sites without enough electricity for days.

Key stages in the Uruguay project May 2023
Tether announces Bitcoin mining operations in Uruguay.

November 2024
The electricity supply dispute is documented by UTE.

May 2025
Microfin stops paying its electricity bills.

July 25, 2025
UTE disconnects electricity to the mining sites.

November 2025
Tether notifies authorities that operations will cease and most employees will be laid off.

Why Negotiations Failed Despite a Revised Electricity Deal The disagreement did not immediately end the project. UTE and Tether attempted to renegotiate the arrangement, and the utility’s board approved both a memorandum of understanding and revised contract documents.

The agreement was never completed.

According to minutes cited by Reuters, Tether representatives did not attend the planned signing. By that stage, Microfin had already stopped paying electricity bills and had informed UTE in June that it intended to terminate the contracts.

With the revised agreement unsigned and bills outstanding, UTE cut electricity to the facilities on July 25, 2025. Microfin subsequently settled its outstanding debt in December, UTE told Reuters.

The chronology matters because the shutdown was not simply the result of Uruguay lacking enough renewable electricity. The underlying problem was whether Tether could secure sufficient power under terms that made expanding its mining operation commercially workable.

Uruguay Exposed the Economics Behind Tether’s Mining Strategy Bitcoin miners effectively convert electricity and computing capacity into BTC. That makes the cost and reliability of power fundamental to profitability.

The April 2024 Bitcoin halving made that calculation harder by reducing the block subsidy from 6.25 BTC to 3.125 BTC. Unless higher Bitcoin prices, transaction fees or improved mining efficiency compensate for the reduction, miners earn fewer coins from the same amount of computational work.

Uruguay offered extensive renewable generation, but renewable electricity is not automatically cheap electricity.

Crypto mining specialist Nicolas Ribeiro told Reuters that Uruguay’s relatively high power costs make the country less competitive for Bitcoin mining, while its stable grid and internet infrastructure may be better suited to AI data centers. Mining operators have fewer reasons to remain in a particular jurisdiction when electricity becomes uneconomical because much of their computing hardware can be relocated.

That flexibility helps explain why the collapse of the Uruguay project has not ended Tether’s broader energy strategy.

Tether Has Already Shifted Mining Investment Elsewhere The failed project also needs to be viewed against the scale of Tether’s wider expansion.

CEO Paolo Ardoino said in June 2025 that Tether had invested more than $2 billion in energy and mining infrastructure across 15 sites in Uruguay, Paraguay and El Salvador, according to The Block. The company has pursued mining infrastructure as one component of a much broader investment portfolio.
Brazil has become one of the next areas of focus.

In July 2025, Tether and agricultural and renewable-energy producer Adecoagro signed a memorandum of understanding to explore using renewable electricity for Bitcoin mining. The proposed model centers on monetizing surplus electricity from Adecoagro’s renewable-energy portfolio rather than reproducing the same arrangement used in Uruguay.

Tether had previously acquired a controlling stake in Adecoagro, giving it a more direct connection to the underlying energy assets.

The distinction could prove significant. Greater control over energy generation may reduce some of the contractual dependency that affected the Uruguay sites, although it does not eliminate mining’s exposure to

Bitcoin prices, network difficulty, equipment efficiency and local electricity economics.

What Changes Next for Tether’s Bitcoin Mining Plans The Uruguay exit leaves Tether with a more fragmented South American mining strategy than originally envisioned. Instead of using one country as a regional testing ground and then replicating the model, the company is pursuing different energy arrangements across individual markets.

For the mining industry, the episode reinforces why headline renewable-energy capacity is not enough when evaluating potential jurisdictions. Large operators need contractual certainty over both electricity prices and how much power they can consume as facilities expand.

Tether’s next South American projects will therefore provide a clearer test of whether greater involvement in energy production can improve those economics. Its partnership with Adecoagro is particularly relevant because the companies plan to examine whether surplus renewable electricity can be converted into Bitcoin mining revenue, giving Tether a different operational model from the one that failed in Uruguay.

Bittensor (TAO) is maintaining its key support levels, as buyers continue to defend the long-term price structure. The project’s recent expansion to the Base network, facilitated by Chainlink‘s Cross-Chain Interoperability Protocol (CCIP), is bringing its decentralized AI technology closer to decentralized finance (DeFi) users and supporting growth in the onchain AI sector.

Bittensor price activity and growthAs of the latest data, TAO is trading at $232.32, with a 24-hour trading volume of $217.92 million and a market capitalization of $2.61 billion. The token’s price recorded a 5.78% gain in the past day, which analysts view as a sign of a potential bullish reversal supported by network expansion and resilient market structure.

Crypto analyst Etherealist noted that TAO consistently finds buyers near the $200 zone, even during extended periods of weak sentiment. This consistent support is seen as a key factor for a possible trend reversal. The analyst suggested that if TAO breaks past its current resistance level, further rallies toward $500 could become likely.

