The easy way to read the AI data-centre boom is to count the gigawatts being announced. The more useful questions are where those gigawatts will come from, who will pay for them and whether the proposed site can ever be connected. That distinction matters because the numbers being discussed are no longer ordinary commercial loads. A single hyperscale data centre campus can draw several hundred megawatts of continuous power, on a par with a mid-sized industrial precinct or a small city, and some of the proposals in this piece are pitched at up to 1GW or 2GW (Figure 1). Networks built for households and existing industry were never sized for that kind of load to arrive in one location almost overnight. That is the pressure behind Australia's shift from simply welcoming data centres to attaching conditions to them. Figure 1: Aerial view of Microsoft’s new AI data centre campus in Mt Pleasant, Wisconsin (Source: Microsoft) The conditions now on the table are specific. New large facilities are expected to help underwrite the new electricity supply they need, rather than draw down capacity meant for everyone else. They are expected to fund their own connection costs, rather than push those costs onto other network users. They are expected to curtail their own power use when the grid is under strain, so a heatwave that stresses household supply does not get worse because a data centre keeps running at full draw. And they are expected to show they use water efficiently, because the cooling systems large computing facilities need can be heavy water users in regions where water is already contested. On top of that, the federal government has directed that new data centres should mostly run on renewable energy. Gas can help firm that supply when the sun is not shining or the wind is not blowing, but is not meant to be the primary source. Put together, this reshapes who is actually well positioned in the sector. It favours companies that already hold a working combination of developable land, grid access, new generation, storage, gas firming and fibre, over companies that simply hold gas in the ground or have attached an "AI" label to a project that predates any of this - and it looks like resources and energy companies are well-positioned to fill this gap. This story discusses five ASX small caps that are well-positioned to host data centres The vocabulary, in plain English A hyperscaler is a very large technology company, such as a cloud computing provider, that builds and runs its own giant data centres. A heads of agreement, or HoA, is an early document that sets out what two parties intend to agree. It can be non-binding, where neither side is legally obliged to proceed, or binding, where the parties have made a legal commitment even before a full contract is signed. A power purchase agreement, or PPA, is a contract to buy electricity from a specific generator over a set period. MW means megawatt and GW means gigawatt, a GW is 1,000 MW, and both measure electricity capacity, not the amount actually used. Freehold land is land the company owns outright, as opposed to a lease or an option to buy. Firming means adding backup generation, such as gas or batteries, so that renewable power stays reliable when the sun is not shining or the wind is not blowing. Voluntary administration is a formal process where an independent administrator takes control of a company that cannot pay its debts, to work out whether it can be restructured, sold, or wound up. How do the ASX Data Centre optionality plays compare? Table 1: Companies with AI data centre optionality Company ASX Market cap Key asset AI-opportunity Stage 1414 Degrees 14D approximately A$43m Aurora Energy Precinct, SA Up to 1GW renewable-powered campus Proposed, feasibility stage, non-binding HoA signed Beetaloo Energy Australia BTL approximately A$346m Carpentaria gas plus Weddell site, Northern Territory 2GW off-grid gas-powered campus concept Pre-FEED, feasibility stage Strike Energy STX approximately A$341m South Erregulla gas plant and precinct, WA Firmed renewable-powered industrial and data precinct Under construction, commissioning, data centre case speculative Frontier Energy FHE approximately A$216m Waroona solar and battery site, WA Renewable-powered campus beside 330kV transmission infrastructure Under construction, early works, data centre case speculative Pilot Energy PGY approximately A$6.8m Cliff Head/Arrowsmith and Three Springs, WA Modular compute plus proposed 50MW solar and battery campus Operating 0.4MW module, proposed 50MW site, company in voluntary administration Prices and market caps above are indicative, mostly from 6 to 7 August 2026, calculated from market price and the latest disclosed or post-raising share count where available, and can differ across delayed data feeds. Approximate prices used were A$0.061 for 14D, A$0.230 for BTL, A$0.095 for STX, A$0.195 for FHE and A$0.054 for PGY. Pilot's market cap uses its last traded price because its securities are currently suspended. Readers should check current prices and suspension status on the ASX before making any decision. 1. 1414 Degrees (ASX: 14D) 14D controls the 15.8-square-kilometre Aurora Energy Precinct near Port Augusta, South Australia, under long-term tenure. Moreover, the company's silicon-based thermal energy storage technology – SiBrick® – safely and efficiently stores renewable electricity as latent heat (Figure 2). Figure 2: SiBrick® - company's silicon-based thermal energy storage technology (Source: 14D Website) The site has highway access, water infrastructure, transcontinental fibre and high-voltage transmission next door. Its approved 140MW/280MWh battery, up to 900MW of solar potential and a proposed data centre campus give it a workable combination of renewable power, storage and land. Management says an initial 17MW can connect through 33kV infrastructure, around 200MW through 275kV infrastructure, and up to 1GW through staged upgrades. These are development pathways, not power that is already contracted. The grid operator AEMO and network owner ElectraNet have accepted generator performance standards for the battery, but final connection terms, an electricity offtake deal and construction funding are still outstanding. In July 2026, 14D signed a non-binding HoA with an unnamed Australian operator. It gives that operator exclusivity over an initial 40-hectare parcel and covers up to 1GW in total. The operator is expected to bring capital and operating expertise, but no lease, PPA, named hyperscaler customer or financial terms have been disclosed. 2. Beetaloo Energy Australia (ASX: BTL) BTL holds exclusive negotiating rights, described as a "not-to-deal" arrangement rather than ownership or a signed lease, over 185 hectares at Weddell, about 30 kilometres from Darwin (Figure 3).The company has proposed a two-campus, up to 2GW AI data centre development there. The flat, cleared site sits near the planned Territory Energy Link and a proposed subsea cable landing point connecting to Asia. Figure 3: Beetaloo's Major holdings across both McArthur and Beetaloo Basins (Source: BTL Website) The concept would run on gas generated on site from BTL's Beetaloo gas holdings, built larger than needed so it does not draw power from the Northern Territory grid. Fibre access, water supply and cooling design have not been made public. The gas backing this concept sits at BTL's Carpentaria project, where the company reports about 1.6 trillion cubic feet of contingent gas resources. Pilot gas production is targeted for late 2026, so the fuel source is not yet in production. The Weddell project remains at concept and pre-FEED stage, meaning early feasibility work before a full front-end engineering design, and still needs a funding consortium, proven gas supply, financing, approvals and signed land and infrastructure agreements. A memorandum of understanding with Halliburton covers technical collaboration only, not project capital or a data centre customer. 3. Strike Energy (ASX: STX) Strike owns 3,500 hectares of freehold, previously cleared farmland at South Erregulla, roughly 280 kilometres north of Perth and 45 kilometres east of Three Springs, Western Australia (Figure 4).The site combines gas resources, planned renewable generation and an 85MW gas-fired power station. The original development case pointed to transmission lines about 15 kilometres away, gas reserves in the highest confidence category, and around 1.3 petajoules of annual fuel demand supporting more than 25 years of operation. Figure 4: Strike’s petroleum acreage in the onshore, northern Perth Basin (Source: STK Website) By 29 June 2026, all 20 generating units at the power station had been commissioned on gas. Grid and network commissioning, along with dedicated connection works, were still continuing at that date, so the power supply is not yet fully proven for a data centre load. The 85MW plant was designed mainly as a peaking and reserve-capacity asset, not dedicated round-the-clock baseload supply, so additional generation or storage would be needed to supply a data centre. Strike has publicly argued that gas, solar and wind in the Mid West region could support data centres, but this is company positioning rather than a signed deal. No data centre developer, hyperscaler customer, site plan, fibre connection, cooling water source or commercial agreement has been disclosed. 4. Frontier Energy (ASX: FHE) Frontier owns approximately 820 to 830 hectares at Waroona, around 120 kilometres south of Perth, about 500 metres from Western Power's Landwehr transmission terminal. Stage One of its project is 132MW of solar generation, an 81.5MW/565MWh battery, and a new 330kV substation. A fixed-price construction contract has been awarded, and lenders have given credit approval for debt of up to A$280 million, though that debt still needs final documentation and conditions to be met. Combined with an equity raising in July, management expects Stage One to be fully funded, with first power targeted for the second half of 2028. A further Stage Two, covering about 120MW of solar and an 80MW battery, has approvals in place. Figure 5: FHE Waroona Infrastructure (Source: FHE June 2026 Presentation) This makes Frontier one of the better renewable-powered land assets in the screen on paper, with freehold scale, a high-voltage connection point and long-duration storage matching the direction of federal policy. Water capacity, carrier-grade fibre and any formal data centre development application have not been publicly verified. Media reports said AI company Sharon AI had considered an equity position in Frontier, and that Frontier receives approaches from interested parties, but Sharon AI denied it had bought the reported stake. No agreement, customer or data centre plan has been announced by Frontier itself. 5. Pilot Energy (ASX: PGY) - In administration Pilot commissioned 0.4MW of a planned 1MW modular high-performance-compute facility at Arrowsmith, WA, in April 2026, using existing gas and part of its approximately 4.4MW of generation capacity. The remaining 0.6MW needs an electrical upgrade. Management has pointed to Starlink and a Vocus fibre route about 10 kilometres away, but a carrier-grade fibre connection, a confirmed water source and the actual workload running on the facility have not been verified. Separately, Pilot signed a binding HoA with a company called SNE, covering land acquisition, solar and battery storage and a proposed 50MW data centre at Three Springs, with milestone payments due to Pilot. Due diligence and the definitive contracts under that HoA were incomplete, and there was no certainty the deal would proceed to completion. Pilot subsequently appointed voluntary administrators, and its securities are now suspended from trading. That means the 0.4MW facility's current operating status, and whether its counterparties still intend to proceed, both need to be reconfirmed. Pilot's larger Mid West infrastructure, including roads, pipelines and an offshore reservoir at Cliff Head, may still hold value in its own right, but that infrastructure does not by itself make a working data centre campus. Which of the five stand out 1414 Degrees offers the widest gap between its roughly A$43 million market value and a potentially large campus of up to 1GW. Aurora already combines a large controlled site, transmission access, fibre, water infrastructure, solar potential and an approved battery project. The HoA is real, if early, commercial evidence. A binding lease, a staged connection agreement or a named creditworthy customer could change how the market sees the company. The current discount reflects that exclusivity is non-binding, grid capacity is not reserved at the full 1GW, and construction funding has not been arranged. Strike Energy's 3,500-hectare freehold precinct, dedicated gas, 85MW of generation and grid works are more tangible physical infrastructure than most companies on this list can show. A renewable partner and a signed data centre tenant could let Strike use its existing investment to move faster than a company starting from nothing. But Strike has no data centre counterparty at all. Its current valuation rests mainly on its energy assets, so upside from AI data centres is possible but unproven. Frontier Energy is the closest company to displacing Strike in this group. It has a better fit with the renewable-power policy direction, but its data centre interest is based on media reports rather than a signed agreement, and first power is not expected until 2028. Pilot Energy is the only one of the five that has both a binding, data centre-specific HoA and equipment already running at a customer-facing scale. On paper, that is more commercial progress than any of the other four can show. In practice, the company's move into voluntary administration overwhelms that progress. Whether the SNE agreement, the operating module, or Pilot's broader infrastructure hold any of their earlier value now depends on the outcome of the administration, not on the underlying data centre case. The risks that apply across this group Building a data centre-scale campus, plus new generation, storage and fibre, can cost far more than these companies are currently worth on the market. 14D has recently raised equity and would need considerably more capital to fund a full-scale campus. Frontier's credit-approved debt and equity package is the most advanced project funding among the five, but it is still subject to conditions. Beetaloo still needs to assemble consortium-scale capital before Weddell can proceed. Pilot's funding position is now a matter for its administrators rather than ordinary equity or debt raising. A nearby transmission line does not mean firm capacity is actually available. Connection studies, equipment needed to maintain grid stability, network upgrades, curtailment rules and queue position can all affect what a company can actually deliver. Announced capacity targets should not be treated as capacity that is already connected. Water rights and cooling design are often not disclosed. Dry cooling reduces water use but can increase construction cost and energy use. Being near a fibre route is not the same as having a confirmed, carrier-grade connection. Data centre, generation, transmission and water approvals are separate approval processes. Gas developments carry additional emissions, Traditional Owner, environmental and social approval risks. Land options or exclusivity periods can lapse before approvals are granted. Forecasts for AI computing demand may prove too high. More efficient chips, liquid cooling and changes in where workloads are located could reduce the value of remote sites. Large technology companies can also bypass smaller companies and contract directly with utilities. Gas price, how much of a contingent gas resource actually converts into a produced reserve, production decline and carbon costs can all reduce the economics of gas-fired, behind-the-meter power. Under Australia's developing policy framework, gas-only campuses face the most exposure, and renewable PPAs and storage are likely to be needed alongside them. An HoA, whether binding or non-binding, is not revenue on its own, and a signed agreement can still fail to complete, as Pilot's own situation shows. Small management teams have to manage energy, property, telecommunications and digital infrastructure work all at the same time, on top of everyday production and funding pressures. A company appointing administrators can leave every other commercial arrangement it has signed in doubt. Samso Concluding Comments 1414 Degrees has the most complete package among the five, on the evidence available. Aurora combines controlled land, high-voltage infrastructure, a fibre corridor, water access, solar potential and an approved battery, and its non-binding HoA is a step beyond a purely theoretical proposal. At a market value of roughly A$43 million, successfully de-risking the project could be significant for the company. It remains a high-risk development, with no confirmed customer, lease, large-scale connection or funding secured yet. Strike Energy has the more tangible, near-operational infrastructure of the remaining companies with an intact balance sheet, but no data centre counterparty at all. Frontier Energy would move ahead of Strike if it announced a credible operator, or a verified data centre fibre and water solution, of its own. Pilot Energy is the clearest illustration in this group of the gap between having signed something and having a viable business. It has the most advanced data centre-specific agreement and the most equipment actually running, and also the single largest company-level risk of the five, now that it is in the hands of administrators. The Samso Way – Seek the Research Here at Samso, we pride ourselves on delivering content for investors that is independent and informed by over three decades of experience in the industry. Our content is well-researched and is only created if I see merit in discussing the company's story. Our mission is simple: cut through the noise and spotlight what matters—genuine stories, grounded insights, and real opportunity. Our content is well-researched and is only created if the team sees merit in discussing the company or concept. Investors can explore our three core platforms: Coffee with Samso Samso Insights Samso News There may be numerous paths to success in investing, but the common thread among successful individuals is that they remain committed to making informed decisions. Equip yourself with the right knowledge and tools, and you will be well on your way to achieving your financial goals. Most importantly, investors need to be absolutely diligent in understanding their own risk-reward tolerance and capabilities. Never bite off more than you can chew. As they say, Rome wasn’t built in a day, and the Great Wall stood because it took centuries to complete. The Samso Philosophy: Stay curious. Stay sharp. And remember—digging deeper always uncovers the real value. In Life, there is no such thing as a Free Lunch. Never bite off more than you can chew is my parting comment. Happy Investing, and the only four-letter word you need to know is DYOR. To support our independent work, please head over to our Support Page and give us a helping hand in any of the ways listed. This is a new initiative for the Samso Platform, and it was always the concept of Samso when we started this journey in 2018. Disclaimer The information or opinions provided herein do not constitute investment advice, an offer, or solicitation to subscribe for, purchase, or sell the investment product(s) mentioned herein. It does not take into consideration, nor have any regard to your specific investment objectives, financial situation, risk profile, tax position and particular, or unique needs and constraints. Read full Disclaimer. About Samso Samso is a trusted platform that equips dedicated investors with up-to-date industry knowledge and insights from top CEOs and thought leaders. By staying informed on business advancements and market trends, investors can enhance their financial decisions through a combination of expert guidance and their own research. Samso News | www.samso.com.au | An Investor Lens on ASX-Listed Companies

Part 1 argued that the world is spending enormous sums to produce AI, whether or not that spending yet buys value. This part follows the money into the physical supply chain the spending pays for, the chips and the buildings, and asks a plain question: which of it can an Australian investor actually own on the ASX, and which of it, including the much-hyped vector databases, is simply not listed here at all. 1.00 — THE PICKS-AND-SHOVELS IDEA, AND ITS LIMIT Selling shovels in a token rush There is an old rule about gold rushes: the steadier money was often made selling picks, shovels and denim to the miners rather than digging for the metal. The equivalent question for artificial intelligence is which businesses get paid no matter which AI application eventually wins. Part 1 made the case that a company's AI spending should be judged on cost per outcome, the cost of a finished, useful result, rather than on how much it spends or how loudly it talks about AI. That test is deliberately hard on the buyers of AI. The suppliers are a different proposition. Whether or not a given bank or retailer ever proves its chatbot was worth it, the tokens (the small chunks of text an AI reads and writes, and the unit it is billed by) still had to be produced somewhere, on physical chips, inside physical buildings, drawing physical power. FIG. 01 lays out that supply chain as a stack, and marks the plain truth of it: an ASX investor can buy the bottom of the stack (the buildings and, speculatively, the chips) but not the fashionable middle (the models and the vector databases). Now the limit, stated before any company is named. Australia is a small part of this supply chain. It has no large-scale chip fabrication, so most of its listed semiconductor names are early-stage technology-licensing hopefuls rather than profitable manufacturers. What follows maps what is here, what it does, and what each name's main risk is. It is not a list of recommendations, and the counterweights are as important as the descriptions. 2.00 — THE CHIP LAYER: SEMICONDUCTORS ON THE ASX Small, speculative, and mostly selling ideas rather than chips A semiconductor is the material, usually silicon, from which computer chips are made, and "the semiconductor industry" is the chain that turns it into working processors and memory. That chain has several rungs, and the handful of ASX names sit on different ones. FIG. 02 places them. The names differ as much in maturity as in what they do. FIG. 03 maps them on two axes an investor cares about: how far each is from real, repeatable revenue, and how speculative it remains. Most cluster at the pre-revenue, high-risk corner. Only DUG Technology and, to a lesser extent, Revasum sit near the "actual revenue" end. Before the profiles, PLATE 3 is a reminder of what this whole layer eventually produces: physical chips. Design and intellectual property BRN · AXE BrainChip Holdings (ASX: BRN) designs a low-power processor called Akida for "edge" AI, meaning AI that runs on a device such as a drone, sensor or piece of defence equipment rather than in a distant data centre. It licenses the design to others. For the year to December 2025, revenue was about US$1.89 million (up from about US$0.4 million a year earlier) against a net loss of about US$20 million, and a next-generation processor (the AKD2500) is expected in prototype during 2026. Risk Revenue is minimal, the licensing model is unproven at commercial scale, and the company has burned cash for years. The share price reflects hope for the future, not current earnings. PLATE 1 shows the Akida design BrainChip licenses to others, with its on-chip processing blocks and "neuron fabric". Archer Materials (ASX: AXE) is a research-stage developer of a quantum computing chip and a separate biochip. Quantum computing is an experimental approach that could one day solve certain problems far faster than today's machines. Archer reported a net loss of about A$2.99 million for the half and generates almost no revenue. Risk Highly speculative, pre-revenue, and any commercialisation is years away and uncertain. Memory WBT · 4DS Both names are chasing the same prize shown in PLATE 2: a memory that is as fast as today's working memory but keeps its data without power. Weebit Nano (ASX: WBT) has developed a memory technology called ReRAM (resistive memory) that aims to be faster and more power-efficient than the Flash memory in today's devices, and licenses it to chipmakers. It reported a record half-year revenue of about A$5.6 million, referenced a licensing deal with Texas Instruments, held about A$82.8 million in cash, and guided to more than A$10 million of revenue in the 2026 financial year. Its market value is above A$1 billion. Risk The valuation rests on future royalties that are still ramping, not on current profits, so it is priced for success that has not yet arrived. 4DS Memory (ASX: 4DS) is developing its own version of resistive memory aimed at high-density storage for mobile devices and data centres, which makes it the closest chip-level name to the "data storage for AI" theme. Revenue is negligible at about A$316,000, with roughly A$3.34 million in cash plus about A$5.45 million in deposits. Risk Very early stage with almost no revenue; it needs a commercial manufacturing partner before it counts for much. Materials BLG · SLX BluGlass (ASX: BLG) makes laser components from gallium nitride, a semiconductor material used in lasers and power electronics, using its own manufacturing process. It reported revenue of about A$5.76 million (up about 14 per cent), a net loss of about A$4.10 million, and referenced a US defence contract and a project pipeline. Risk Loss-making and facing the challenge of scaling a specialised process to volume. SILEX Systems (ASX: SLX) belongs here only with a clear caveat. Its main business is laser uranium enrichment technology, not chips. Its semiconductor-relevant activity is a "Quantum Silicon" plant producing highly purified silicon-28, a material used to build quantum computing chips. That is a genuine but secondary part of the company. Risk This is primarily an enrichment-technology company; anyone buying it for semiconductor exposure is really buying the enrichment story with a small chip option attached. Equipment and high-performance computing RVS · DUG Revasum (ASX: RVS) is the most literal "shovels" name on this list. It makes the grinding, polishing and planarising machines used to manufacture semiconductor wafers, including silicon carbide, and so sells tools to the industry rather than betting on any single chip winning. Risk A small supplier into a cyclical capital-equipment market; demand rises and falls with chipmakers' own spending. Current financials should be checked directly. DUG Technology (ASX: DUG) is different again: an established provider of high-performance computing (very large-scale number crunching), cloud services, geoscience work and software. Unlike the chip hopefuls it earns real revenue, and reported strong growth with its high-performance computing demand rising and cash flow turning positive. Risk Small in a field where the giants of cloud computing set the price, so it must keep finding niches they do not serve. 3.00 — DATA STORAGE FOR AI: THE DATA CENTRE, NOT THE DATABASE The vector database everyone talks about is not listed here Ask where AI keeps its data and the fashionable answer is the vector database. The realistic answer for an ASX investor is that you cannot buy one here, so the real local exposure is the building the data sits in. First, what a vector database is, in plain terms. Traditional databases find things by exact match, the way a phone book finds a name you spell correctly. A vector database stores information as strings of numbers that capture meaning, so it can find things that are similar in sense rather than identical in spelling. It is the technology behind most "chat with your documents" features, where an AI answers using a company's own material. It is central to how businesses are putting AI to work. It is also not something you can own on the ASX. The specialist vector database companies are privately held, and the listed ways to touch the technology are large overseas firms that offer it as one feature among many. One of the more prominent independents, Marqo, has Australian roots but is private (a point worth confirming before relying on it). FIG. 04 sets out where the category actually trades. So the ASX way to own "storage for AI" is the physical layer: the data centre. A data centre is a large, secure, heavily powered building full of computers that store data and run software for other companies. Its size is measured in megawatts (MW) of power, because power, not floor space, is what limits how much AI computing a site can run. FIG. 05 explains how a data centre turns power into profit, and where the cost-per-outcome test from Part 1 bites. Australia has a genuine, investable cluster of listed operators. FIG. 06 compares the capacity figures the main names have disclosed, with an important caveat in the caption about how those figures are defined. The listed data-centre names NXT · MAQ · DGT · GMG · IFT NextDC (ASX: NXT) is described as the purest data centre business on the ASX: it builds and runs the buildings that house AI computing and storage. It referenced 667 megawatts of contracted capacity (as at April 2026, up from 416.6 megawatts in December 2025), a A$1.5 billion equity entitlement offer forming part of a A$2.2 billion capital plan announced in April 2026, and a partnership with OpenAI to build sovereign AI infrastructure in Sydney (data centres kept on Australian soil under Australian control). Risk It spends enormous sums upfront; returns depend on filling that capacity with paying tenants at prices above its cost of capital. PLATE 4 is an artist's impression of NextDC's S4 Sydney campus, one of the hyperscale sites behind its contracted capacity. Macquarie Technology (ASX: MAQ) runs data centres alongside cloud and telecommunications businesses, and hosts a large share of Australian Federal Government computing, which is a demanding, sticky customer. A new 47 megawatt facility was referenced for opening in September 2026, with more in the pipeline. Its market value is around A$1.7 billion. Risk Smaller than NextDC, so expansion carries more execution and funding risk relative to its size. DigiCo Infrastructure REIT (ASX: DGT) is a data centre real estate investment trust (a listed property trust). In May 2026 it announced the sale of its Chicago data centre for about US$750 million to strengthen its balance sheet and fund growth, including an approved 88 megawatt expansion of its SYD1 Sydney campus. It has traded below its A$5.00 December 2024 listing price. Risk Recently listed and already below its issue price; as a property trust it is sensitive to interest rates. Goodman Group (ASX: GMG) is a large industrial property group that has pivoted hard towards data centres, referencing a development pipeline of about A$14.4 billion that is roughly 73 per cent weighted to them, and multi-gigawatt power capacity. It is the most diversified, and therefore lower-risk, way to touch the theme. Risk Data centres are only a slice of a business still exposed to the broader property cycle, so the AI exposure is diluted. PLATE 5 shows one of Goodman's data centre developments, the kind of asset its pipeline is now weighted towards. Infratil (ASX: IFT) owns about half of CDC Data Centres, a major operator that referenced a 555 megawatt contract with a large cloud customer, taking its contracted capacity above one gigawatt. Infratil is a New Zealand-based, dual-listed infrastructure investor. Risk A diversified infrastructure owner, so the AI exposure is real but indirect and shared with airports, energy and other assets. The connective tissue MP1 Megaport (ASX: MP1) is not storage, but it is worth naming as the plumbing between clouds and data centres. It sells "network-as-a-service", on-demand connections that let businesses link up cloud and data-centre services, and has been positioning towards AI workloads. Risk It provides connectivity rather than storage or compute, and competes in a fast-moving market; include it as an adjacency, not a core storage play. 4.00 — HOW THE TWO HALVES CONNECT The cost-per-outcome test applies to the shovels too It would be easy to read Part 1 and Part 2 as opposites: one urging caution on AI spending, the other pointing at ways to buy into it. They are the same discipline applied twice. The test from Part 1 was cost per outcome, not inputs. Turn it on the picks-and-shovels names and it still bites. A data centre operator is worth it when it signs paying tenants at a return above its cost of capital, not when it announces a headline number of gigawatts. A chip hopeful is worth it when it converts its intellectual property into recurring royalties, not when it runs an impressive demonstration. A high-performance computing provider is worth it when it wins work the cloud giants cannot serve more cheaply. In each case the announcement is the bill; the signed contract or the royalty cheque is the value. There is also a real tension between the two parts that a clear-eyed investor should hold. Part 1's cost discipline, if it takes hold, would slow the growth in tokens per task, which is part of what drives demand for all these shovels. The counter is that adoption has been growing faster than efficiency: Google's token volumes rose roughly seven-fold in a year even as the cost per token collapsed. That pattern, more use swamping cheaper units, is the bull case for the infrastructure. It is a reasonable bet, not a law, and Part 1's bearbox applies here too. References & sources Company activities and figures below are drawn from the secondary sources listed and are as at the companies' most recently reported results referenced there, broadly the first half of the 2026 financial year. Every company figure is market-sensitive and should be confirmed against each company's own ASX announcements and refreshed on publication day. All figures shown in the visuals are original Samso illustrations of the data named in each caption. Figures in this note were checked against the companies’ own ASX presentations and announcements where available (to July 2026); the NextDC OpenAI partnership, DigiCo’s US$750 million Chicago sale, Weebit’s Texas Instruments licence, Goodman’s pipeline, Macquarie’s 47MW facility and Infratil’s 555MW CDC contract were all confirmed against primary sources. This note names companies to illustrate where exposure exists; it is not a recommendation on any of them. Photographs and company diagrams labelled PLATE are reproduced from the companies' public ASX release presentations, with attribution in each caption, and are distinct from the original Samso illustrations labelled FIG. Veye — "Top 5 ASX semiconductor stocks 2026": Weebit Nano (WBT), Archer Materials (AXE), BluGlass (BLG), 4DS Memory (4DS), BrainChip (BRN) descriptions and half-year figures. Stocks Down Under — "Marvell... 3 ASX AI chip stocks WBT AXE BRN" and related coverage: Weebit Nano, Archer Materials and BrainChip technology and status; BrainChip market value and AKD-series roadmap. Stockhead — "Revasum leads a pack of ASX semiconductor stocks...": Revasum (RVS) grinding, polishing and CMP wafer equipment; SILEX (SLX), BrainChip, Weebit Nano, 4DS and BluGlass activity summaries. Silex Systems (SLX) company materials and Stocks Down Under coverage: primary business in laser uranium enrichment technology; "Quantum Silicon" (Q-Si) plant producing enriched silicon-28 for quantum computing. DUG Technology (DUG) FY26 half-year results release and Stocks Down Under coverage: high-performance computing, cloud and geoscience services; revenue growth and cash flow. Stocks Down Under — "The ASX AI infrastructure boom... data centre stocks": NextDC (NXT) 667 MW contracted, A$1.5 bn raise and OpenAI partnership; Macquarie Technology (MAQ) government hosting and 47 MW facility; DigiCo (DGT) Chicago sale and Sydney project; Goodman Group (GMG) pipeline and power capacity; Infratil (IFT) / CDC 555 MW contract and 1 GW-plus contracted capacity. Kalkine and Megaport (MP1) investor materials: network-as-a-service cloud connectivity and AI-workload positioning. AI Magazine — "Top 10 vector databases for AI"; MarketsandMarkets vector database market overview: specialist vector database vendors (Pinecone, Weaviate, Qdrant, Milvus/Zilliz, Marqo) and listed providers offering vector search as a feature (MongoDB, Microsoft, Amazon); no ASX-listed pure-play identified. Reproduced company visuals (PLATES): BrainChip Holdings 2026 AGM presentation (6 May 2026); 4DS Memory Capital Raising Presentation (January 2025); NextDC Investor Presentation (20 April 2026); Goodman Group 1H FY26 Results Presentation (19 February 2026). Each reproduced with attribution in its caption. The Samso Way – Seek the Research Here at Samso, we pride ourselves on delivering content for investors that is independent and informed by over three decades of experience in the industry. Our content is well-researched and is only created if I see merit in discussing the company's story. Our mission is simple: cut through the noise and spotlight what matters—genuine stories, grounded insights, and real opportunity. Our content is well-researched and is only created if the team sees merit in discussing the company or concept. Investors can explore our three core platforms: Coffee with Samso Samso Insights Samso News There may be numerous paths to success in investing, but the common thread among successful individuals is that they remain committed to making informed decisions. Equip yourself with the right knowledge and tools, and you will be well on your way to achieving your financial goals. Most importantly, investors need to be absolutely diligent in understanding their own risk-reward tolerance and capabilities. Never bite off more than you can chew. As they say, Rome wasn't built in a day, and the Great Wall stood because it took centuries to complete. The Samso Philosophy: Stay curious. Stay sharp. And remember—digging deeper always uncovers the real value. In Life, there is no such thing as a Free Lunch. Never bite off more than you can chew is my parting comment. Happy Investing, and the only four-letter word you need to know is DYOR. To support our independent work, please head over to our Support Page and give us a helping hand in any of the ways listed. This is a new initiative for the Samso Platform, and it was always the concept of Samso when we started this journey in 2018. Disclaimer The information or opinions provided herein do not constitute investment advice, an offer, or solicitation to subscribe for, purchase, or sell the investment product(s) mentioned herein. It does not take into consideration, nor have any regard to your specific investment objectives, financial situation, risk profile, tax position and particular, or unique needs and constraints. Read full Disclaimer. About Samso Samso is a trusted platform that equips dedicated investors with up-to-date industry knowledge and insights from top CEOs and thought leaders. By staying informed on business advancements and market trends, investors can enhance their financial decisions through a combination of expert guidance and their own research. Samso Insights | www.samso.com.au | An Investor Lens on ASX-Listed Companies

I use AI in my own work, and I can see how far it extends what one person can do. But when this industry talks about itself, the words get in the way. This Insight takes one real presentation by a leading AI company, quotes the jargon exactly as spoken, and translates it piece by piece, so that by the end you can read any AI announcement and know what is actually being said. 1.00 — WHY I AM WRITING THIS The language got between me and the industry - The AI Token Mystery I am writing this Insight for a simple reason: the language of AI still confuses me. I use AI, and I can see how far it extends what one person can do. But when this industry talks about itself, the words get in the way. To understand any industry, you have to understand what its people are talking about. Nobody in resources would expect a newcomer to know what JORC means, yet AI announcements assume the whole world speaks their language. What settled it for me was a recent presentation by Kimi, one of the Chinese AI companies now competing with the biggest American labs. I watched it wanting to learn, and the words were mind-numbing: token efficiency, optimisers, context length, agent swarms. If I was going to understand what these companies are building, and what the market is paying for, I had to decode the language first. This Insight is that decoding, for me as much as for you. The presentation was given by Yang Zhilin, the founder of Moonshot AI, the Beijing company behind the Kimi models, at NVIDIA's GTC conference in San Jose on 19 March 2026. The session was called "How We Scaled Kimi K2.5". Here is a taste of it, quoted as spoken: If that reads like another language, this piece is for you. By the end, you will be able to re-read that paragraph and follow every word. And because one company from Beijing is not the whole industry, each section pairs the Kimi talk with a second voice, OpenAI, Anthropic, Google, Microsoft, NVIDIA, Meta or xAI, saying the same kind of thing in their own words. The jargon is not one company's habit. It is the working language of a global industry that now carries trillion-dollar price tags. 2.00 — THE UMBRELLA WORDS: AI, MACHINE LEARNING AND WHAT AN LLM ACTUALLY IS Five nested words, and the engine behind the app Most confusion starts above the jargon, with words that sound interchangeable but are not. They nest inside each other like Russian dolls, and FIG. 01 draws the nesting. Artificial intelligence (AI) is the whole field: any attempt to make computers do things that normally need human intelligence. Machine learning is the approach that now dominates the field: instead of programmers writing rules, the computer works out its own rules from examples. Deep learning is machine learning done with a neural network, a program loosely inspired by the brain, built from many layers of simple connected units. Generative AI is the branch that produces new material, text, images, video and computer code, rather than just sorting or scoring things. And a large language model (LLM) is the kind of generative AI behind the chat tools: a very large neural network trained on enormous amounts of text to predict what comes next. The letters in GPT describe exactly that: Generative, Pre-trained, Transformer, where the transformer is the 2017 model design that made today's AI work at scale. You do not need to know how a transformer works, only that when you read the word, it means the standard engine design the whole industry builds on. The other distinction worth having is between the engine, the product and the company. The model is the engine. The chatbot or app is the car built around it. The company is the maker. OpenAI makes the GPT models that power ChatGPT. Anthropic makes Claude. Google makes Gemini. xAI makes Grok. Meta built Llama. Moonshot AI makes Kimi. When a headline says "GPT-5.6 beats Gemini", it is comparing engines, not apps, and certainly not businesses. 3.00 — THE TOKEN: THE UNIT EVERYTHING ELSE IS PRICED IN A small piece of a word, and the industry's whole accounting system Now the word in this article's title. A token is a small chunk of text, roughly three-quarters of an English word. Before an AI model reads anything, the text is chopped into these chunks, a step called tokenisation, and everything the model reads and writes is counted in them. OpenAI's own documentation gives the rule of thumb: 1,000 tokens is about 750 words. FIG. 02 shows a sentence cut into tokens; FIG. 03 puts sizes in scale. Why does one small word carry so much weight? Because the token plays three roles at once, and each role turns up in a different kind of announcement. The price role first. FIG. 04 compares what the major labs charged per million tokens in late July 2026. Two things stand out. Output tokens cost more than input tokens everywhere, because reading is done in one parallel pass while writing happens one token at a time and takes more computing effort. And the spread is enormous: from about US$0.28 per million output tokens at the cheapest Chinese model to US$90 at OpenAI's dearest professional tier, a gap of roughly 320 times. Price is a choice about quality, speed and brand, not a fixed cost of the technology. Then the usage role, which produced the most quoted number of 2026. Google's chief executive Sundar Pichai told the company's I/O conference in May that Google's products went from processing 9.7 trillion tokens a month two years ago to about 480 trillion a year ago, to "over 3.2 quadrillion per month" now, with its developer services handling "roughly 19 billion tokens per minute". FIG. 05 draws that curve. NVIDIA's founder Jensen Huang put the same idea in one sentence at GTC, the same conference where the Kimi talk was given: "Tokens are the building blocks of AI... your data center is now a factory to generate tokens." 