Analysts emphasized that while TAO often struggles to attract attention at current levels, a robust move toward $500 could change market sentiment, with $1,000 and even $1,500 becoming realistic targets if upward momentum continues.

ForeverMoney reported that Bittensor’s decentralized AI network now includes Base, connecting the project’s AI offerings such as compute, code, vision, and prediction with Layer-2 DeFi infrastructure and a rapidly expanding user base.

Deeper integration through Chainlink CCIPThe use of Chainlink CCIP enables secure and reliable cross-chain communication between Bittensor and Base. This integration allows Bittensor’s decentralized AI features to be more accessible for DeFi users and traders active on Base, broadening the project’s reach within the crypto ecosystem.

By bridging Bittensor and Base, Chainlink CCIP allows for transactions and data flow between the networks, enhancing interoperability between AI and DeFi sectors. The collaboration signals a growing intersection of blockchain interoperability, artificial intelligence, and decentralized finance.

Looking at market dynamics, traders are reminded that, in a landscape where a single Federal Reserve policy move or an unexpected altcoin listing can rapidly shift valuations, switching between multiple apps for price charts, news, and portfolio management creates missed opportunities. Increasingly, investors are using all-in-one platforms like CryptoAppsy, which offer privacy-first access to real-time charts, smart price alerts, coin-specific news, and macro data on a single screen, often without the need for an account.

Support levels and future outlookTAO’s key support remains at $200. If this level holds and the token breaks above the $500 resistance, the path toward higher targets, such as $1,000 and potentially $1,500, may open. However, falling below $200 would likely challenge any bullish reversal and signal renewed downward pressure.

Further integration with Base and continued network growth could strengthen adoption within the DeFi and AI ecosystems.

The collaboration between Bittensor, Base, and Chainlink CCIP could accelerate onchain AI use case development and enhance cross-chain connectivity, potentially supporting long-term value for the TAO token.

While current indicators point to a possible bullish trajectory, analysts caution that future rallies depend on sustained demand and broader market conditions. Investors are urged to remain vigilant in a volatile environment.

Disclaimer: The information contained in this article does not constitute investment advice. Investors should be aware that cryptocurrencies carry high volatility and therefore risk, and should conduct their own research.

Given nothing more than the headline number, it would be easy to believe electric vehicle maker Tesla (TSLA +5.14%) is firing on all (proverbial) cylinders...at least within the United States. Although down slightly from the first quarter's 54.2% share of the U.S. electric vehicle market, Cox Automotive reported that the iconic EV brand accounted for 50.5% of the country's second-quarter EV sales -- as measured in units -- holding onto an industrywide majority reclaimed in the final quarter of last year for the first time since 2023.

Today's Change

(

5.14

%) $

17.73

Current Price

$

362.86

Now read the fine print. Tesla is only enjoying a market share advantage because its domestic rivals are suffering bigger EV sales setbacks than Tesla did. Total electric vehicle sales in the U.S. fell 20% during the second quarter, whereas Tesla's total unit sales fell 13% from 143,535 automobiles in the second quarter of last year to 124,800 units in Q2 of this year.

Losing share in other markets The United States isn't Tesla's only market. Europe and China are key electric vehicle markets as well, and the company's worldwide second-quarter total deliveries improved 25% year over year, to 480,126 automobiles.

Image source: Getty Images.

Even so, Tesla is losing market share in both of those markets, largely to China's BYD, but also to Chinese EV manufacturers Geely and Changan in China, and Volkswagen in Europe.

Of course, electric vehicles could soon be a secondary business for Tesla anyway. The company continues developing AI-powered humanoid robots that CEO Elon Musk has suggested could begin commercial production before the end of next year.

James Brumley has no position in any of the stocks mentioned. The Motley Fool has positions in and recommends Tesla. The Motley Fool recommends BYD Company. The Motley Fool has a disclosure policy.

Late last month, Tilray Brands (TLRY +3.64%) released its latest fiscal results and guidance updates. The market reacted positively to both, resulting in a modest post-earnings rally.

Since then, however, the bull run for one of the most-followed marijuana stocks has run its course. This is especially interesting, given that the U.S. legalization catalyst seems to be strengthening at the same time. Still, considering several factors, it is not surprising that investors appear hesitant to bid up Tilray shares.

Image source: Getty Images.

Tilray's earnings were not much of a game changer Take a look at Tilray's latest quarterly financials, released on July 28, and you'd think that the Canada-based cannabis company had turned a corner. In the earnings release, management touted the company's "record revenue and adjusted EBITDA" and provided promising guidance for the coming fiscal year.