4.00 — TRAINING AND INFERENCE: BUILDING THE PLANT AND RUNNING IT The two lives of a model, and the words that live inside each Every AI model has two lives. Training is the first: the model is shown trillions of tokens of text and slowly adjusts itself until it can predict language well. It happens once, takes months, and costs an enormous amount. Inference is the second: the finished model answers questions, every day, for everyone. It is the recurring cost of every token sold. FIG. 06 draws the split, in a shape any resources investor will recognise. Inside the model sit its parameters, also called weights: the billions of numbers the model adjusted during training, which hold everything it learned. Model sizes are quoted in parameters the way deposits are quoted in tonnes. Kimi K2, the model in our presentation, has one trillion of them. During training, progress is tracked by a single error score called the loss: the lower the loss, the better the model predicts. The arithmetic runs on GPUs, graphics processors from NVIDIA and others that turned out to be ideal for AI, or on TPUs, the equivalent chips Google designs for itself. And then there is the word that made me stop the Kimi video: the optimiser. The optimiser is the recipe the model follows while it learns, the set of rules deciding how much to adjust after each mistake. Almost the whole industry uses a 2014-era recipe called Adam. Moonshot's claim at GTC was that its new optimiser, called Muon (and MuonClip once stabilised), gets roughly twice the learning out of the same data. That is why Yang Zhilin said the line quoted in Section 1.00: with a two-times token efficiency, "50 trillion tokens" of data behaves like 100 trillion. Not magic; a better recipe. The claim is the vendor's own, from its own paper, and is best read that way. 5.00 — SCALING LAWS, TOKEN EFFICIENCY AND THE DATA WALL Why "bigger" has rules, and why the data is running out The AI boom rests on an empirical observation called the scaling laws: make the model bigger, feed it more data and give it more computing power, all together, and its results improve along a curve you can predict in advance. The landmark papers behind this were published by OpenAI researchers in 2020 and DeepMind researchers in 2022, and the industry has bet hundreds of billions of dollars that the curve keeps holding. OpenAI's Sam Altman stated it plainly in a 2025 essay: "The intelligence of an AI model roughly equals the log of the resources used to train and run it." But one of those three inputs is not like the others. Chips can be bought and models can be enlarged, yet the supply of good training text is finite. The industry calls this the data wall: the highest-quality human-written text, roughly speaking the useful part of the internet, has largely already been used. Anthropic's chief executive Dario Amodei put the underlying puzzle well in a February 2026 interview: "Humans don't see trillions of words. There is an actual sample efficiency difference here", meaning people learn language from far less data than models need. Two responses dominate. One is synthetic data: text generated by AI models to train other AI models. The other is token efficiency: squeezing more learning out of every token you already have, which is exactly what Moonshot's optimiser claim is about, and why a dry-sounding piece of mathematics was the headline of a major conference talk. 6.00 — CONTEXT WINDOWS: HOW MUCH THE MODEL CAN HOLD IN ITS HEAD The working memory, measured in tokens The context window is the model's working memory: the maximum number of tokens it can consider at once, covering both what you give it and what it writes back. Once a conversation or document exceeds the window, the earliest material falls out of view. Window sizes have become a headline specification, and FIG. 07 compares the current flagships. Why the arms race for bigger windows? Because the window sets what kind of work the model can take on. A model that holds a paragraph can answer a question. A model that holds a data room can audit it. In the Kimi talk, this is the second of the three things Moonshot says it scaled: their Kimi Linear design mixes a cheaper form of attention (the mechanism the model uses to weigh which earlier tokens matter) with the standard, expensive kind, in a three-to-one ratio, precisely to make very long working memories affordable. The stated goal: an AI that can run "for days or even weeks" on a single task without losing the thread. Anthropic's marketing makes the same promise from the other side of the Pacific: its Claude Sonnet 4.5 launch boasted that the model "maintains focus for more than 30 hours on complex, multi-step tasks". 7.00 — GETTING ANSWERS OUT: PROMPTS, REASONING, MULTIMODAL AND LOOKING THINGS UP The words you meet when you actually use the tools The vocabulary so far describes how models are built. This section covers the words that describe using them, the ones an investor meets in product launches rather than research papers. A prompt is simply the instruction you type, and prompt engineering is the craft of writing instructions that get better results. A reasoning model is the industry's newest product category: a model that can spend extra time working through a hard problem step by step before answering, producing an internal chain of thought. The extra thinking happens at answer time, so the industry calls it test-time compute, and the crucial commercial detail is that the thinking is billed in tokens like everything else. NVIDIA's Jensen Huang explained it at CES in January 2026 in one line: "We also have test-time scaling, which is another way of saying thinking. You think in real time." And in another: "Instead of a one-shot answer, inference is now a thinking process." Multimodal means one model handling more than one kind of material: text, images, audio and video together. The Kimi presentation is a worked example: Moonshot says K2.5 was the first open model trained on vision and text jointly "from day one" rather than bolting image skills onto a finished text model, and reported a result worth noticing: training the model on vision tasks alone improved its text results too, and a strong text model lifted its vision results. The claim is the vendor's own, but the direction, single models that see and read, is industry-wide. Then there is the machinery for making a model useful on your information. RAG, retrieval-augmented generation, lets a model look answers up in a nominated set of documents instead of relying on what it absorbed in training, which sharply reduces invention. The look-up usually runs on a vector database, a special store that searches by meaning rather than by exact words. Samso has covered vector databases before, in the AI & Capital piece "How to buy the AI build-out on the ASX", where our research found there is no pure-play way to buy one on the ASX. Guardrails are the rules wrapped around a model to stop harmful or off-limits outputs. Which brings us to the industry's most famous word for failure. A hallucination is a confident, fluent, wrong answer. It happens because a language model is a prediction machine: it produces the most plausible next tokens, and plausible is not the same as true. The rate has fallen with each generation, but it has not reached zero, and no vendor claims it has. 8.00 — AGENTS AND AGENT SWARMS: FROM ANSWERING TO DOING The industry's favourite word of 2026 An agent is an AI model given tools and permission to take steps on its own: search the web, run computer code, file the result, then decide what to do next. The adjective agentic now appears in nearly every AI announcement, and it marks the industry's core shift: from software that answers questions to software that completes tasks. The Kimi presentation's third big idea takes this one step further. In an agent swarm, an orchestrator (a lead agent) breaks a large job into pieces and hands them to a team of sub-agents working in parallel, then collects and assembles the results. Yang Zhilin's own analogy was a company: a chief executive decomposing work across departments, with researchers, developers and fact-checkers each doing their part. FIG. 08 redraws the structure he described. Moonshot's stated ambition is swarms of hundreds or a thousand agents finishing tasks fast enough "to produce real economical value". Every major lab is saying a version of the same thing. Google's Sundar Pichai told I/O 2026 "we're firmly in our agentic Gemini era". Microsoft's Satya Nadella told Build 2026 that "agents effectively are a new execution environment... a new paradigm", and described a security system where "over 100 specialized agents are working together to discover, debate and prove exploitable vulnerabilities end to end". Anthropic's Claude launches lead on "long-running agents" that keep checking their own work. When you read "agentic" in an announcement, this is the whole of what is meant: the software acts, in steps, with tools, increasingly in teams. 9.00 — THE WORDS OF THE RACE: OPEN WEIGHTS, FRONTIER, AGI AND THE REST The vocabulary of ambition, and what it is worth The last group of words describes not the machinery but the race itself, and these are the words that move markets. Open weights versus closed is the industry's deepest divide. An open-weight model's parameters can be downloaded and run by anyone, free: this is Kimi, DeepSeek and Alibaba's Qwen, and historically Meta's Llama. A closed or proprietary model can only be used through the maker's paid service: OpenAI's, Anthropic's and Google's flagships. Note what "open" does not include: usually not the training data or the full recipe, so open weights is not quite "open source" in the traditional software sense. The Kimi talk's framing, that open models are "closing the gap" with proprietary ones, is now conventional wisdom even at NVIDIA, whose chief executive told Axios in July 2026: "These Chinese models are excellent... Free AI should be great for chips." That the divide is strategy rather than ideology was proven in April 2026, when Meta, the company that built its AI reputation on giving Llama away, launched the first model from its Superintelligence Labs, Muse Spark, as a closed product, saying only that "we hope to open-source future versions of the model". The open champion went closed the moment it believed it had a lead; the challengers stay open because free distribution is how a challenger wins users from an incumbent. Then the ambition words. A frontier model is simply one of the handful of most capable models in the world at a given moment. AGI, artificial general intelligence, means an AI able to match humans across essentially all thinking work, and superintelligence means exceeding us. These are the words the money is raised on, and the industry's leaders now use them as near-term forecasts, on the record: Mark Zuckerberg in July 2025: "Developing superintelligence is now in sight." Sam Altman in June 2025: "Humanity is close to building digital superintelligence." Google DeepMind's Demis Hassabis on AGI, in 2026: "2030 is when I expect it to arrive, either plus or minus a year." Elon Musk went furthest in a January 2026 podcast, saying he expected AGI "next year". Alignment and safety name the work of keeping such systems doing what we intend; Anthropic in particular builds its brand on it. Three more words complete the set, all live in 2026. Distillation is training a cheap model on an expensive model's answers, the student copying the master's homework; in July 2026, Washington accused Moonshot of doing exactly this to American models, an accusation that is reported and denied, not proven. Sovereign AI is a country deciding it needs its own models, data and data centres rather than renting someone else's, the phrase behind national AI programmes worldwide, including the partnership between OpenAI and Australia's NextDC. And a wrapper is investor slang for a thin product built on someone else's model: worth remembering, because a company whose "AI product" is a wrapper owns a user interface, not the technology. 10.00 — FIVE QUESTIONS TO ASK OF ANY AI ANNOUNCEMENT The vocabulary, put to work Learning the words is only useful if it changes how you read. Here are five questions the vocabulary now lets you ask of any AI announcement, product launch or "AI strategy" slide. What is the unit? Tokens, parameters, users, gigawatts or dollars? Each unit flatters a different thing. A company quoting tokens is bragging about volume; parameters, about size; neither is revenue. Is that a usage number or a money number? "3.2 quadrillion tokens a month" measures work done, not dollars earned, and says nothing at all about profit. Who measured it? Vendor benchmarks are chosen by the vendor, from a rotating menu, after the fact. "Best on X" means best on the test they picked. Independent verification is rare and valuable. Which life of the model is the money for? Training (the build) or inference (the running)? Capital raised "for compute" can mean either, and they have completely different payback profiles. Who pays whom? Agents consume tokens; token sellers promote agents. Chipmakers invest in labs that buy chips. Follow the direction of the enthusiasm and check whether it matches the direction of the money. And now, the test. Here is the passage from Section 1.00 again: "Suppose you have 50 trillion high-quality tokens, and then you apply this new optimizer... all of a sudden you have a two times token efficiency... We introduce this new learning paradigm of agent swarms... at the end of the day, we're going to have a swarm of agents that each of them have a super long context." Read it once more. A company with a fixed supply of training text says its new learning recipe doubles what that text teaches its model; and its software now works as coordinated teams, each member able to hold enormous amounts of material in working memory. That is all it ever said. The ideas are ambitious; the language was the only barrier. The condensed decoder below covers the terms this piece taught. The full glossary, more than 45 terms in plain English, lives permanently on the Samso website, with the downloadable Samso AI Decoder PDF, and every future Samso piece on AI links back to it. References & sources This piece is built on one primary transcript (supplied) and the public statements, documents and pricing pages listed below. Quotes are reproduced verbatim from official transcripts or company pages except where noted as reported through a third-party transcript. Market-sensitive and dated figures (prices, context windows, token volumes) are quoted as at the dates shown and must be refreshed on publication day. No locator map appears because the subject is not geographic (house rule 4.2, waived). All figures are original Samso illustrations. Yang Zhilin (founder, Moonshot AI) — "How We Scaled Kimi K2.5", NVIDIA GTC, San Jose, 19 March 2026. Transcript supplied by Samso; session corroborated by Constellation Research, China Biz Insider and 36Kr coverage. Source of: the opening quotation; MuonClip optimiser and the two-times token efficiency claim; Kimi K2's one trillion parameters; Kimi Linear and the three-to-one attention mix; the agent swarm design and company analogy; K2.5's joint vision-and-text training claims. All capability claims are the company's own account. Jensen Huang (NVIDIA) — GTC 2026 keynote, 16 March 2026, official NVIDIA transcript ("Tokens are the building blocks of AI"; "your data center... is now a factory to generate tokens"; "three scaling laws"); CES 2026 keynote, January 2026, Rev.com transcript ("test-time scaling... another way of saying thinking"; "inference is now a thinking process"); interview with Axios, 22 July 2026 ("These Chinese models are excellent"; "Free AI should be great for chips"). Sundar Pichai (Google) — I/O 2026 opening keynote, 19 May 2026, official written version at blog.google. Source of: 9.7 trillion / ~480 trillion / ~3.2 quadrillion monthly tokens; "roughly 19 billion tokens per minute"; "we're firmly in our agentic Gemini era". Satya Nadella (Microsoft) — Build 2026 opening keynote, 2 June 2026, official Microsoft transcript. Source of: "tokens per dollar per watt"; "agents effectively are a new execution environment"; the 100-plus-agent security example. Sam Altman (OpenAI) — essays at blog.samaltman.com: "Three Observations" (February 2025; the log-of-resources scaling line), "The Gentle Singularity" (June 2025; "Humanity is close to building digital superintelligence"). Dario Amodei (Anthropic) — interview on the Dwarkesh Podcast, February 2026, quoted from the published transcript ("Humans don't see trillions of words..."); Anthropic, "Claude Sonnet 4.5" launch post, 29 September 2025 ("maintains focus for more than 30 hours"). Mark Zuckerberg / Meta — "Personal Superintelligence" letter, meta.com, 30 July 2025 ("Developing superintelligence is now in sight"); "Introducing Muse Spark", about.fb.com, 8 April 2026 ("we hope to open-source future versions of the model"). Demis Hassabis (Google DeepMind) — interview with Fast Company, 22 May 2026 (AGI "2030... plus or minus a year"); Yann LeCun — Davos panel remarks as reported by Fortune, 23 January 2026 ("completely LLM-pilled"; "digging the same trench"). Elon Musk (xAI) — post on X, 17 January 2026 (Colossus 2 "First Gigawatt training cluster in the world"); Moonshots podcast, January 2026, quoted from the published transcript (AGI "next year"). Podcast-transcript quotes to be re-verified against audio before publication. Pricing and specifications — official pricing/documentation pages of OpenAI, Anthropic, Google, xAI, Moonshot AI and DeepSeek, all fetched 26 July 2026. Known conflicts noted in the FIG. 04 and FIG. 07 captions (Claude context-window tiers; OpenAI's newest window unconfirmed on an official page; DeepSeek's launch-versus-current prices). Refresh all on publication day. OpenAI Help Center — "What are tokens and how to count them" (the 1,000-tokens-per-750-words rule). Scaling-law literature — Kaplan et al., "Scaling Laws for Neural Language Models" (2020); Hoffmann et al. ("Chinchilla"), "Training Compute-Optimal Large Language Models" (2022). Moonshot distillation accusation — reported by US press in July 2026 following the Kimi K3 release; an accusation, denied and unproven, and described as such in Section 9.00. Photographs (PLATE 1 to PLATE 5) — Unsplash, used under the Unsplash licence: Maximus Beaumont (taxi meter); Taylor Vick (data centre); Vishnu Mohanan (circuit board); Susan Q Yin (library); Frames For Your Heart (construction cranes). All retrieved 27 July 2026. Samso archive — "Tokens Are the Bill, Not the Value" (AI & Capital, Part 1, published 25 July 2026; source of the volume-versus-value framing revisited in Sections 3.00 and 10.00, and first publication of the Google token-volume figures redrawn here as FIG. 05) and "How to buy the AI build-out on the ASX" (AI & Capital, Part 2; source of the ASX vector-database finding cited in Section 7.00; link to be added on publication). The Samso Way – Seek the Research Here at Samso, we pride ourselves on delivering content for investors that is independent and informed by over three decades of experience in the industry. Our content is well-researched and is only created if I see merit in discussing the company's story. Our mission is simple: cut through the noise and spotlight what matters—genuine stories, grounded insights, and real opportunity. Our content is well-researched and is only created if the team sees merit in discussing the company or concept. Investors can explore our three core platforms: Coffee with Samso Samso Insights Samso News There may be numerous paths to success in investing, but the common thread among successful individuals is that they remain committed to making informed decisions. Equip yourself with the right knowledge and tools, and you will be well on your way to achieving your financial goals. Most importantly, investors need to be absolutely diligent in understanding their own risk-reward tolerance and capabilities. Never bite off more than you can chew. As they say, Rome wasn't built in a day, and the Great Wall stood because it took centuries to complete. The Samso Philosophy: Stay curious. Stay sharp. And remember—digging deeper always uncovers the real value. In Life, there is no such thing as a Free Lunch. Never bite off more than you can chew is my parting comment. Happy Investing, and the only four-letter word you need to know is DYOR. To support our independent work, please head over to our Support Page and give us a helping hand in any of the ways listed. This is a new initiative for the Samso Platform, and it was always the concept of Samso when we started this journey in 2018. Disclaimer The information or opinions provided herein do not constitute investment advice, an offer, or solicitation to subscribe for, purchase, or sell the investment product(s) mentioned herein. It does not take into consideration, nor have any regard to your specific investment objectives, financial situation, risk profile, tax position and particular, or unique needs and constraints. Read full Disclaimer. About Samso Samso is a trusted platform that equips dedicated investors with up-to-date industry knowledge and insights from top CEOs and thought leaders. By staying informed on business advancements and market trends, investors can enhance their financial decisions through a combination of expert guidance and their own research. Samso Insights | www.samso.com.au | An Investor Lens on ASX-Listed Companies

In the previous Samso Insights - Understanding Tungsten - Part 1- "Reading the Skarn — where tungsten hides in the contact", calcium was the hero: it trapped tungsten as scheelite and built the orebody. In this discussion, it becomes the villain. The same calcium leaves scheelite surrounded by look-alike minerals — calcite, fluorite, apatite — that are almost impossible to tell apart in a flotation cell. This is why a high-grade skarn can still be a hard business. Samso Insights Research Note Understanding Tungsten - Part 2 Geology Series 1.00 — WHY THE EFFORT IS WORTH IT Tungsten in sixty seconds Before the challenge of processing, comes the reward. Tungsten stands out in metallurgy due to its exceptional properties: it has the highest melting point of any metal and a density close to that of gold. This makes it a preferred choice for cutting tools, armor, and aerospace applications. Additionally, tungsten is rare, difficult to recycle, hard to replace, and approximately 80% of it is sourced from China. This is why the EU considers it a critical raw material. Figure 1: Tungsten's value sits in extreme physical properties and concentrated supply, not volume — global mine output is only ~82,000 t/yr. Source: Foucaud et al. (2020), citing USGS (2019) and Audion & Labbé (2012).

Roughly half the world's tungsten lives in skarn: a hard, ugly, calc-silicate rock that forms where a hot granite leans against limestone. Strip away the market noise and this is a story about chemistry at a contact. Here is how it works — in pictures. Samso Insights Research Insights Understanding Tungsten - Part 1 Geology Series 1.00 — THE SHORT VERSION A skarn is a chemical handshake between two rocks The theory : a rising granitic magma will naturally proceed to a cooling phase as it makes contact against a limestone, leading to the granite exhaling hot, metal-bearing fluid into a rock that is almost pure calcium carbonate (Figure 1). Where the two meet, the limestone is rebuilt into a coarse, dense rock packed with calcium-silicate minerals. That rebuilt rock is the skarn. When the fluid is carrying tungsten, the calcium it strips from the limestone locks the tungsten up as scheelite — calcium tungstate, CaWO₄. The short version of what a skarn is now explained. Just remember that a tungsten skarn needs four ingredients in the same place: a fertile granite, a carbonate host, heat, and a plumbing system to move fluid. Miss one and you get no ore. The diagrams below walk through each of them. Skarn A coarse-grained calc-silicate rock formed by the chemical replacement of carbonate rock (limestone or dolomite) near an igneous intrusion. The ore-grade, fluid-built variety is sometimes called tactite. Scheelite CaWO₄ — calcium tungstate, the dominant tungsten ore mineral in skarns. About 80% WO₃ by weight. Heavy (SG ≈ 6), pale, and famously glows blue under ultraviolet light. 2.00 — THE SETTING The recipe, in cross-section - the creation of a tungsten skarn This is the picture to carry around. A granite pluton rises into a layered sequence of sediments (Figure 1). The heat alone bakes the surrounding rock into hornfels. But where the carbonate layers sit against the granite — and especially where faults give fluid a path — the limestone is chewed out and replaced by skarn. Scheelite concentrates right at that reaction front. Figure 1: The tungsten-skarn system. Heat bakes a metamorphic aureole; fluid from the granite (blue arrows), guided by faults, replaces the carbonate host with calc-silicate skarn (orange). Scheelite concentrates at the reaction front. Schematic by Samso, after standard skarn models (USGS Open-File 02-195; Meinert et al.). Samso take Notice what does the work: it is the carbonate, not the granite, that makes a tungsten skarn. The limestone supplies the calcium that traps tungsten as scheelite. No carbonate at the contact, no skarn — which is exactly why explorers map the marble, not just the granite. 3.00 — TWO PROCESSES, ONE ROCK Baking versus rebuilding Two different things happen near the intrusion, and geologists keep them separate. Contact metamorphism is heat alone — the rock recrystallises but its chemistry barely changes. Metasomatism is the active one — fluid floods in and physically swaps elements in and out of the rock. Ore-grade skarn is built by metasomatism. Figure 2: Heat alone turns limestone into marble (left). Fluid-driven metasomatism rebuilds it into calc-silicate skarn and drops scheelite (right). Schematic by Samso, after Geology Science & ITIA descriptions. 4.00 — THE TWO ACTS Prograde, then retrograde — and tungsten moves twice Skarns build in two acts as the system heats then cools. The hot prograde stage grows anhydrous minerals — garnet and pyroxene — and drops the first, fine scheelite. As the system cools, the wetter retrograde stage overprints it with amphibole, mica and fluorite, and crucially re-dissolves and re-concentrates the scheelite into coarse, high-grade masses. The best ore is usually retrograde. Figure 3: Tungsten precipitates twice. Early prograde scheelite is fine and molybdenum-rich; cooling retrograde fluids re-dissolve it and re-deposit it as the coarse, high-grade ore that mines chase. Schematic by Samso, after Newberry & Einaudi (1981) and Kwak & Tan (1981), King Island. Why does the ore move twice? Because the two processes produce two distinct types of scheelites. The prograde process — characterized by heat and dryness — crystallizes garnet and pyroxene, depositing the first scheelite at the marble front. This scheelite is fine-grained, dispersed throughout the rock, and typically molybdenum-rich, as molybdenum easily integrates into scheelite formed from these early high-temperature fluids. The retrograde process is where the value lies. As the pluton cools and the fluid becomes water-rich and more acidic, hydrous minerals such as amphibole, mica, epidote, and fluorite overprint the anhydrous framework. The initial scheelite that isn't trapped within garnet dissolves and re-precipitates as coarse, molybdenum-poor masses, usually accompanied by sulfides. The USGS skarn model clearly states: early scheelite near the marble front is generally remobilized and redeposited in coarse-grained, high-grade masses within the zones of hydrous alteration. Three factors make retrograde ore highly desirable, all of which are commercial. Grade where the richest tungsten is typically found with the hydrous, retrograde minerals. In the Processing stage, the coarse scheelite separates cleanly and can be recovered through inexpensive gravity separation, whereas the fine prograde scheelite is the sub-10-micron material that eludes gravity (the issue addressed in Part 2). The third factor is quality. The molybdenum, which was a disadvantage in the early scheelite, is removed during remobilization, resulting in cleaner retrograde scheelite. THE KILBA TELL You can read the retrograde signature straight off the metallurgy. Research is showing that the Kilba deposit (ASX: TGN) — a skarn / calc-silicate system against the Kilba granite in WA's Ashburton carries its tungsten as coarse-grained scheelite that responds to conventional gravity separation, and 2015 testwork produced an extremely high-grade concentrate. Coarse, gravity-recoverable and high-grade is the fingerprint of retrograde-remobilised ore: Kilba sits at the good-ore end of the curve above. The same story, across Australia The model is played out in the Australian deposits. The textbook case is the King Island (Dolphin) mine, where a classic 1981 study traced the exact sequence . The early Mo-rich scheelite with garnet and pyroxene, then coarse Mo-poor scheelite and molybdenite during the retrograde overprint. Read Table 1 below as a context map for what "prograde, then retrograde" means on the Australian project. Table 1: Australian tungsten deposits with Prograde and Retrograde examples. that Sources: Kwak & Tan (1981), Economic Geology 76(2), King Island (Dolphin); USGS skarn model (OF 02-195); JORC resource/reserve figures from company disclosures — Tungsten Mining NL (Kilba, Mt Mulgine, Big Hill, Watershed), Group 6 Metals (Dolphin), Venture Minerals (Mount Lindsay), EQ Resources (Mt Carbine). The Mt Mulgine molybdenum interpretation is Samso's, built on the reported Mo grades. Deposit (ASX) Where Style Resource & grade The prograde → retrograde tell King Island / Dolphin (G6M) King Island, TAS Scheelite skarn (Grassy Granite) Dolphin ~3.0 Mt @ 0.73% WO₃ (probable reserve) Kwak & Tan (1981) traced early Mo-rich scheelite with garnet–pyroxene, then coarse Mo-poor scheelite + molybdenite in the retrograde overprint — the model deposit. Mt Mulgine (TGN) Murchison, WA Porphyry / vein-stockwork W–Mo (greenstone) ~72.2 Mt @ 0.18% WO₃ (~186–220 ppm Mo) The molybdenum credit is the prograde-scheelite signature; large, lower-grade disseminated system. Kilba Ashburton, WA Skarn / calc-silicate 7.2 Mt @ 0.19% WO₃ Coarse, gravity-amenable, very high-grade concentrate — the retrograde end. Big Hill Pilbara, WA Scheelite (calc-silicate) ~11.5 Mt @ 0.15% WO₃ Same skarn-style scheelite. Mount Lindsay (CRI) NW Tasmania Sn–W magnetite skarn ~45 Mt @ 0.2% Sn, 0.1% WO₃ Polymetallic skarn with the same prograde (magnetite-garnet-pyroxene) → retrograde architecture. Watershed (TGN) ~130 km N of Cairns, QLD Calc-silicate-hosted scheelite TGN development project Another calc-silicate / skarn-style scheelite system. Mt Carbine (EQR) Far North QLD Wolframite in quartz veins (greisen) Australia's primary W producer The counter-example — a vein / greisen deposit, not a skarn, so it does NOT follow the prograde → retrograde model. 5.00 — THE FINGERPRINT What a tungsten skarn is made of A skarn announces itself by its minerals apart from the tungsten minerals that we have discussed such as scheelite (Figure 5) and wolframite. The diagnostic pair is garnet plus pyroxene and in tungsten skarns garnet usually dominates. Add the retrograde overprint and a few accessories, and you have the assemblage a geologist logs at the drill rig (Figure 4). Figure 4: The skarn assemblage. Anhydrous garnet and pyroxene form first; a hydrous retrograde overprint carries the high-grade scheelite and the sulphides. Compiled by Samso from USGS, mindat.org and Meinert skarn classifications. Those legend entries aren't abstractions — name a line and a famous deposit shows it. China's Shizhuyuan (Hunan), a world-class W–Sn–Bi–Mo–F skarn, carries exactly the prograde set: garnet, pyroxene, vesuvianite and wollastonite, with scheelite and bismuthinite between the grains. Its ore, though, sits mostly in the retrograde overprint of fluorite, epidote, amphibole and magnetite. Korea's giant Sangdong deposit is the textbook in miniature — you can walk from an outer, low-grade pyroxene–garnet zone, through an amphibole–biotite zone, into a quartz–mica core where the grade is highest. Both say what Part 1's curve does: the rich ore tracks the hydrous, retrograde minerals. Figure 5: Scheelite, the orange-brown pseudo-octahedral crystals, on muscovite. Its high density (SG ≈ 6) is a field giveaway. Specimen located from Mt Xuebaoding, Pingwu County, Mianyang Prefecture, Sichuan Province, China. (Source: Wikimedia Commons ) The garnet-versus-pyroxene balance also flags the skarn's "flavour" (section 7.00). Garnet-dominant, andradite-rich systems like Shizhuyuan are oxidised and tend to carry more molybdenum in their scheelite; pyroxene-dominant, hedenbergite-and-pyrrhotite systems like Canada's Cantung and Mactung are reduced, and their highest grades again sit in the retrograde amphibole and biotite facies. The minerals you log are already telling you the deposit type, the molybdenum penalty, and where the best ore should be. Named deposits that show the fingerprint A spread of well-studied tungsten skarns — two from China, three from elsewhere — and the assemblages that define them: Table 2: Sources: Shizhuyuan — Lu et al. (2003, Economic Geology 98(5)); Yang/Newberry Shizhuyuan studies; Xianglushan — comparative scheelite geochemistry, Shizhuyuan vs Xianglushan (oxidising vs reducing fluids). Sangdong — Almonty Korea project geology; Seo et al. genetic model (prograde wollastonite–garnet–clinopyroxene → pyroxene–garnet; retrograde amphibole–biotite → quartz–mica). Cantung / Mactung — Elongo et al. (2020) and CanTung E-Zone studies (reduced grossular–hedenbergite–pyrrhotite; highest grade in hydrous retrograde facies). King Island — Kwak & Tan (1981). Framework: Newberry & Einaudi (1981); USGS OF 02-195. Deposit (country) Skarn type Prograde assemblage Retrograde / ore assemblage The tell Shizhuyuan (Hunan, China) Oxidised, garnet-rich (W–Sn–Bi–Mo–F) Garnet · pyroxene · vesuvianite · wollastonite (+ interstitial fluorite, scheelite, bismuthinite) Fluorite · epidote · amphibole · chlorite · magnetite + scheelite, wolframite, molybdenite Ore mostly in the retrograde skarn; higher-Mo scheelite from oxidising fluids. Xianglushan (Jiangxi, China) Reduced-type W skarn Garnet–pyroxene skarn + layered sulphide–scheelite Scheelite (lower-Mo), sulphides The reduced counterpart to Shizhuyuan — cleaner, lower-Mo scheelite. Sangdong (South Korea) Zoned giant W(–Mo) skarn Wollastonite–garnet–clinopyroxene → pyroxene–garnet (outer, low grade) Amphibole–biotite → quartz–mica core (highest grade); scheelite >95% of W, minor wolframite, molybdenite Grade climbs inward with the hydrous overprint (~0.3% → ~3% WO₃). Cantung / Mactung (Canada) Reduced (high-grade) Grossular garnet + hedenbergite pyroxene (+ abundant pyrrhotite) Amphibole + biotite skarn (highest grade); scheelite tracks pyrrhotite; chalcopyrite, bismuth Textbook reduced W skarn — ore in the hydrous retrograde facies. King Island / Dolphin (Australia) Garnet–pyroxene skarn (Grassy Granite) Garnet · pyroxene hornfels; early Mo-rich scheelite Coarse Mo-poor scheelite + molybdenite (retrograde amphibole) The Kwak & Tan (1981) model deposit (also in §4.00). 6.00 — ZONATION Distance from the granite is a thermometer Walk away from the pluton and the skarn changes in an orderly way, because the fluid cools as it travels. At King Island's Dolphin mine, scheelite-bearing fluids ran at roughly 500°C at the granite contact and cooled to about 300°C some 500 m out. Garnet sits proximal (andradite-rich nearest the contact), pyroxene takes over distally, and the richest ore tends to sit near the marble front where the chemistry tips over. Figure 6: Skarn zonation. Proximal garnet gives way to distal pyroxene and finally marble; the high-grade scheelite clusters near the reaction front. Temperatures from the King Island (Dolphin) study. After Kwak & Tan (1981), Economic Geology, King Island (Dolphin) zoning study. 7.00 — TWO FLAVOURS Reduced versus oxidised — and why it matters In 1981 Newberry and Einaudi split tungsten skarns into two families, and the split still frames how geologists think. Reduced skarns form deeper, in carbon-bearing host rocks, with hedenbergite-rich (ferrous) mineralogy and often higher grade. Oxidised skarns form shallower, in hematite-bearing hosts, with andradite-rich (ferric) mineralogy. The distinction predicts grade, by-products and metallurgy. Figure 7: The reduced/oxidised division of Newberry & Einaudi (1981), based on host-rock chemistry, iron oxidation state and depth. It is shorthand for what grade and metallurgy to expect. After Newberry & Einaudi (1981), via the EarthSci / USGS skarn literature. 7.50 — The Mo tell: molybdenum writes the conditions into the crystal Section 7 sorted skarns into reduced and oxidised by the company they keep. But there is a finer signal hiding inside the scheelite — its molybdenum content — and it records the fluid one pulse at a time. Here is the catch that trips people up, and it looks at first like it contradicts the table above. The reduced/oxidised split tells you whether a system carries molybdenum — and reduced tungsten skarns very often are W–Mo deposits. But whether that molybdenum ends up locked inside the scheelite is a different question, and the answer runs almost the opposite way. Molybdenum only slips into the scheelite lattice — substituting for tungsten — when the fluid is oxidising enough to carry it as Mo⁶⁺. Turn the fluid reducing and the molybdenum walks off to build its own mineral, molybdenite, leaving the scheelite clean. So a high-Mo scheelite is the fingerprint of an oxidising pulse, not a reduced one. Figure 7.5: Where molybdenum goes. Oxidising fluids load Mo⁶⁺ into the scheelite lattice; reducing fluids divert it into molybdenite and leave the scheelite low-Mo. The same deposit can record both as it cools — at Yaogangxian the molybdenum in scheelite falls roughly an order of magnitude from the early oxidising stage to the late reducing one, and both generations are ore. Schematic by Samso, after Hsu (1977) and Li et al. (2023), Ore Geology Reviews. You can watch this play out inside a single orebody. At Yaogangxian in South China, the early retrograde scheelite grew from an oxidising fluid and carries around 1,800 ppm molybdenum (Figure 7.5). The later sulphide-stage scheelite — grown as the system cooled and turned reducing, with pyrrhotite everywhere — carries roughly 240 ppm. Same deposit, same granite, an order of magnitude less Mo in the crystal, purely because the redox shifted. And crucially, both generations are ore. The redox state changed the flavour of the scheelite without ever stopping it forming — which is the quiet lesson that oxygen state is not the master switch on whether tungsten drops out, just on what comes with it. Samso take This is where the geology reaches the bank. Molybdenum in scheelite is a penalty element in a tungsten concentrate — APT buyers want it low — so an oxidised, high-Mo orebody forces a choice: strip the Mo out, or sell it as a molybdenite credit. Western Australia's Mt Mulgine is the local case in point: its scheelite carries molybdenum, which is exactly why the flowsheet there is built to pull two products rather than one. Read the Mo number in the rock and you are reading a processing-and-pricing decision that is fixed long before the plant is designed. So is lattice Mo a bad thing for getting the tungsten out? It depends entirely on the form, and this is the crux. Molybdenum sitting as its own mineral, which is molybdenite is the easy case. Molybdenite floats off cleanly in processing and can even be sold separately. Molybdenum dissolved into the scheelite lattice is the hard case. Because it is built into the crystal, no amount of gravity, magnetic or flotation work can reject it — it rides with the tungsten straight into the concentrate. It does not lower how much tungsten you recover, but it contaminates the concentrate with a penalty element, forcing an extra chemical step downstream: the Mo dissolves alongside the tungsten during alkaline digestion and then has to be precipitated back out as a sulphide to reach APT specification. So a high-Mo scheelite is genuinely the awkward one — you carry a de-molybdenisation cost you cannot avoid with physics, and because the Mo is dispersed in the lattice rather than present as discrete molybdenite, you cannot bank it as a by-product credit either. The severity tracks the grade: a few hundred ppm is usually manageable, while a properly powellitic (Mo-rich) scheelite becomes a real concentrate problem. 8.00 — THE EXPLORER'S TRICK Why scheelite glows — and how it gets found Scheelite has a gift for explorers: under short-wave ultraviolet light it fluoresces a bright sky-blue. Swap some tungsten for molybdenum and the glow shifts toward cream-yellow — so the colour itself hints at composition. Geologists have mapped skarns at night with UV lamps for a century, and the trick still re-opens old mines by lighting up ore left in the walls. Figure 8: The same rock, two lights. Pure scheelite glows blue under short-wave UV; molybdenum shifts it to yellow. This single property turns a UV torch into a tungsten exploration tool. Schematic by Samso; fluorescence behaviour per Wikipedia (Scheelite), mindat.org and Geology.com. In my previous life working on a tungsten project in New Zealand, the identification of scheelite was indeed a sight to see (Figure 8.5). In my opinion, the Kirwan Hill project still remains one of the untapped tungsten occurrences that could be a game changer if the project was worked on in 2026. Figure 8.5: A scheelite core sample form the Kirwan Hill Tungsten Project, New Zealand. The sample shows scheelite in brecciated unit and occurs as coarse grains as well as in a disseminated form. Samso take Fluorescence is not a party trick — it is a low-cost mapping vector. A handheld UV lamp run over outcrop, drill core or mine walls converts an invisible problem into a visible one. For a small-cap explorer, that is meaningful: cheap nighttime traverses can rank targets before a single expensive metre is drilled. 