Today's Change

(

3.64

%) $

0.17

Current Price

$

4.84

Yes, last fiscal year, revenue increased by 11%, to around $915 million, signaling that Tilray's getting close to hitting its $1 billion annual revenue target. Adjusted earnings before interest, taxes, depreciation, and amortization (EBITDA) increased 11%, to $61.1 million. Adjusted net income, rising from $6.5 million to $12.2 million, nearly doubled as well. Even so, adjusted earnings fell short of sell-side forecasts. Worse yet, on a GAAP basis, Tilray once again reported heavy losses, with net losses attributable to Tilray shareholders totaling $49.6 million, or negative 43 cents per share.

Other factors keep investors hesitant about the stock For fiscal year 2027, Tilray's management expects adjusted EBITDA of $68 million to $75 million, yet it's unclear whether this will translate into a swing to positive GAAP earnings. Management may also be touting how it's cut Tilray's debt to effectively zero, but it's doing so in a dilutive manner: through debt-for-equity swaps.

Even as the U.S. federal government's marijuana rescheduling efforts continue, Tilray has relatively limited exposure to this catalyst. Now diversified into areas such as alcoholic beverages and pharmaceutical distribution, cannabis accounts for just 29% of overall sales. Barring an end to share dilution, a significant improvement in results next quarter, or a big pivot back toward recreational cannabis, ho-hum price action will likely persist.

Cover image via U.Today Disclaimer: The opinions expressed by our writers are their own and do not represent the views of U.Today. The financial and market information provided on U.Today is intended for informational purposes only. U.Today is not liable for any financial losses incurred while trading cryptocurrencies. Conduct your own research by contacting financial experts before making any investment decisions. We believe that all content is accurate as of the date of publication, but certain offers mentioned may no longer be available.

Solana's RWA value has surpassed $4 billion, a new ATH for the network. The increase comes as the total supply of tokenized equities on Solana continues to grow.

This growth remains significant as what began as a negligible market in early 2024 has become a multibillion-dollar ecosystem spanning tokenized Treasuries, public equities, private credit, reinsurance, sovereign debt, commodities, liquidity funds, and stablecoin settlement infrastructure.

Tokenized equities are where Solana's RWA growth is accelerating fastest. As of late July 2026, 97% of all onchain tokenized equities spot volume to date had settled on Solana. This reflects Solana's increasing role as a leading venue for tokenized equity distribution and secondary activity.

RWA value doublesOne of the defining industry trends over the past year has been the growth of RWAs, with Solana sustaining its momentum. The recent increase of RWA value to $4 billion marks a nearly 100% increase from that seen at Q1 close. Solana non-stablecoin RWA value doubled, rising from $2.01 billion at the end of Q1 2026 to a current $4 billion.

HOT Stories

You Might Also Like

Solana currently has 347,944 RWA holders according to rwa.xyz data. The network has become a significant venue for tokenized Treasury products, with several large global issuers already on it.

These include Circle's USYC tokenized money market fund, one of the largest tokenized Treasury products worldwide, which launched on Solana in October 2025.

Others include BlackRock's BUIDL, the USD Institutional Digital Liquidity Fund, which expanded to Solana in March 2025 through Securitize.

Ondo Finance runs two Treasury-linked products on Solana. USDY is a tokenized note backed by short-term U.S. Treasuries and bank demand deposits, and OUSG provides exposure to short-term U.S. government bonds and is backed significantly by BlackRock's BUIDL fund.

VanEck's VBILL, a short-term U.S. Treasury product, launched on Solana in May 2025 via Securitize. Franklin Templeton's BENJI, the token of the Franklin OnChain U.S. Government Money Fund (FOBXX), the first U.S.-registered mutual fund to use a public blockchain as its system of record, launched on Solana in February 2025.

TLDR: The proposal could cut projected six-year SOL supply by 18.9M tokens, equal to about $1.81B at $95.70. Modeled staking yields could fall from 5.84% now to 4.34% after one year and 2.25% after three years. SGP-0002 needs one-third stake participation and two-thirds support, making it an early governance test. A 30% disinflation rate could bring Solana to its 1.5% inflation floor in 2029 instead of around 2032. Solana has opened an on-chain vote on SGP-0002, putting a token issuance change before validators and stakers. The proposal would double annual disinflation from 15% to 30%, accelerating how quickly new SOL issuance declines without changing the network’s 1.5% inflation floor.

⚡️LATEST: Solana’s proposal to double its disinflation rate from 15% to 30% is now live for governance voting.

If approved, $SOL inflation would fall twice as fast, meaning fewer new tokens entering circulation and less dilution for holders. pic.twitter.com/pPTt5iXQ5p

— CryptosRus (@CryptosR_Us) August 23, 2026

The vote is active under the network’s new stake-weighted governance framework, where validators and native stakers signal support or opposition. Delegators can override their validator’s choice, giving stakers a direct role in the outcome. The decision therefore combines monetary policy with an early governance test.