9.00 — KNOW YOUR ORE MINERAL Scheelite (skarns) vs wolframite (veins) Tungsten comes in two ore minerals, and which one you have tells you a lot about the deposit. Skarns and their carbonate-bound calcium overwhelmingly carry scheelite. The other major ore, wolframite, belongs more to granite-hosted veins and greisens. They look, behave and process differently. Figure 9: The two tungsten ore minerals. Scheelite dominates skarns and gives up its calcium-tungstate identity under UV; dark wolframite belongs to the vein/greisen world. Compiled by Samso from ITIA, mindat.org and mineralexpert.org. The main importance for investors is to understand that Wolframite is the Fe cousin of Scheelite who is a Calcium variant. Discovering these tungsten minerals separately is ok, however, if they are associated in the same ore body, it will create issues in the processing part of the business. Wolframite will be magnetic and as thee ore bodies are commonly associated with magnetic minerals, the separation using a magnetic separator will be hampered. These are some aspects that investors need to keep in mind. IN part 2 of the Understanding Tungsten series, we will introduce the issues for Scheelite as the calcium variant of a tungsten mineral. 10.00 — IN PRACTICE, AT HOME The Australian skarn: King Island and beyond Australia's exemplary tungsten skarn is Dolphin, located at Grassy on King Island. It features scheelite in garnet-pyroxene skarn surrounding the Devonian-Carboniferous Grassy Granite, discovered in the early 1900s and intermittently mined for a century. It exemplifies the classic characteristics: carbonate host, granite contact, fault control, and retrograde high-grade ore. In Western Australia, Mt Mulgine (Tungsten Mining NL, ASX: TGN) contains scheelite with molybdenum, highlighting that the predicted Mo content in the geology becomes a significant factor in processing and pricing downstream. Figure 10: Selected Australian tungsten settings. Skarn-hosted scheelite (Dolphin, Mt Mulgine) contrasts with vein/greisen wolframite (Mt Carbine). Positions indicative only. Compiled by Samso from Group 6 Metals, Tungsten Mining NL and Geoscience Australia disclosures. But Dolphin is neither the biggest nor the only Australian skarn. The country's largest tungsten resource is O'Callaghans (Greatland Resources, ASX: GGP), about 10 km south of Telfer in WA — a flat-lying polymetallic skarn where scheelite- and wolframite-bearing amphibole–pyroxene–garnet rock replaces limestone above the reduced O'Callaghans granite. Its December 2025 resource is ~70 Mt at 0.35% WO₃ with copper, zinc and lead by-products, and the proposed flowsheet is pure Part 2: flotation for the base metals, then magnetic separation and gravity to recover the tungsten. Two more wear the classic fingerprint. Kilba (TGN, Ashburton WA) is scheelite in skarn / calc-silicate against the Kilba granite — and, as section 4 noted, its coarse, gravity-amenable scheelite is a textbook retrograde signature. Molyhil (Thor Mining, ASX: THR), 220 km north-east of Alice Springs, is a scheelite–molybdenite–magnetite skarn — a small but genuine W–Mo skarn that wears its molybdenum on its sleeve. Is Watershed a skarn? Not quite — and that's the point The most useful example is the awkward one. Watershed (TGN, 130 km north of Cairns) sits in the Hodgkinson Formation — a pile of deformed metasediments (conglomerate, psammite, slate), not a clean limestone. It carries genuine skarn minerals (garnet, clinopyroxene, actinolite), so the rock is calc-silicate / skarn-altered; but the scheelite was largely emplaced in shear-controlled veins and then upgraded during later metamorphism. The published work describes it as a "metamorphic tungsten upgrade after a magmatic-hydrothermal event" — a hybrid that sits between a contact skarn and a metamorphic / vein deposit. Its resource is large and low-grade (~49–70 Mt at ~0.11–0.14% WO₃). Cookes Creek is a similar matter of degree. It isn't a single named skarn but a granite-and-workings cluster in WA's Gascoyne; the defined resource on its margin — Big Hill (~11.5 Mt at 0.15% WO₃) is vein-hosted scheelite within a tremolite-rich calc-silicate unit: skarn-affiliated, but vein-controlled rather than a classic replacement body. So I would call this a hybrid and not a yes/no in regards to being a skarn deposit. Table 3: Australian Tungsten Deposits nd their relationship to being a skarn mineralisation. Sources: O'Callaghans — Greatland Resources (ASX: GGP) December 2025 Mineral Resource (~70 Mt @ 0.35% WO₃) and Newcrest skarn geology; Dolphin — Group 6 Metals / Kwak & Tan (1981); Kilba, Big Hill, Watershed, Mt Mulgine — Tungsten Mining NL disclosures; Molyhil — Thor Mining (ASX: THR) / NT Geological Survey; Mount Lindsay — Venture Minerals (ASX: VMS); Mt Carbine — EQ Resources (ASX: EQR). Watershed geology — Hodgkinson Formation / Mossman Orogen studies (the deposit is described as a metamorphic tungsten upgrade of a Carboniferous magmatic-hydrothermal system). Deposit (owner) Setting Skarn? O'Callaghans (Greatland, GGP) Polymetallic W–Cu–Zn–Pb replacement of limestone above the O'Callaghans granite, Telfer WA ✓ Classic skarn — largest in Australia Dolphin / King Island (G6M) Scheelite in garnet–pyroxene skarn at the Grassy Granite contact, TAS ✓ Classic skarn Kilba Scheelite in skarn / calc-silicate against the Kilba granite, Ashburton WA ✓ Skarn / calc-silicate Molyhil (Guardian Metals) Scheelite–molybdenite–magnetite skarn, Aileron Province NT ✓ Skarn (W–Mo) Mount Lindsay (Critica Limited) Sn–W magnetite skarn, NW Tasmania ✓ Skarn (Sn–W) Watershed (TGN) Scheelite in calc-silicate / skarn-altered metasediments + shear veins, Hodgkinson Fm, FNQ ~ Skarn-altered — not a classic carbonate skarn Big Hill / Cookes Creek Vein scheelite in a tremolite-rich calc-silicate unit on the Cookes Creek granite margin, WA ~ Calc-silicate-hosted, vein-controlled Mt Mulgine (TGN) Porphyry / vein-stockwork W–Mo in greenstone + greisen, Murchison WA ✗ Not a skarn (porphyry / vein) Mt Carbine (EQR) Wolframite in quartz veins / greisen, Far North QLD ✗ Not a skarn (vein / greisen) — SAMSO CONCLUDING COMMENTS The geology is the moat Like all mineral deposits, when it comes to classificating a deposit style type to define deposits as occurring the same way, I find that no single deposits ever sits in the same framework that an institution demands. My experience has shown me that a broad framework of conditions will work but when a discussion starts, everything seems to just falls apart and the final conclusion always seem to fall back to individuality rather than a group classification. Hence if we strip a tungsten skarn back and it is almost elegant in a Type case of a skarn tungsten deposit. It is a granite heat source, source rock, that breathes fluid into a limestone, the limestone gives up its calcium, and tungsten is locked away as scheelite at the reaction front. I think for reader, it is away to remember all the factors we have discussed in this Samso Insight, but I would not hold these points as a hard and fast rule to understanding skarn tungsten deposits. Everything an investor cares about downstream, grade, by-product molybdenum, how hard the ore is to process, whether a UV torch can cheaply find more is written into that geological setup before any mining decision is made. That is the reason Samso keeps coming back to the rocks and the process that creates the minerals and hence, the metal abundances. In a critical-minerals cycle full of noise, the contact between a granite and a carbonate is one of the few things that does not move with the headlines. Understand the skarn, and the tungsten story stops being a ticker and starts being a deposit. References & sources U.S. Geological Survey — Environmental Geochemistry of Skarn and Polymetallic Deposits, Open-File Report 02-195 (Chapter H). Skarn classification and W-skarn behaviour. Newberry, R.J. & Einaudi, M.T. (1981) — reduced vs oxidised tungsten skarn division; summarised in the EarthSci.org / USGS skarn literature. Meinert, L.D. et al. — World skarn deposits classification (resource split ~48% skarn). Via tungsten beneficiation review literature. Kwak, T.A.P. & Tan, T.H. (1981) — "The geochemistry of zoning in skarn minerals at the King Island (Dolphin) Mine," Economic Geology 76(2): 468–497. Bowman, J.R. et al. (1985) — CanTung E-Zone scheelite skarn isotope study, Economic Geology 80(7): 1872–1895. Geology Science — "Skarn deposits" overview; "Scheelite: properties, formation, occurrence." Wikipedia — Scheelite (CaWO₄, fluorescence, occurrence); cross-checked with Britannica and mindat.org (min-3560). Group 6 Metals Limited (ASX: G6M) — Dolphin Tungsten Project overview & exploration pages. Mineral Resources Tasmania — King Island geology documents; Callaghan (Resource & Exploration Geology) Dolphin/Bold Head report. International Tungsten Industry Association (ITIA); mineralexpert.org — scheelite vs wolframite mineralogy & WO₃ contents. Tungsten Mining NL (ASX: TGN) — Mt Mulgine (scheelite + molybdenum) disclosures. Hsu, L.C. (1977) — “Effects of oxygen and sulfur fugacities on the scheelite–tungstenite and powellite–molybdenite stability relations,” Economic Geology 72: 664–670. Li, W.-S., Ni, P., Pan, J.-Y., et al. (2023) — “The genetic association between vein and skarn type tungsten mineralization in the Yaogangxian tungsten deposit, South China,” Ore Geology Reviews 159: 105544 (open access, CC BY-NC-ND). The Samso Way – Seek the Research Here at Samso, we pride ourselves on delivering content for investors that is independent and informed by over three decades of experience in the industry. Our content is well-researched and is only created if I see merit in discussing the company's story. Our mission is simple: cut through the noise and spotlight what matters—genuine stories, grounded insights, and real opportunity. Our content is well-researched and is only created if the team sees merit in discussing the company or concept. Investors can explore our three core platforms: Coffee with Samso Samso Insights Samso News There may be numerous paths to success in investing, but the common thread among successful individuals is that they remain committed to making informed decisions. Equip yourself with the right knowledge and tools, and you will be well on your way to achieving your financial goals. Most importantly, investors need to be absolutely diligent in understanding their own risk-reward tolerance and capabilities. Never bite off more than you can chew. As they say, Rome wasn't built in a day, and the Great Wall stood because it took centuries to complete. The Samso Philosophy: Stay curious. Stay sharp. And remember—digging deeper always uncovers the real value. In Life, there is no such thing as a Free Lunch. Never bite off more than you can chew is my parting comment. Happy Investing, and the only four-letter word you need to know is DYOR. To support our independent work, please head over to our Support Page and give us a helping hand in any of the ways listed. This is a new initiative for the Samso Platform, and it was always the concept of Samso when we started this journey in 2018. Disclaimer The information or opinions provided herein do not constitute investment advice, an offer, or solicitation to subscribe for, purchase, or sell the investment product(s) mentioned herein. It does not take into consideration, nor have any regard to your specific investment objectives, financial situation, risk profile, tax position and particular, or unique needs and constraints. Read full Disclaimer. About Samso Samso is a trusted platform that equips dedicated investors with up-to-date industry knowledge and insights from top CEOs and thought leaders. By staying informed on business advancements and market trends, investors can enhance their financial decisions through a combination of expert guidance and their own research. Samso Insights | www.samso.com.au | An Investor Lens on ASX-Listed Companies

Cheaper by the unit, dearer by the invoice Start with tokens. Everything in this note runs on them. A token is a small chunk of text, roughly a short word or part of a word. It is the basic unit an AI model reads and writes. Tokens count for three reasons. With that in hand, here is the puzzle that frames everything else: the price of that raw input has collapsed, and yet company spending on AI has gone up, not down. On the price, the direction is not in doubt. Stanford University's 2025 AI Index reported that the cost of running a model at the level of GPT-3.5 fell from about US$20 per million tokens to about US$0.07 through 2024, a fall of more than 99 per cent. Over the same broad period, the venture capital firm Menlo Ventures estimated that enterprise spending on large language models (the kind of AI behind chat assistants and coding tools) tripled in the twelve months to the end of 2025. The price per unit fell through the floor; the total bill still multiplied. FIG. 01 sets the two lines side by side. The instinct is to call this waste. That instinct is worth resisting. A falling unit price alongside rising total spending is the normal signature of a technology being adopted quickly, not proof that money is being burned. Cars became cheaper to run per kilometre over the twentieth century, and total spending on motoring rose all the same because people drove far more. The real question is not whether AI spending is rising. It is whether each dollar of it is buying something worth having. Holding that question in mind is the whole discipline this note is about. The Supply Side: What a Quadrillion Tokens Looks Like The scale the bill is paying for If the first section is the demand side, the clearest picture of the supply side came from Google's I/O 2026 developer conference. The numbers are difficult to hold in your head. Google reported that the volume of tokens processed across its products went from 9.7 trillion a month two years ago, to about 480 trillion a month a year ago, to about 3.2 quadrillion a month now. A quadrillion is a thousand trillion, so the latest figure is roughly seven times the level of a year earlier. FIG. 02 shows the shape of that climb. The line barely lifts off the floor for most of the period, then goes near vertical, which is the visual an investor should associate with the current phase of AI adoption. The demand behind that curve is not only consumer. A companion figure Google showed, redrawn here as FIG. 03, tracks tokens processed through its model APIs, the pipes that developers and businesses build on. That volume rose from about 7 billion tokens per minute in September 2025 to about 19 billion per minute by April 2026, roughly six times higher year on year. This is the line to watch most closely for investors, because it is the closest proxy for businesses actually wiring AI into their operations, which is where the enterprise bill from Section 1 comes from. That demand rides on enormous distribution. Google said it now runs 13 products with more than a billion users each, five of them with more than three billion users (Search, Gmail, Android, Chrome, and YouTube). Its Gemini AI app alone roughly doubled its monthly users in a year, from about 400 million to more than 900 million. To serve it, the company's capital expenditure, the money spent building data centres and buying chips, is guided to rise about six-fold in four years, as FIG. 04 shows. That capital line is the bill from Section 1, seen from the side of the company sending the invoices. The tokens counted so far are produced inside physical facilities like the one in PLATE 1: powered, cooled halls filled with computers. Where the Money Actually Goes Why a cheap token still adds up to a fortune If tokens are so cheap, how does the bill get so large? McKinsey's answer is that the cost of AI has moved out of the price per token and into the way the work is done, especially once the software starts acting on its own. The relevant shift is from a model that answers a single question to an AI agent: software that chains many AI steps together, uses tools, and checks its own work to finish a task without a human at each step. That autonomy is where the money goes. McKinsey set out several reasons, summarised in FIG. 05. In plain terms: an agent re-sends its full context to the model again and again as it works; much of what it does is checking and re-doing its own output; its freedom to choose a path means the same job can cost very different amounts on different runs; powerful, expensive models get used on trivial steps; coordinating several agents multiplies the traffic; and messy data or long-winded prompts inflate the count before any useful work begins. Google's own keynote gave the perfect worked example, shown in FIG. 06. Its engineers set an agent system loose to build a working computer operating system from an empty project. Over 12 hours, 93 sub-agents made more than 15,000 model requests and processed 2.6 billion tokens, and produced the core of a functioning operating system for under US$1,000 of usage. Read that two ways at once. It is astonishingly cheap for work that would take skilled people many months. It is also 2.6 billion tokens for a single task, and multiplied across millions of tasks that is exactly how you get to a US$180 billion capital bill. Tokens Are the Bill, Not the Value The one line to carry through every AI announcement Here is the idea the title promises, stated plainly. Tokens, models, and dollars spent are inputs. They are the bill. Whether any of it produced something worth having is a separate question, and it is the only one that counts. McKinsey's phrase for the mistake is judging AI by its inputs instead of its outcomes. The number they argue companies should track is cost per outcome: the cost of a finished, useful result, a resolved customer complaint, a cleared invoice, a piece of code shipped, rather than the cost of the tokens burned along the way. FIG. 07 draws the distinction. Most organisations cannot yet produce that number. In a McKinsey survey on AI financial operations from May 2026, 93 per cent of respondents said they had exceeded their AI budgets, and in a separate McKinsey survey about one in five said their organisation had limited its use of AI because of running costs. Those figures sound alarming, and they need their counterweight. Budgets set at the dawn of a new technology are guesses, so overshooting them is not by itself evidence of failure. It is evidence that very few organisations yet measure the thing that would tell them whether the spend was worth it. That gap, between spending confidently and measuring plainly, is the gap an investor can look through. Six Questions Before You Believe an AI Story Turning the framework into investor diligence McKinsey wrote its recommendations for chief executives. Turned around, they become a short checklist for anyone deciding whether a company's AI story deserves to move the share price. None of these need technical knowledge to ask. What did the spend buy? Ask for the outcome in cost-per-outcome terms, not tokens, licenses, or headcount "freed up". A company that can answer this is rare and worth noting. Where does AI actually change the economics? A serious operator can name the few parts of its business where machine work alters the cost or quality in a way that counts, rather than claiming AI everywhere. Who owns it, and what are they measured on? Look for clear accountability for AI spending and results, with real targets, not a vague "innovation" mandate. Is the spend concentrated or sprayed? Heavy AI use tends to cluster in a few high-value places. At McKinsey's own firm, about 10 per cent of users drove roughly 65 per cent of token use. Concentration can be healthy focus or hidden dependence; the point is that management should know which. Is the tool matched to the task? Using the most powerful, most expensive model for everything is a red flag. McKinsey found that simply encouraging shorter prompts and answers cut token use by 30 to 40 per cent in some workflows without materially hurting quality. Can they separate the bill from the value? If management cannot tell you, in plain numbers, what the AI cost and what it returned, you are being asked to take the value on faith. One caution keeps these questions fair. Most companies cannot yet answer most of them, so a string of "no" answers is common today and not automatically damning. The tools to measure this are themselves new. The signal to pay for is a company that answers them well. There is also a subtler point from McKinsey worth carrying: as AI makes clever workflows easy to copy, a slick "we use AI" process is a weaker moat than it looks. Proprietary data a rival cannot copy and the discipline to run AI cheaply become the real, and quieter, advantages. A company that never headlines AI may hold more of it than one that never stops talking about it. References & Sources This note draws on one primary article and one primary event transcript, listed below. The AI Index, Menlo Ventures, and arXiv figures reach us as reported within the McKinsey article; where a reader wants them stood behind their original sources, those originals should be consulted directly. All figures shown in the visuals are original Samso illustrations of the data named in each caption. Market-sensitive and time-stamped figures (token volumes, survey dates, capital-expenditure guidance) are quoted as at the dates shown and should be refreshed on publication day. Photographs labelled PLATE are reproduced from a company's public ASX release, with attribution in the caption, and are distinct from the original Samso illustrations labelled FIG. McKinsey & Company (QuantumBlack) — "Is that AI agent worth it? Agentic economics and the modern operating model" (2026). Source of: the token price fall (US$20 to US$0.07 per million tokens through 2024, attributed to the Stanford HAI 2025 AI Index); enterprise LLM spending tripling to end 2025 (attributed to Menlo Ventures); the six cost drivers; the ~1,000× token multiple, ~60 per cent refinement share, and factor-of-30 variation (attributed to arXiv research); the "cost per outcome" concept; 93 per cent exceeding AI budgets (McKinsey Enterprise AI FinOps survey, May 2026); about one-fifth constraining AI use on cost (McKinsey State of AI survey); the roughly 10 per cent of users driving about 65 per cent of consumption; and the 30 to 40 per cent saving from more concise prompts. Google I/O 2026 keynote — transcript supplied by Samso, with the "Monthly Tokens Processed" and "Tokens Per Minute Processed" slides. Source of: 9.7 trillion, ~480 trillion, and ~3.2 quadrillion monthly tokens; ~19 billion tokens per minute across model APIs; 13 products with 1 billion-plus users and five with 3 billion-plus; the Gemini app rising from ~400 million to more than 900 million monthly users; capital expenditure of ~US$31 billion (2022) rising to ~US$180 billion to US$190 billion (2026); and the agent demonstration (93 sub-agents, 15,000-plus model requests, 2.6 billion tokens, under US$1,000, 12 hours). The 3.2 quadrillion monthly-token figure is corroborated by third-party coverage of I/O 2026 (e.g. Crypto Briefing; Thurrott; Google's own I/O 2026 blog posts). DigiCo Infrastructure REIT (ASX: DGT) — 1H FY26 Results Presentation (20 February 2026): source of the data-hall photograph reproduced in PLATE 1, with attribution. The Samso Way – Seek the Research Here at Samso, we pride ourselves on delivering content for investors that is independent and informed by over three decades of experience in the industry. Our content is well-researched and is only created if we see merit in discussing the company's story. Our mission is simple: cut through the noise and spotlight what matters—genuine stories, grounded insights, and real opportunity. Our content is well-researched and is only created if the team sees merit in discussing the company or concept. Investors can explore our three core platforms: Coffee with Samso Samso Insights Samso News There may be numerous paths to success in investing, but the common thread among successful individuals is that they remain committed to making informed decisions. Equip yourself with the right knowledge and tools, and you will be well on your way to achieving your financial goals. Most importantly, investors need to be absolutely diligent in understanding their own risk-reward tolerance and capabilities. Never bite off more than you can chew. As they say, Rome wasn’t built in a day, and the Great Wall stood because it took centuries to complete. The Samso Philosophy: Stay curious. Stay sharp. And remember—digging deeper always uncovers the real value. In Life, there is no such thing as a Free Lunch. Never bite off more than you can chew is my parting comment. Happy Investing, and the only four-letter word you need to know is DYOR. To support our independent work, please head over to our Support Page and give us a helping hand in any of the ways listed. This is a new initiative for the Samso Platform, and it was always the concept of Samso when we started this journey in 2018. Disclaimer The information or opinions provided herein do not constitute investment advice, an offer, or solicitation to subscribe for, purchase, or sell the investment product(s) mentioned herein. It does not take into consideration, nor have any regard to your specific investment objectives, financial situation, risk profile, tax position and particular, or unique needs and constraints. Read full Disclaimer. About Samso Samso is a trusted platform that equips dedicated investors with up-to-date industry knowledge and insights from top CEOs and thought leaders. By staying informed on business advancements and market trends, investors can enhance their financial decisions through a combination of expert guidance and their own research. Samso Insights | www.samso.com.au | An Investor Lens on ASX-Listed Companies

Tusker Minerals Ltd (ASX: TSK) has entered a binding agreement with AuKing Mining Limited (ASX: AKN) for the proposed sale of 100% of the Machinga Rare Earth Elements Project in southern Malawi. The project comprises two exclusive prospecting licences, EPL 0529 and EPL 0705, held by Green Exploration Limited, a wholly owned Malawian subsidiary of Tusker. The total consideration is stated as up to A$4 million. Tusker frames the proceeds as non-dilutive funding, meaning money that comes in without the company issuing new Tusker shares and diluting existing holders. The stated purpose is to redirect capital to the rutile and heavy mineral sands portfolio in Cameroon and Malawi. The Vocabulary, in Plain English Rare earth elements (REE) are a group of seventeen metals used in permanent magnets, electronics and defence hardware. TREO stands for Total Rare Earth Oxide, the standard way of expressing rare earth grade. Heavy rare earth elements (HREE) are the scarcer subset, including dysprosium, terbium and yttrium. They command far higher prices than the light rare earths because supply is concentrated in very few places. Rutile is a naturally occurring titanium dioxide mineral. It is the premium feedstock for titanium pigment and titanium metal, and it commands a higher price than ilmenite, the more common titanium mineral. Heavy mineral sands (HMS) deposits contain rutile, ilmenite and zircon in loose sand or weathered material, which usually makes them cheap to mine. An exclusive prospecting licence (EPL) is the Malawian equivalent of an Australian exploration licence. It gives the holder the exclusive right to explore a defined area. Transferring one to a new owner requires regulatory approval. A JORC-compliant inferred Mineral Resource is the lowest confidence category of formally estimated tonnage and grade under the Australian reporting code. Inferred means the estimate is based on limited sampling and geological inference. It is the first formal resource milestone an explorer reaches. A cut-off grade is the minimum grade at which material is counted in a resource estimate. A resource of 10Mt at 0.65% TREO applying a 0.5% TREO cut-off means only material grading above 0.5% TREO is included in that tonnage. Performance shares are securities that convert into ordinary shares only if a defined milestone is achieved. Until then they carry no value and no dividend. VWAP is the volume weighted average price, an average share price over a stated period weighted by the volume traded at each price. It is used so that a single day's trading cannot set the value of a share issue. Voluntary escrow is an agreement not to sell shares for a stated period. Escrowed shares cannot be turned into cash while the restriction applies. Non-dilutive funding means funding raised without issuing new shares in the company receiving it. It says nothing about how quickly that funding arrives, or in what form. The Terms, Broken Into Four Pieces The announcement sets out the consideration in four parts. Splitting them by certainty and by timing is the most useful thing an investor can do with this document (Table 1). Table 1: Machinga consideration, sorted by form, timing and certainty. Component Amount Form Timing Certainty Cash at completion A$750,000 Cash At completion Conditional on completion AuKing ordinary shares A$750,000 (30,000,000 shares) Equity, 12-month escrow At completion Conditional on completion and any AKN shareholder approval Deferred cash A$1,250,000 Cash 12 months after completion Conditional on completion AuKing performance shares A$1,250,000 (50,000,000 performance shares) Contingent equity Within 3 years, on a resource milestone Contingent on exploration success Total A$4,000,000 Three observations follow. First, the cash at completion is A$750,000. That is a little under a fifth of the headline. A further A$1.25 million in cash follows twelve months later. Both are real, and together they make up the cash half of the deal. They simply arrive at different times, which is worth keeping in mind for a company planning exploration spend over the next year. Second, the equity is held rather than spent. The 30,000,000 AuKing ordinary shares carry a 12-month voluntary escrow, so they cannot be sold during that period. That is a normal feature of a deal like this, and it also signals that Tusker intends to stay a holder rather than an immediate seller. The practical effect is that of the A$2.75 million not contingent on exploration success, A$750,000 is available to fund work in the next twelve months. Third, the last A$1.25 million is tied to exploration success. The 50,000,000 performance shares convert on publication, within three years of their issue, of a JORC-compliant inferred Mineral Resource for Machinga of at least 10Mt at 0.65% TREO, applying a 0.5% TREO cut-off. Machinga has no resource today, so reaching that milestone requires AuKing to fund and complete a drilling program, assay it, model it and publish it inside the three-year window. That is a normal sequence of work for an explorer, and it is also the part of the consideration with the widest range of outcomes. Staged and contingent consideration is standard practice when an explorer sells an undrilled project to another explorer, and there is nothing unusual in this structure. The headline is accurate. It is just worth reading alongside the detail, because "up to A$4M" and "A$2.75 million payable on completion, of which A$750,000 is cash on the day" describe the same transaction from different ends. The Performance Share Cap Runs One Way On satisfaction of the resource hurdle, AuKing will issue ordinary shares with a value of A$1,250,000, calculated using the 90-day VWAP of AuKing shares at the date the hurdle is satisfied, subject to a maximum of 50,000,000 ordinary shares. Divide A$1,250,000 by 50,000,000, and you get A$0.025 per share. That is Samso's arithmetic, and it sets the point at which the cap starts to matter. If AuKing's 90-day VWAP is above A$0.025 when the hurdle is met, Tusker receives fewer shares but the full A$1.25 million of value. If AuKing's VWAP is below A$0.025, the cap engages, Tusker receives the maximum 50,000,000 shares, and those shares are worth less than A$1.25 million. The same A$0.025 reference sits behind the completion equity. The deal values 30,000,000 AuKing shares at A$750,000, which is also A$0.025 per share. Same Buyer, Same Country, Same Regulator This is Tusker's second sale to AuKing in four months. The Tundulu Rare Earth Project was sold under a binding agreement announced on 17 April 2026 for a total consideration of up to A$5.55 million. This one was with Auking too. The Tundulu experience is directly relevant to how investors should read the Machinga timetable, and it is the piece of context the announcement does not supply. In early June 2026, AuKing had lodged all documentation required for the transfer of the Tundulu exploration licence to its subsidiary with Malawi's Mining and Minerals Regulatory Authority, and the transfer remained under review. To keep exploration moving in the meantime, AuKing and Tusker entered an earn-in agreement in early June 2026, giving AuKing contractual authority to explore the licence and an alternative pathway to acquire 100% of it if the formal transfer process were delayed further. The consideration for Tundulu was paid, and AuKing has been drilling. The licence transfer, which is the legal core of the transaction, took longer than the parties planned and needed a contractual workaround. Machinga is subject to approval from the same regulator, for two licences rather than one. The word "binding" in the headline refers to the agreement between the two companies. It does not bind the Malawian mining authority to anything. The Strategic Case, and the Question It Raises The rationale Tusker gives is coherent. The company holds district-scale rutile and heavy mineral sands ground in Cameroon and Malawi; it believes that is where near-term value creation sits, and it does not want to fund rare earth exploration at the same time. Selling to a dedicated owner lets Machinga be advanced by somebody whose primary business it now is, at no further cost to Tusker. The rutile portfolio has been producing results. In July 2026 the company reported a JORC exploration target at the Diwong South deposit within the Douala Basin project in Cameroon, and in March 2026 it reported high-grade rutile from reconnaissance sampling at Mzimba in northern Malawi. Samso covered the Mzimba result in March. Those figures come from the relevant company releases and secondary reporting, not from the announcement under discussion here. "This proposed transaction builds on the successful divestment of our Tundulu project and reflects our disciplined approach to capital allocation and portfolio optimisation. It enables Tusker to prioritise advancement of its high-grade rutile and heavy mineral sands assets in Cameroon and Malawi, where we see the strongest near-term value creation, while retaining meaningful exposure to rare earth sector upside through our equity position in AuKing." — Cliff Fitzhenry, Chief Executive Officer, Tusker Minerals Ltd, ASX Release, 4 August 2026 Cameroon Portfolio During the June quarter, Field activities advanced across the Cameroon Central Portfolio, with hand-auger drilling undertaken at the Bounde and Nganda licences along the historically mapped high-grade rutile corridor. Systematic soil sampling was also completed at Yaoundé West, supported by the Company’s in-country Yaoundé heavy mineral sands laboratory. Within the Douala Basin, a JORC (2012) Exploration Target of 2.1–2.6 billion tonnes at 2.1–2.3% total heavy minerals has been established. This includes indicative grades of 0.3–0.35% rutile and 0.06–0.07% zircon, positioning Douala as a globally significant emerging rutile-bearing mineral sands system. Table 2: Douala Basin HMS Project - JORC Exploration Target The potential quantity and grade of the Exploration Target are conceptual in nature. There has been insufficient exploration to estimate a Mineral Resource, and it remains uncertain whether further exploration will result in the estimation of a Mineral Resource. The Exploration Target has been prepared and reported in accordance with the JORC Code. Figure 2: Map of Cameroon Tusker's interests in Cameroon (Source: ASX Announcement) Malawi Portfolio – Mzimba Rutile Project The Company completed the first modern, systematic rutile exploration programme across the approximately 710 km² Mzimba Project on time and within budget. The programme generated the project’s first project-wide, rutile-focused geochemical dataset. Figure 3: Mzimba Project in Malawi ( Source ASX Announcement) The exploration work included regional mapping, six exploration pits, 31 channel samples and 159 soil samples. Well-developed regolith, mottled clays and visible rutile grains observed at surface and in panned concentrates support the potential for a large-scale residual rutile system. Samples are currently being processed through Tusker’s in-country infrastructure in Malawi, with initial rutile assay results expected in the coming weeks. The Company remains fully funded for this phase of exploration. Near-Term Milestones to Watch Execution of the definitive transaction documentation Any AuKing shareholder approval required for the share issue Malawian mining authority approval for the transfer of EPL 0529 and EPL 0705 Completion, and receipt of the A$750,000 cash Tusker's next quarterly cash flow report, which will show what actually arrived and what the cash position is The deferred A$1.25 million cash payment twelve months after completion AuKing's exploration and drilling program at Machinga, which determines whether the performance shares ever convert Assay and drilling results from Mzimba and the Douala Basin, which is where the redirected capital is supposed to go Samso's Concluding Comments Taken at face value, this is a junior explorer passing a project it has chosen not to fund to a buyer that has already taken on its neighbour, on staged terms. That is a common and sensible transaction in this part of the market, and