Solana Vote Puts 18.9M SOL Issuance Cut Before Stakers SGP-0002 is linked to SIMD-0550, authored by Lostin and 0xIchigo of Helius. Their June model placed the inflation rate at 3.82% under the existing schedule. At the current 15% annual disinflation rate, inflation would fall to about 3.24% after one year, reaching the 1.5% floor around the first half of 2032.

Source: X

However, the proposed 30% schedule would move faster. Inflation would decline to roughly 2.86% after one year and reach the same floor around early 2029. That acceleration would reduce cumulative issuance. The model projects total supply of 708.54 million SOL after six years, versus 727.43 million under the current schedule.

The 18.9 million-token difference equals about 2.6% of projected supply. At $95.70 per SOL on Aug. 23, that amount is worth about $1.81 billion. Nonetheless, the proposal does not immediately halve inflation. It only doubles the pace at which the inflation rate declines toward the unchanged long-term floor.

Lower issuance would also reduce staking rewards. At 68% modeled staking participation, nominal yield would fall from about 5.84% currently to 4.34% after one year. The model then places staking yield near 3.00% after two years and 2.25% after three years. Those figures exclude commissions, MEV, and block-related revenue.

Validator economics also weaken gradually. Among 738 validators, two additional operators become unprofitable or move from breakeven after one year. That number rises to 13 after two years and 30 after three years. Still, the authors modeled the overall validator impact as relatively limited.

Lower Staking Yields Put Validator Economics in Focus The economics have already drawn institutional opposition. Nasdaq-listed Solana Company said Aug. 21 that it would vote against SGP-0002. The company supports lower issuance as a long-term goal. However, it argued that changing established economics during the first governance cycle could reduce institutional predictability.

The vote also follows the failed SIMD-228 debate in 2025. That proposal sought dynamic issuance tied to staking participation rather than the existing fixed schedule. About 74% of staked SOL participated. However, only 61.4% of non-abstaining votes supported the proposal, below the required two-thirds threshold.

Under the new SGP process, at least one-third of network stake must participate. Two-thirds of participating stake must then vote in favor for passage. SGP-0002 therefore places two questions before the network. One concerns how quickly SOL dilution should decline, while the other tests whether governance can produce decisive consensus.

Axsome Therapeutics (AXSM -0.13%), a biotech company, has performed fairly well this year. The company's shares are up 16% versus the S&P 500's 11%. Could the stock rise even higher? Wall Street certainly thinks so. Axsome Therapeutics' average price target (according to Yahoo! Finance) of $285.42 implies about a 37% upside from its current level. Let's find out why Wall Street is bullish on the stock and whether it's time to buy it.

Image source: The Motley Fool.

Axsome Therapeutics' recent financial results Axsome Therapeutics has made significant clinical and regulatory progress over the past few years. The company earned approval for Auvelity as a treatment for depression, and more recently won a label expansion for the medicine in treating Alzheimer's disease (AD) agitation. Axsome Therapeutics' approved portfolio also includes Symbravo, a migraine therapy, and Sunosi, which treats daytime sleepiness due to narcolepsy. Axsome Therapeutics' sales are growing rapidly. In the second quarter, the company's revenue increased by 46% year over year to $218.4 million. Auvelity is the main growth driver.

It contributed $180.3 million in net product sales, up 51% compared to the year-ago period. The good news is that Auvelity's indication in AD agitation is still brand-new. The U.S. Food and Drug Administration granted it in April. So, as the medicine gains significant traction in this market, it will help boost its sales. And there is a large addressable opportunity here. AD agitation affects more than five million patients in the U.S., and treatment options are limited.

So, Auvelity could fill an unmet need. Eventually, it could also earn another label expansion in smoking cessation, since Axsome is gearing up to start a phase 2/3 study in that indication. Axsome Therapeutics' Sunosi could also receive label expansions. It is currently being investigated in phase 3 studies as a potential treatment for depression, ADHD, and binge eating disorder. If it can land these additional approvals, it would become more successful and help improve Axsome Therapeutics' financial results.

Today's Change

(

-0.13

%) $

-0.26

Current Price

$

207.77

Why the future looks bright Axsome Therapeutics' pipeline doesn't end with Auvelity and Sunosi. In fact, the company boasts a rich lineup of mid and late-stage candidates that could transform its approved portfolio over the next few years. Here are some of these products. One of them is AXS-12, an investigational medicine for cataplexy in narcolepsy. This candidate has already cleared phase 3 studies and is currently under review by U.S. regulators.

Axsome Therapeutics estimates peak sales for AXS-12 between $500 million and $1 billion in narcolepsy. Then there is AXS-14, an investigational therapy for fibromyalgia. It is currently undergoing late-stage studies, as is the biotech's AXS-20, which is being developed to treat schizophrenia. Between these and other products in development, Axsome Therapeutics estimates that it could generate peak sales above $16 billion.