it lets both companies put their money where their conviction is. Tusker gets funding without issuing shares. AuKing gets a second Malawian rare earth asset alongside the one it is already drilling. What we would be watching for are: Conditions still to clear. Definitive documentation, several categories of approval and Malawian regulatory sign-off for two licence transfers all sit between the announcement and completion. None of that is unusual for a cross-border sale. The Tundulu experience does suggest the regulatory step is worth allowing time for. The buyer’s funding. AuKing is a small explorer. It raised A$3 million in April 2026 in connection with Tundulu and is now drilling there. Machinga adds A$2 million of cash obligations on top. How AuKing funds that is a matter for AuKing, and the company has raised capital for the first transaction already. The timing of the deferred A$1.25 million is still the item Tusker shareholders will want to see land. No grade for Machinga in this release. The announcement describes it as a heavy rare earth prospect with demonstrated potential through prior exploration activities. It carries no assay result, no drill intercept and no resource estimate. That is a reasonable choice for a transaction announcement rather than an exploration one. It does mean a reader relying on this document alone cannot form a view on whether A$4 million is a fair price, and would need to go back to the earlier Machinga releases to do so. What non-dilutive means here. The term is accurate. No Tusker shares are being issued, and existing holders are not diluted. It describes the form of the funding rather than its speed, so it is worth pairing with the timing set out above when thinking about what is available to spend in Cameroon in the near term. The retained exposure sits with one company. Tusker will hold AuKing shares and AuKing performance shares from two separate transactions. That gives it continued leverage to rare earths, as the company says. The exposure runs through a single small ASX explorer rather than the sector as a whole, so it will move with AuKing’s own progress and share price. Jurisdiction. Malawi and Cameroon are both frontier exploration jurisdictions, and Samso has made that point about this company before. It cuts both ways. Frontier ground is where under-explored, district-scale positions are still available at reasonable cost, and it is also where approvals and timelines are harder to predict. The Tundulu licence transfer is a useful worked example of the second half of that. The case in favour. Focus is worth a great deal to a company this size. Two years ago this was a six-project explorer in Malawi. It is now a rutile and heavy mineral sands company with a defined exploration target in Cameroon and a laboratory of its own in Yaoundé. That is a clearer proposition to fund and a clearer one to explain to a market. Shareholders who bought this as a rare earth story have a fair question to ask about the change. Shareholders who wanted the company to concentrate on one thing and finish it have had their answer. About Tusker Minerals Ltd Tusker Minerals Ltd (ASX: TSK), formerly DY6 Metals Ltd, is an African-focused explorer advancing critical mineral assets across Cameroon and Malawi. Its projects comprise the Central Rutile Project and Douala Basin HMS Project in Cameroon, and the Mzimba Rutile and Salambidwe Projects in Malawi. The company's Malawian rare earth portfolio has been progressively divested during 2026, with Tundulu sold to AuKing Mining under a binding agreement announced on 17 April 2026 and Machinga the subject of the binding agreement announced on 4 August 2026. The Samso Way – Seek the Research Here at Samso, we pride ourselves on delivering content for investors that is independent and informed by over three decades of experience in the industry. Our content is well-researched and is only created if I see merit in discussing the company's story. Our mission is simple: cut through the noise and spotlight what matters—genuine stories, grounded insights, and real opportunity. Our content is well-researched and is only created if the team sees merit in discussing the company or concept. Investors can explore our three core platforms: Coffee with Samso Samso Insights Samso News There may be numerous paths to success in investing, but the common thread among successful individuals is that they remain committed to making informed decisions. Equip yourself with the right knowledge and tools, and you will be well on your way to achieving your financial goals. Most importantly, investors need to be absolutely diligent in understanding their own risk-reward tolerance and capabilities. Never bite off more than you can chew. As they say, Rome wasn’t built in a day, and the Great Wall stood because it took centuries to complete. The Samso Philosophy: Stay curious. Stay sharp. And remember—digging deeper always uncovers the real value. In Life, there is no such thing as a Free Lunch. Never bite off more than you can chew is my parting comment. Happy Investing, and the only four-letter word you need to know is DYOR. To support our independent nature of our work, please head over to our Support Page and give us a helping hand in any of the ways listed. This is a new initiate for the Samso Platform, and it was always the concept of Samso when we started this journey in 2018. Disclaimer The information or opinions provided herein do not constitute investment advice, an offer, or solicitation to subscribe for, purchase, or sell the investment product(s) mentioned herein. It does not take into consideration, nor have any regard to your specific investment objectives, financial situation, risk profile, tax position and particular, or unique needs and constraints. Read full Disclaimer. If you find this article informative and useful, please help me share the information. I try to write about topics that are interesting and have the potential to be of investment value. It is not easy to find stories that fit those parameters. If you or your organisation sees the benefit of what Samso is trying to achieve and has a need to share your journey, please contact me at noel.ong@samso.com.au. About Samso Samso is a trusted platform that equips dedicated investors with up-to-date industry knowledge and insigh0ts from top CEOs and thought leaders. By staying informed on business advancements and market trends, investors can enhance their financial decisions through a combination of expert guidance and their own research. Samso News | www.samso.com.au | An Investor Lens on ASX-Listed Companies

The iron ore market is splitting in two. The product that pays today's dividends, 62% Fe fines (62% iron content) into Chinese blast furnaces, carries a softening price outlook on every published forecast. The product with the strongest structural story, high-purity magnetite feed for direct-reduction steelmaking, sits mostly in projects that are the hardest to finance. That is the dilemma. This Insight lays out both sides, maps the exposure on the Australian Securities Exchange (ASX), and closes not with a directive but with the what-ifs a reader needs to think through for their own path. Samso Insights Investor Awareness Commodity Series Samso Market Strategy 1.00 — THE DILEMMA Two markets are hiding inside one commodity For thirty years the iron ore story has been a single sentence: dig haematite in the Pilbara, ship it to China, collect the margin. That sentence built the largest export industry Australia has ever had. It is now becoming two different sentences, and they point in different directions. The first market is bulk iron units: products grading 58% to 62% Fe that feed conventional blast furnaces (BF). It remains enormous, and it will not disappear. But its marginal pricing is set by declining Chinese steel output, a weak Chinese property sector, and a wave of new supply from Simandou in Guinea, Brazil and Australia itself. The Australian Government's June 2026 Resources and Energy Quarterly (REQ) expects iron ore export earnings to fall from A$117 billion in 2025-26 to A$77 billion in real terms by 2030-31. The second market is premium low-gangue iron units for direct-reduction (DR) steelmaking: 65% Fe fines, DR-grade pellet feed, DR pellets, briquettes and eventually hot briquetted iron (HBI), the feed that DR plants and electric-arc furnaces (EAF) need. Gangue is the waste material in ore, mostly silica and alumina, and in an electric furnace it does not burn away. It becomes slag, consumes electricity and destroys productivity. That is why direct-reduction customers pay for purity, not just for iron. The tension in this Insight's title sits between those two markets. The companies with the safest cash flows are concentrated in the first one, which has the weakest price trajectory. The strongest structural growth story belongs to the second, yet most of the vehicles offering exposure to it are development projects that need billions of dollars, years of approvals and customer qualification programs that can outlast an investor's patience. The superior product does not automatically produce the superior investment. FIG. 01 below maps where the assets behind both markets sit, from the Labrador Trough, Sweden and the Gulf to the Australian register that Section 7 works through in detail. 2.00 — WHERE THE PRICE SITS A controlled softening, not a collapse At the time of writing, mid-July 2026, the benchmark price for 62% Fe fines delivered into China, quoted on a CFR basis (cost and freight, meaning the price includes ocean shipping to the destination port), is trading at approximately US$100 to US$105 per tonne, having dipped briefly below US$100 in late June before recovering. That is a healthy price for the big, low-cost Australian producers, whose average cash costs the Australian Government estimated at roughly US$33/t in 2025, before sustaining capital, royalties, freight and corporate costs. The forecasts, set out in TABLE 01, all lean the same way. One note on the pricing bases in the table: CFR prices include ocean freight to the destination, while FOB (free on board) prices are quoted at the export port and exclude the ocean freight leg. The decks are a base case, not a guarantee. Chinese stimulus, a bad cyclone season, mine disruptions or a delay at Simandou could hold the benchmark above US$100/t for longer than any of them implies. A softening benchmark also does not make low-cost producers uninvestable; it means their returns lean on volume, cost control and capital management rather than on a rising price. That distinction runs through the rest of this note. 3.00 — TWO MARKETS, ONE ORE What "DR grade" actually means The most common mistake in this theme is treating high-grade concentrate, pellet feed, DR pellets and HBI as interchangeable. They are separate products with separate customers, and the difference is where the risk and the margin sit. TABLE 02 sets out the ladder from mined product to furnace metal. Chemistry is the entry ticket. Representative Midrex specifications put DR pellets (FIG. 03) at a minimum of about 67% Fe, with silica commonly 1.0% to 3.0%, alumina 0.2% to 3.0% and sulphur around 0.01% maximum, and the long-run aspiration for premium feed is more than 67% Fe with under roughly 2% combined gangue. These are operating ranges, not universal contract limits; individual plants can demand tighter numbers. Chemistry alone, however, is not the product. A pellet must survive handling, reduction and discharge without breaking down, swelling or sticking together, and the reduced iron that comes out of the furnace may hold only about 25% to 35% of the parent pellet's strength. A concentrate with a spectacular headline grade can still fail as DR feed. Hydrogen reduction does not remove these requirements either: Midrex trials of commercial pellets at around 67.6% to 68.3% Fe found hydrogen reduction faster than natural-gas reduction, with similar or slightly better product strength, but pellet structure remained decisive. Every new source still faces plant-specific qualification that can take years. FIG. 04 traces the full chain from orebody to electric furnace, and the four questions an investor should be able to answer at each step. Magnetite deposits hold a natural advantage in this market. Grinding and magnetic separation can strip out most of the waste before the ore ever leaves the site, producing concentrates of 67% to 71% Fe with controllable chemistry. Haematite direct-shipping ore cannot usually be engineered that way. The price of that advantage is capital: concentrators, power, water, tailings storage and often a pellet plant. Fine grinding is energy hungry, and a high-grade product made with expensive or emissions-intensive electricity is not automatically a low-carbon product. 4.00 — THE DEMAND QUESTION Real growth, uncertain timing Global direct reduced iron production reached approximately 140.8 million tonnes (Mt) in 2024, a fourth consecutive record and roughly 30% above 2019 (FIG. 05). The direction is not in dispute. The timing and the scale are. The market a new entrant actually has to sell into is far smaller and far more concentrated than the headline suggests. Of the pellets sold openly to DR plants in 2024, a pool of about 47.7 million tonnes, three suppliers accounted for almost three-quarters (FIG. 06). Midrex's December 2025 scenario work models merchant-ore DRI production more than doubling from about 32.9 Mt in 2024 to 72.4 Mt in 2034 (FIG. 07), but nearly every tonne of that growth comes from plants not yet built. On the pellet side, its 2034 scenarios range from a 5.6 Mt surplus to a 16.4 Mt deficit (FIG. 08) depending on which demand case and which supply case eventuate. Midrex is explicit that these are scenarios, not forecasts. Other published scenarios run hotter. BloombergNEF work reported by the Institute for Energy Economics and Financial Analysis (IEEFA) implies a deficit of roughly 15 Mt by 2030 growing toward 133 Mt by 2040; S&P Global work reported by the same source points to a 70 to 85 Mt shortfall by 2035; and the International Iron Metallics Association has estimated merchant DR-pellet demand rising from about 47 Mt in 2022 to 58.5 Mt in 2026 and 117.3 Mt by 2033, with around 31.6 Mt of the 2033 figure not covered by identified supply. These studies use different definitions, some covering total DR-grade requirements and some only the merchant seaborne pool, and they should not be averaged into a consensus. 5.00 — THE PRICE OF PURITY What the premium is, and what it is not The cleanest public reference for DR-pellet pricing comes from Labrador Iron Ore Royalty Corporation, which reported the Platts DR pellet premium averaging approximately US$42 per tonne in the first quarter of 2026, down from US$45/t a year earlier. That premium is quoted over the 65% Fe high-grade fines index, which itself averaged about US$121/t in the quarter, not over the 62% benchmark. It softened year on year because steelmakers under margin pressure chose immediate cost savings over the efficiency benefits of premium feed. FIG. 09 shows the arithmetic, and it carries a warning: premium products remain cyclical. High-grade premiums widen when steel margins, environmental enforcement and productivity incentives are strong, and compress when mills are simply trying to survive a quarter. High grade is no hedge against a falling benchmark, because DR pricing is built on top of an index that moves with the same steel cycle. The premium is not profit either. A pellet producer pays for conversion, fuel, binders, maintenance, freight and plant utilisation out of it (FIG. 10). Headline premium figures can also mislead. Vale reported a portfolio-wide "all-in premium" of just US$0.90/t in the fourth quarter of 2025, a number that covers its entire sales book of fines, blends and lower-premium products, and says almost nothing about what a qualified DR pellet earns under a long-term contract. The DR market is dominated by such contracts, with premiums often settled quarterly; spot indications can diverge from what established suppliers actually receive. 6.00 — THE VALE VARIABLE Mega Hubs, briquettes, and a warning inside an opportunity Vale is the most important company in this story, because it is not simply planning to sell more pellets. It is building a system: produce or source ore, concentrate it into high-grade feed, agglomerate it near cheap energy and deep ports, and partner with steelmakers who build the DRI and HBI plants next door (FIG. 11). Its proposed Mega Hubs in Oman, Saudi Arabia and the United Arab Emirates (UAE), with studies in Brazil, follow exactly that template. The strategic point of Sohar is uncomfortable for anyone holding a remote deposit purely for its grade: Vale does not need every tonne of DR feed to start life as premium ore. It can ship a broader product to a hub and upgrade it next to the customer. The second Vale variable is the cold-bonded briquette. Vale commissioned its first commercial briquette plant at Tubarão in December 2023 (FIG. 12), has a second plant scheduled before 2027, is testing a DR-specific briquette in cooperation with Midrex, and has studied a 1.5 Mtpa DR briquette plant in Louisiana. Vale claims its low-temperature process can involve lower cost, lower capital intensity and approximately 80% lower production emissions than conventional pelletising. If that qualifies at industrial scale across different shaft furnaces, the briquette becomes a genuine alternative to the indurated pellet, and a ceiling on how far DR-pellet premiums can stretch. 