The company won't get there in a year or two. But considering its current market cap of roughly $11 billion, Axsome Therapeutics may be an attractive stock to buy even if it can land $16 billion in annual revenue by 2040. Here is why. Assuming a price-to-sales ratio of 3 by 2040 -- which is a reasonable assumption in the biotech industry -- if Axsome Therapeutics generates $16 billion in revenue by then, the company's market cap will be $48 billion. From its current level of $11 billion, it would grow at a compound annual rate of 11.1% over the next 14 years, which is a pretty strong return.

Of course, there are risks. Axsome Therapeutics could fail to launch one or more of its current pipeline projects due to clinical or regulatory setbacks. It could also face stronger competition in markets where it currently performs well, such as depression treatments, where several pharmaceutical giants dominate. However, the $16 billion projection only takes into account the company's more advanced programs, not several earlier projects that may also contribute to the top-line down the road.

So, even with the risk of clinical setbacks, Axsome Therapeutics has a large enough pipeline to generate competitive returns over the long run. Wall Street is right to be bullish, and investors should seriously consider buying this healthcare stock.

The Dutch Data Protection Authority is fining Uber €825 million (around $966 million) — the second largest penalty issued so far under Europe’s General Data Protection Regulation, according to Reuters.

The Dutch regulator was investigating complaints that Uber had deactivated driver accounts through an automated process without sufficient warning or human oversight. In a statement, deputy chair Monique Verdier said that the company had “committed serious infringements.”

“A computer should not make decisions on its own that have [such] major consequences,” Verdier said.

Uber, however, argued that most driver suspensions are brief, that no permanent deactivations take place without human review, and that drivers have the ability to appeal. (Dutch regulators said some drivers were permanently deactivated without human review, which Uber disputes.) The company said it will appeal the decision.

“We strongly disagree with this decision ​and disproportionate fine,” an Uber spokesperson told Reuters. TechCrunch has reached out to the company for additional comment.

Brahim Ben Ali, a former Uber driver in France, told the Dutch newspaper de Volkskrant that after his account was deactivated in 2019, he collected testimonies from 170 other Uber drivers and eventually brought his complaint to the Netherlands, where Uber’s European headquarters are located.

Ben Ali was assisted in this effort by a Swiss nonprofit focused on digital rights called PersonalData.io, which helped the drivers collect data about how the deactivation decisions were made. Founder Paul-Olivier Dehaye said a driver “can complete a thousand journeys with satisfied passengers, but if just one person reports a very serious problem, the consequences can be enormous.”

Dehaye told me that this is the third fine that the Dutch regulator has levied on Uber, following a €290 million fine over its handling of drivers’ personal data and a €10 million fine stemming from related issues. He also said he plans to start a class action suit through which drivers can seek compensation.

In fact, Dehaye said these fines all originate with complaints made by the same group of drivers. And he’s starting a new company called StartClaims to support the litigation and other regulatory action — first against Uber and then eventually expanding to other gig economy cases, as well as related areas like adtech.

While discussing the case with Dehaye (who I’ve known casually since college), I brought up a blog post by Daring Fireball’s John Gruber, in which Gruber worried that this fine makes it “unlawful in the EU for Uber to monitor its drivers for pulling scams against customers, or just never picking riders up, leaving them stranded.”

Gruber also took issue with Verdier’s statement, arguing, “Saying that ‘a computer’ made these decisions is like saying that when a company suspends or fires a habitually late employee, that ‘the time clock’ made the decision. Managers at the company set the policies, and the devices measure employee compliance.”

Dehaye countered that Gruber “misses the point.”

“Uber is free to use humans to punish drivers who scam, but then [it] has to take responsibility for this decision making (like ‘being an employer’, not ‘being a marketplace’),” he said.

When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.

Anthony Ha is TechCrunch’s weekend editor. Previously, he worked as a tech reporter at Adweek, a senior editor at VentureBeat, a local government reporter at the Hollister Free Lance, and vice president of content at a VC firm. He lives in New York City.

You can contact or verify outreach from Anthony by emailing [email protected].

This post may contain links from our sponsors and affiliates, and Flywheel Publishing may receive compensation for actions taken through them.

Alibaba (NYSE:BABA | BABA Price Prediction) just made its capital allocation priorities unmistakable. In the June 2026 quarter, the company repurchased 13.4 million ordinary shares (approximately 1.7 million ADSs) for US$162 million. A year earlier, in the same fiscal quarter, it bought back 56 million ordinary shares (7 million ADSs) for US$815 million. That is roughly an 80% cut in ADS repurchases at a company that still had US$19.3 billion of authorization remaining as of June 30, 2025.

The cash was rerouted into silicon and concrete.

Where the Buyback Money Went Capital expenditures rose 75% year over year to RMB 67,678 million for AI infrastructure. Free cash flow deteriorated to negative RMB 44,670 million from negative RMB 18,815 million a year earlier. The newly disclosed AI Labs and Applications segment posted an adjusted EBITA loss of RMB 13,861 million, up from RMB 3,224 million. On top of that, the quarter absorbed a EUR 550 million European Commission fine and RMB 4,458 million goodwill impairment.

Internal cash is not covering the buildout alone. During fiscal 2026, Alibaba raised approximately US$3.2 billion in convertible notes and HK$12 billion in exchangeable bonds to fund cloud and international commerce, and total debt to adjusted EBITDA doubled to 2.29x. Full-year FY26 repurchases came in at just US$1.046 billion, a fraction of prior years.

Management Frames It as an ROIC Bet CEO Eddie Wu was direct about the shift. “AI has become Alibaba’s most certain growth engine,” he told analysts on the August 20 call. CFO Toby Xu argued the math works: “Our AI plus cloud investment has a clear path to attractive ROIC.” Management said AI hardware typically reaches break-even within three years on a five-year useful life, with AI compute supply expected to remain constrained industry-wide until at least 2030. That constraint is the whole reason the power, cooling, and networking suppliers behind the buildout keep drawing capital, a group we profiled in a free report on seven AI infrastructure names that aren’t chipmakers.

The revenue side supports the case. AI Cloud and Compute Services revenue grew 45%, and AI-related product revenue posted triple-digit growth for the twelfth consecutive quarter. Cloud external growth hit a 22-quarter high, and MaaS annual run rate surpassed RMB 16 billion as of August, tracking a year-end target above RMB 30 billion.

What Investors Should Watch Next Shares closed at $130.53 on August 20, up 1.26% on the day and 6.88% over the past week, though still down 10.09% year to date. The setup is straightforward: if Qwen monetization and Zhenwu chip deployments compound as guided, the buyback cut looks like disciplined reinvestment. If AI Labs losses keep widening past RMB 13,861 million without matching cloud margin expansion, the balance sheet, now carrying US$46.5 billion in net cash, becomes the shock absorber. Wu made the trade-off explicit: “It’s only possible to monetize when you have that compute capacity in place.”

Contact [email protected] for any questions or corrections.

Dividends are great, but what's even better for long-term investors is knowing that they're holding shares of a company that's a true dividend stock, not just a stock that pays a dividend.

Companies become true dividend names by showing unwavering commitment to steadily increasing their payouts. One of the world's largest oil companies, ExxonMobil (XOM -0.63%), is certainly in that camp. ExxonMobil is on a 43-year run of increasing its payout. Those are increases shareholders can set their clocks by, and for those wondering, pencil in the energy stock's next dividend lift. It's likely to arrive in October, as it has over the past several years.

Image source: Getty Images.

Each of the company's 2023 through 2025 increases was $0.04 per share quarterly. That's not much, but those boosts add up over time. That consistency may have some on Wall Street banking on another increase of $0.03 to $0.04 a share, but ExxonMobil can deliver an "October surprise" -- and a positive one at that.

ExxonMobil can enhance dividend excellence In addition to the 43-year payout increase streak, ExxonMobil is the second-largest dividend payer in the S&P 500. Fortunately, a yield of 2.5% and a payout ratio of 52.5% imply two pivotal factors. First, the energy company isn't burdened by its dividend obligations. Second, there's room for payout growth.

How much growth? That's the $64,000 question, but there are credible reasons ExxonMobil could deliver a larger-than-expected dividend increase later this year. As the company noted last December, it was on pace to buy back $20 billion of its shares in 2025 and expected to maintain a similar cadence this year. Retire $40 billion worth of stock over two years, and any company's dividend tab will decline, making it easier to juice payouts to the upside.

ExxonMobil's status as an oil dividend stock royalty is further supported by cold, hard cash. Under its 2030 plan, the oil behemoth raised its 2024 to 2030 earnings and cash flow growth targets to $25 billion and $35 billion, respectively.

Today's Change

(

-0.63

%) $

-1.04

Current Price

$

165.11

Perhaps shortening the odds of a dividend surprise is ExxonMobil's expectation of $145 billion in "surplus cash flow" through 2030. That's based on $65-per-barrel Brent crude prices. Brent closed at nearly $89 on Aug. 20. If that oil contract remains elevated into the fourth quarter, it's possible (not promised) that ExxonMobil could put a little something extra in dividend investors' Halloween goody bags (the dividend increase is often announced around that holiday).

Competitive considerations Corporations are always competing with each other, but the competition isn't limited to business and generating sales. It extends to captivating investors' attention and their dollars. This is particularly true with dividend investors, and ExxonMobil likely knows as much.

These days, there's plenty of competition. Bond yields are high, and a slew of energy companies sport dividend yields well beyond ExxonMobil's 2.5%. Some of those companies boost payouts several times a year.

So while ExxonMobil's yield is more than double that of the S&P 500, that's not saying much, and the energy company may not want to rest on those "laurels." Amid stiff competition for dividend investors' capital, it might be prudent for ExxonMobil to go the extra mile with its next payout increase.

Advanced Micro Devices (AMD +0.81%) plans to invest more than $10 billion in Taiwan. But this does not mean that Taiwan Semiconductor Manufacturing (TSM +0.71%) will be the only beneficiary.

Instead, the money will go across Taiwan's broader semiconductor ecosystem, including advanced packaging, chip substrates (the base materials used in advanced chip packaging), and manufacturing capacity for complete artificial intelligence (AI) systems. These investments are expected to run through 2029 and help its partners scale production of next-generation products such as its Helios AI racks.

Hence, CEO Lisa Su is investing now to ensure AMD can manufacture enough hardware if its rapidly growing AI demand translates into large-scale deployments.

Image source: Getty Images

AMD's next AI bottleneck may not be the GPU AMD's new Venice EPYC server CPU is already ramping production using TSMC's advanced 2-nanometer process technology. The company also uses TSMC's SoIC-X and CoWoS-L advanced packaging technologies for some of its AI and data center chips.

The chipmaker is also expanding its supplier ecosystem beyond TSMC. The company is developing next-generation Elevated Fanout Bridge (EFB) chip packaging with ASE Technology and Siliconware Precision Industries. The company has completed testing of a panel-based version of its EFB packaging technology with Powertech Technology. AMD is also working with Taiwanese substrate suppliers and manufacturers that will help produce Helios AI systems at high volume.

The extra capacity could be critical. In July 2026, TSMC CEO C.C. Wei claimed that tight advanced-packaging capacity was limiting customers' growth. AMD could therefore win AI customers but still miss sales if it cannot package and assemble enough chips.

Need to expand manufacturing capacity Data Center revenue reached increased 107% year-over-year to $6.7 billion in the second quarter. This business accounted for about 58% of AMD's total revenue. Management expects Data Center revenue to grow at a compound annual growth rate (CAGR) of above 60% over the next three to five years, including CAGR of more than 80% for data center AI.

AMD will need a much larger supply chain if it comes close to those growth targets. A Helios AI rack contains 72 Instinct MI455X GPUs and 18 Venice CPUs. Meanwhile, OpenAI, Meta Platforms, and Anthropic have announced AMD deployments that could total as much as 14 gigawatts. However, those deployments will occur over several years.

Today's Change

(

0.81

%) $

3.80

Current Price

$

473.25

Being fabless no longer means being capital-light AMD purchased only $1.2 billion of property and equipment during the first half of 2026. Yet, the company exited the second quarter with $30.3 billion of broader unconditional commitments, primarily covering wafers, substrates, components, cloud capacity, software, and technology licenses. AMD also recorded a roughly $1 billion increase in prepaid expenses and other assets, mainly due to advance payments under supply agreements in the first half of 2026.

Hence, while AMD does not operate chip manufacturing factories, it still needs to commit significant capital to secure supply.

The added manufacturing capacity will matter only if it leads to profitable AI growth. AMD's non-GAAP operating margin was 27% in the second quarter, significantly lower than the management's target of more than 35% over the next three to five years. CEO Lisa Su is spending billions to make sure AMD can produce enough AI hardware if customer demand grows as expected.

Oracle (ORCL +3.10%) stock has plummeted by 56% from last year's record high, but I'm not convinced this is a good buying opportunity. Although the company operates some of the world's best data centers for processing artificial intelligence (AI) workloads, investors are concerned about its substantial debts, especially because some of its top customers could struggle to fulfill their financial obligations over the next few years.

Oracle will have an opportunity to ease some of those jitters in early September when it releases its financial results for its fiscal 2027 first quarter (ending Aug. 31), but here's why the report -- expected on Sept. 8 -- probably won't turn sentiment around.

Image source: The Motley Fool.

There is a problem with Oracle's $638 billion order backlog Oracle has a diverse business spanning enterprise software, database systems, AI infrastructure, and more. All eyes are on the cloud infrastructure segment right now, which is where the company logs the revenue it earns from renting data center computing capacity to AI customers.

Oracle's data centers are filled with thousands of advanced chips from suppliers like Nvidia and Advanced Micro Devices, connected by proprietary random direct memory access networking (RDMA) technology that moves information between components faster than traditional Ethernet networks. Moreover, Oracle's infrastructure is highly automated by software, so the company can bring new locations online faster than competitors that rely on human-led processes.

Those features combine to provide AI developers with fast processing speeds at an affordable price, which is why companies like OpenAI, Elon Musk's xAI, and Meta Platforms are lining up to use Oracle's infrastructure.

Today's Change

(

3.10

%) $

4.40

Current Price

$

146.47

The company generated $19.2 billion in total revenue during its fiscal 2026 fourth quarter (ended May 31), a 21% increase from the year-ago period. Cloud infrastructure accounted for $5.8 billion of that revenue, and it grew at a significantly faster pace of 93%. But Oracle's remaining performance obligations (RPO) were the headline number in the fourth-quarter report, soaring by 363% to a record $638 billion.

RPO reflects the value of signed contracts for services that haven't been delivered yet, so it's like an order backlog, and it's often used as an indicator of future revenue. Most of the $638 billion is from AI customers waiting for more data center infrastructure to come online -- but therein lies a big problem. According to a report by The Wall Street Journal from last September, around $300 billion of Oracle's RPO was from OpenAI alone.

However, OpenAI currently has just $40 billion in annualized revenue and is losing truckloads of money, so there is a serious question mark over the start-up's ability to fulfill its financial commitment to Oracle over the next few years. To make matters worse, OpenAI has made similar commitments to other cloud providers like Microsoft, making it even more unlikely that Oracle will see the entire $300 billion.

A cheap stock isn't always a good stock The main reason Oracle's RPO conundrum is so concerning is because the company is taking on a mountain of debt to build more data centers, so if it can't convert its backlog into revenue, it could find itself in a dire financial position in the future. As of May 31, it was sitting on $122 billion in long-term debt, and it has since announced plans to raise a further $40 billion through a mix of debt and equity.

That risk is now being reflected in Oracle's valuation. Its stock was trading at a price-to-earnings (P/E) ratio of 24.7 as of the market close on Wednesday, Aug. 19, making it cheaper than both the S&P 500 and Nasdaq-100 indexes, which had P/E ratios of 26.5 and 35, respectively. Simply put, investors seem unwilling to pay a market multiple for the stock, despite the incredible growth in the AI infrastructure business.

ORCL PE Ratio data by YCharts

When Oracle releases its fiscal 2027 first-quarter financial results in early September, management might be able to ease Wall Street's concerns by providing an update on the composition of its RPO. If the backlog is less concentrated than before, investors might feel better about the company's ability to convert most of it into revenue. Management might also commit to avoiding additional debt.

Despite Oracle's seemingly attractive valuation, I personally don't feel comfortable buying its stock ahead of its upcoming report, because the severe risks facing the business can't be fully resolved in a single quarter.

PennantPark Investment (NYSE:PNNT) and PennantPark Floating Rate Capital (NYSE:PFLT) both reported Q3 FY26 results on Aug. 10, run by the same manager, Art Penn. They share a middle-market credit playbook, but the portfolios and payout math diverge in ways that matter for a retiree writing checks off these yields.

Same Manager, Two Very Different Books PFLT is the larger and cleaner vehicle. Total investment income reached $66.09 million, the debt book is 99% floating rate and 89% first lien senior secured, and core NII of 26-cent per share covered the 24-cent quarterly base dividend. PNNT is smaller, broader, and softer. Revenue of $24.77 million fell 16.2% year over year, the portfolio shrank to $1,193.2M, and the mix still leans on subordinated debt and equity co-investments alongside first liens.

Metric PNNT PFLT Portfolio yield on debt 8.9% 9.8% Floating-rate exposure 87% 99% NAV per share $6.56 $10.26 Non-accruals (cost) 2.5% 1.0%
Where the Dividend Story Really Splits PFLT already took its medicine. Management cut the monthly base from 10 cents to 8 cents in mid-2026 and layered on a variable supplemental equal to 50% of NII above the base. The reset is painful, but core NII now sits above the base, and Penn told investors “our mission remains consistent to deliver a stable and well-covered dividend while preserving capital.”

PNNT tells a harder story. Core NII of 14 cents does not cover the 24-cent quarterly distribution. CFO Rick Allorto confirmed the gap is being filled by spillover income, which he expects to decline to about $0.40 per share by year-end 2026 from a peak of “a little over a dollar per share not that long ago.” The current supplemental is only communicated through the end of calendar 2026. After that, coverage must come from equity rotations and the PSLF refinancing that dropped its cost of capital from SOFR+266bps to SOFR+169bps. A distribution funded out of a draining reserve is exactly the setup we flagged in a free guide to the seven warning signs of a dividend at risk.

What I’m Watching Into 2027 With Fed funds parked at 3.75% since December, the tailwind that floated these BDCs to peak yields is gone. PFLT’s PSSL II joint venture at $390 million, targeted to exceed $1 billion over 12 to 18 months, is the accelerant. For PNNT, the real question is whether equity exits keep pace once the spillover buffer thins.

Why I Lean PFLT for Retirement Income For an investor funding retirement checks, PFLT screens as the more defensible income vehicle. The dividend has already been rebased, coverage is real at $0.26 core NII against a $0.24 base, and the book is 89% first lien with a lower non-accrual rate. Analysts agree, sitting at 3 Strong Buys, 2 Buys, 2 Holds. PNNT is the deeper-value play, trading at 0.57 times book, but the coverage math relies on a reserve that is visibly draining. A turnaround investor comfortable with the 37.7% one-year drawdown and confident in the equity rotation might still buy it. For a retiree, dividend reliability is what carries the most weight.

Contact [email protected] for any questions or corrections.