7.00 — THE ASX REGISTER What this means on the Australian market Australia is the world's dominant iron ore exporter, yet almost all of that dominance sits on the wrong side of the grade divide. The ASX exposure to the premium story is real but narrow, and it spans the full spectrum from operating cash flow to pure development option. Each company below gets the case for and the case against. The incumbents: BHP, Rio Tinto and the haematite base The major producers are the first horn of the dilemma in corporate form. Their Pilbara haematite operations are among the lowest-cost mines on earth, and at US$100/t they generate extraordinary cash. On the REQ deck, that cash compresses steadily through 2031, and their returns become a function of volume, cost and capital discipline rather than price. Rio Tinto also holds an important, often overlooked position on the other side of the divide: through the Iron Ore Company of Canada (IOC) it is already one of only three suppliers of DR pellets to the seaborne merchant market, alongside Vale and Samarco. The incumbents are not absent from the premium story. They are simply not priced on it. Fortescue (ASX: FMG): the cautionary success Iron Bridge is Australia's flagship magnetite project (FIG. 13) and the most instructive single data point in this note. It produces a genuine 67% to 68% Fe concentrate, exactly the product the DR transition wants. It also cost approximately US$3.9 billion, arrived a year late, and its ramp-up has been repeatedly extended: after plant reliability problems, Fortescue guided shipments of 10 to 12 million tonnes for the 2026 financial year against a 22 Mtpa nameplate, with full capacity now expected in financial year 2028, a five-year ramp from first production in 2023. Even a company with a fortress balance sheet, existing port and rail, and world-class operating teams found magnetite harder, slower and more expensive than planned. And yet the product is real, it ships, and it gives Fortescue a premium stream its haematite peers lack. Both halves of that sentence are the dilemma. Grange Resources (ASX: GRR): the quiet incumbent Grange is the company the magnetite hopefuls want to become, and the market barely prices it. Savage River in Tasmania has mined magnetite since 1967, and the Port Latta plant produces more than two million tonnes of low-impurity blast-furnace pellets a year, realising prices at a substantial premium to the fines benchmarks; its reported average realised price in the March 2026 quarter was approximately US$126/t. The case against is equally concrete: its pellets are BF-grade rather than DR-grade, its customer base is narrow, Tasmanian costs are rising, and the transition of the North Pit to underground mining is a genuine execution risk even after a positive feasibility study. Grange also holds Southdown near Albany, a large undeveloped magnetite project that is effectively a free option on this entire theme at the company's current valuation, and an option is all it is until somebody funds it. Champion Iron (ASX: CIA): the proof that Tier 1 exists Champion is listed on the ASX but operates in Canada's Labrador Trough (FIG. 14), and it has just done the thing every developer promises. In late June 2026 it announced first production of DR-quality iron ore from its Bloom Lake direct-reduction pellet feed (DRPF) project, completed within its estimated $500 million budget, upgrading half the operation toward a product of up to 69% Fe with combined silica and alumina below 1.2%. Commissioning is still under way: an inaugural Capesize sale of at least 160,000 wet tonnes is expected in the third quarter of 2026, a commercial agreement already covers part of near-term capacity, and commercial production is expected toward the end of the company's financial year, around March 2027. This is what the strongest risk-adjusted version of the DR theme looks like: an existing producer spending incremental capital to upgrade a product it already sells, with real costs, real logistics and real customers. What remains is the ramp and the order book: reaching commercial production on schedule, the premium the market actually pays through a soft steel cycle, and how quickly the rest of the volume finds binding offtake. Iron Bear Resources (ASX: IBR): the giant with a giant bill Iron Bear near Schefferville, also in the Labrador Trough (FIG. 15), is the most ambitious DR story on the ASX. Its owner renamed itself from Cyclone Metals to Iron Bear Resources in January 2026, changing its ticker from CLE to IBR, so the company now carries the name of its defining asset. The numbers: a mineral resource of 13.6 billion tonnes at about 30% Fe under the company's June 2026 update, and a 2025 scoping study outlining a 25 Mtpa operation producing about 16 Mt of roughly 70% Fe blast-furnace concentrate and 9 Mt of 71% Fe DR pellets with around 1.1% silica, pre-production capital of approximately US$4.6 billion and a post-tax net present value (NPV) of around US$9.8 billion. Vale's involvement, a staged agreement under which Vale can invest up to about US$138 million to earn up to 75% of the project, with a decision to mine targeted around 2028, is exactly the kind of strategic validation Section 8 says to look for. The other side of the ledger: the capital requirement is many multiples of the company's market value, the Vale agreement is staged rather than guaranteed, a pre-feasibility study is still in progress, and the history of this sector says the distance between a spectacular study and a financed construction decision is where most shareholder value goes to die, or gets diluted to death. Iron Bear can be a company-maker and a value trap from the same spreadsheet, depending entirely on the funding path. The Champion yardstick: what Iron Bear's discount is really pricing The obvious question follows: if Champion can support a market capitalisation of roughly A$2.4 billion on the back of Labrador Trough iron ore, and Iron Bear Resources trades at roughly A$83 million, is Iron Bear a cheap entry into the same business? The comparison is genuinely worth making, because the similarities are real. The two projects sit about 200 km apart on the same iron formation, would ship through the same rail-and-port corridor to Sept-Îles, target the same DR-quality end market, and both hold major-company validation: Champion through a commercial agreement with a global steel group, Iron Bear through its staged agreement with Vale. On product, Iron Bear's pilot work at 71% Fe with about 1.1% silica sits alongside, and on paper slightly above, the 69% Fe product Champion has just produced. On scale, the 25 Mtpa scoping concept would out-produce Bloom Lake's 15 Mtpa nameplate. Judged purely as orebodies feeding the same future market, a valuation gap of roughly thirty times looks extreme. The gap, however, is pricing real differences, and an investor should be able to name them. Champion's value sits on cash flow that exists: about 15 Mtpa of production, roughly $1.8 billion in annual revenue, and a DR upgrade paid for out of earnings. Iron Bear's value sits on a study. Ownership differs too. Champion owns 100% of Bloom Lake, having bought out the Québec government's 36.8% for C$211 million; if Vale funds Iron Bear through its full earn-in, Iron Bear Resources shareholders could hold as little as 25% of the project. The honest comparison is therefore not company against company. It is a minority share of an unbuilt US$4.6 billion project against full ownership of an operating one. The history behind Champion's valuation is the part the cheap-entry argument most needs to confront. Champion never built Bloom Lake from a small-cap base. Consolidated Thompson built it; Cliffs Natural Resources paid about C$4.9 billion for it near the top of the 2011 market, shut it within four years, and Champion bought the constructed mine and its rail assets out of creditor protection in 2016 for C$10.5 million cash plus about C$43 million in assumed liabilities. Champion's billions were created by buying distressed, already-sunk capital for a fraction of a cent in the dollar and restarting it well. Iron Bear has to do the harder thing: fund a first build. The Australian precedents for greenfield magnetite at that scale are Sino Iron, Karara and Iron Bridge, and in each case the asset eventually worked while early equity carried heavy losses or dilution. TABLE 03 sets the two companies side by side. What could a 25% share be worth? Two pieces of illustrative arithmetic frame the question, and both need their assumptions read as carefully as their answers. The first takes the scoping study at face value: 25% of the US$9.79 billion post-tax net present value is about US$2.4 billion, or roughly A$3.7 billion at an assumed exchange rate of about US$0.66. The second anchors to what the market pays today for producing Labrador Trough tonnes: Champion's approximately A$2.4 billion capitalisation over its 15 Mtpa of capacity implies about A$160 million per producing million tonnes of annual capacity, and Iron Bear's attributable 6.25 Mtpa (25% of the 25 Mtpa concept) on that yardstick is worth about A$1.0 billion. TABLE 04 sets both results against today's market capitalisation. Against a current capitalisation of about A$83 million, either end-state figure is more than ten times today's price, and that is exactly why the assumptions deserve the emphasis. These are values for a delivered project in the 2030s, not fair prices for today. A scoping study carries an accuracy range of tens of per cent in either direction. Pre-production companies almost never trade at the face value of their study NPV; a fraction of it is normal. And the largest unknown is funding: if Iron Bear Resources must fund its 25% share of the roughly US$4.6 billion build, that is in the order of US$1.15 billion against an A$83 million equity base, and the terms on which that money is raised, or the extent to which the Vale agreements carry Iron Bear through construction, will determine how much of any end value today's shareholders actually keep. Readers should verify the carry and funding terms in the definitive agreements before leaning on either number. This is illustrative arithmetic, not a valuation, a forecast or a price target. Read this way, the Champion comparison supports one specific conclusion and no more: the market demonstrably pays billions for qualified, producing Labrador Trough DR-grade iron ore, so the destination Iron Bear is pointed at has proven value. What the comparison cannot do is shortcut the path. Iron Bear today is a low-priced option on that destination, and options of this kind re-rate on milestones rather than on production: the pre-feasibility study due through 2026, each stage of Vale's funding elections, and the decision to mine targeted around 2028. Each Vale dollar validates the project, and each stage also transfers more of it to Vale. Magnetite Mines (ASX: MGT) and Hawsons Iron (ASX: HIO): the Braemar options South Australia's Braemar iron formation hosts two listed development options. Razorback, 240 km northeast of Adelaide (FIG. 16), targets a DR-grade concentrate of about 68.5% Fe from a 3.8 billion tonne resource, sits near open-access rail and a renewables-heavy grid, and in February 2026 became the only iron ore project on the Federal Government's Major Project Status list, a designation that helps with approvals but funds nothing. Hawsons, 70 km from Broken Hill (FIG. 17), targets an ultra-premium concentrate around 69% to 70% Fe using a dry-processing flowsheet, and has a completed pre-feasibility study. Hawsons Iron is also the cautionary tale of this pair. In late 2022 the company paused its own bankable feasibility study, the final study before a build decision, because cost inflation had pushed the projected capital and operating numbers beyond what the project could support. It later re-scoped the project around a smaller, staged development. The episode shows exactly where the fragility in development-stage magnetite sits. The orebody did not change and the product target did not change, but the cost environment moved and the project had to stop. Both hold genuinely DR-capable geology. Both need capital measured against their market values in multiples, plus infrastructure, water solutions, offtake and years of qualification. They are long-duration options on the second market, and options can expire. The history the sector would rather forget Balance requires the record. CITIC's Sino Iron project in the Pilbara was conceived at a capital cost of roughly US$2.5 billion and ended up costing in the order of US$10 to 12 billion, years late. Karara in Western Australia's Mid West consumed well over A$2.5 billion, and Gindalbie Metals shareholders were substantially wiped out before Ansteel consolidated the asset. Both projects, note carefully, now operate and ship premium concentrate. The orebodies were real and the product was good. The equity path was catastrophic anyway. That is the sharpest possible statement of the dilemma: in magnetite, the asset can win while the shareholder loses. TABLE 05 summarises the seven exposures and the case for and against each. 8.00 — HOW TO READ A MAGNETITE STORY The framework before the valuation Before assigning any DR-grade value to a project, an investor should be able to tick through six layers of evidence, in roughly this order of difficulty. Complete chemistry, not just iron. A company reporting "70% Fe concentrate" has started the work, not finished it. Silica, alumina, phosphorus, sulphur, alkalis and titanium determine whether the product qualifies, and the grade must be achievable at a commercial recovery, grind size and energy cost across the whole deposit, not from a hand-picked composite. Pelletising and reduction testwork. Green-ball formation, fired strength, abrasion, reducibility, swelling, clustering and post-reduction strength, under both natural-gas and hydrogen-rich conditions. Bench-scale chemistry alone does not make DR pellet feed. Delivered cost, not mine-gate cost. A cheap concentrate at the plant can be uncompetitive after grinding power, pelletising, rail, port, shipping and royalties. Compare the cost landed at the customer, and note that the largest variables, grid power price, water, rail distance, port capacity, usually sit outside the mine gate entirely. Conservative pricing. A robust study works at a conservative 65% index, a conservative grade spread and a conservative DR premium. A project that needs both a permanently high iron ore price and a historically elevated premium is a bet on two cycles at once. Capital intensity per annual tonne. Champion's roughly $500 million to upgrade about 7.5 Mtpa of existing capacity and a greenfield project's US$4 billion-plus to build 10 to 25 Mtpa are not the same investment proposition, whatever the resource sizes say. Large resources can hide poor capital efficiency. Funding and dilution. For a small company, the deposit can succeed while the shareholder fails. How much must be raised before a construction decision? Will existing holders retain meaningful ownership? FIG. 18 condenses all of this into the Samso screening frame: product qualification on one axis, execution readiness on the other. TABLE 06 pairs each archetype with what an investor has to believe and what breaks the thesis. 9.00 — THE WHAT-IFS Samso concluding thoughts: a dilemma is not resolved by a tip - the concluding iron ore story investment case. Samso's usual habit is to end with a view. This Insight ends differently, on purpose. The iron ore dilemma does not have a single correct answer, because the answer depends on variables nobody controls and on each reader's own horizon, risk tolerance and need for income against growth. What we can do is lay the plausible futures side by side, with who benefits and who is exposed in each, and let the reader decide which future they are actually prepared to underwrite. The four futures share one thing. In every one of them, an operating asset with a qualified product and a funded balance sheet lands somewhere between fine and excellent, while an unfunded promise lands somewhere between stranded and diluted. The futures disagree violently about which product earns the premium. They agree almost completely about who survives to collect it. 10.00 — REFERENCES & SOURCES References & sources Every figure and claim in this Insight is sourced below. The figures fall into two categories. FIG. 03, 10, 12, 13 and 14 are photographs, and FIG. 15, 16 and 17 are company maps, reproduced from publicly released company materials with the source credited in each caption. All remaining charts and diagrams (FIG. 01, 02, 04 to 09, 11 and 18) are original Samso illustrations of data and concepts from these references. The locator map (FIG. 01) is an original Samso illustration drawn on a public-domain Natural Earth basemap, with deposit positions plotted at approximate true coordinates from company disclosures. Company figures are as at the publication dates shown in their captions. Market-sensitive numbers (spot prices, index averages and premiums) were current at the time of writing in mid-July 2026 and are to be refreshed on publication day. Midrex Technologies, Inc. 2024 World Direct Reduction Statistics (global DRI production of 140.8 Mt in 2024; regional and merchant supply data, including the 47.7 Mt merchant DR-pellet supply pool and producer shares). Midrex Technologies, Inc. "Iron Ore for Direct Reduction: The Challenge Updated", Direct From Midrex, December 2025 (2034 merchant pellet demand scenarios of 104.9 Mt and 110.7 Mt; supply scenarios of 94.3 Mt and 110.5 Mt; merchant-ore DRI base case to 2034; European project delays affecting roughly 17.5 to 18 Mt of pellet demand; DR-grade chemistry guidance; hydrogen reduction trial results). Labrador Iron Ore Royalty Corporation. First quarter 2026 results (Platts DR pellet premium averaging approximately US$42/t in Q1 2026 versus US$45/t in Q1 2025, quoted over the 65% Fe index; 65% Fe index average of approximately US$121/t; commentary on steelmakers prioritising cost savings over premium feed). Vale S.A. Public disclosures and releases (Tubarão briquette plant inauguration, December 2023; second plant and mobile unit; Midrex technical cooperation on DR briquettes; Louisiana 1.5 Mtpa DR briquette plant study; Mega Hub descriptions in Oman, Saudi Arabia and the UAE; Q4 2025 portfolio all-in premium of US$0.90/t; claimed briquette cost and emissions advantages). Vale S.A. and Jinnan Iron & Steel. Sohar concentration project disclosures (approximately 18 Mtpa feed producing approximately 12.6 Mtpa of high-grade concentrate; targeted start around mid-2027; Vale infrastructure commitment of approximately US$227 million; Jinnan investment reported at approximately US$400 million). Australian Government, Department of Industry, Science and Resources. Resources and Energy Quarterly, June 2026 (iron ore export earnings of A$117 billion in 2025-26 declining to A$77 billion in real 2025-26 dollars by 2030-31; 61% Fe FOB price deck of US$91/t in 2026 to US$72/t nominal in 2031; Chinese import projections; Indian steel demand growth of roughly 5.4% annually; Australian average cash costs of about US$33/t in 2025). World Bank. Commodity Markets Outlook (iron ore averaging US$97/t in 2026 and US$95/t in 2027). Fitch Ratings. Metals and mining price assumptions, June 2026 (62% Fe CFR China at about US$100/t for 2026 and US$90/t for 2027). International Energy Agency (hydrogen-based steelmaking cost premium of approximately 50% to 140% over conventional BF-BOF, and natural-gas start-up for most announced near-zero-capable capacity). IEEFA reporting of BloombergNEF and S&P Global DR-grade supply-demand scenarios; International Iron Metallics Association merchant DR-pellet demand estimates (approximately 47 Mt in 2022, 58.5 Mt in 2026 and 117.3 Mt by 2033, the latter with around 31.6 Mt not covered by identified supply, as cited in Fastmarkets pricing notices, April 2026). Champion Iron Limited (ASX/TSX: CIA). ASX/TSX release, 28-29 June 2026 (first production of DR-quality iron ore from the Bloom Lake DRPF Project; completion within the estimated $500 million budget; product of up to 69% Fe with combined silica and alumina below 1.2%; inaugural Capesize sale of at least 160,000 wmt (wet metric tonnes) anticipated in calendar Q3 2026; commercial agreement covering part of near-term capacity; commissioning ongoing, with commercial production expected toward the end of the company's financial year), and quarterly activities reporting. Fortescue Ltd (ASX: FMG). Iron Bridge disclosures and market reporting (first production April-May 2023 at greater than 67% Fe; 22 Mtpa nameplate; joint venture 69% FMG Magnetite and 31% Formosa Steel IB; May 2025 revised ramp-up guiding FY2026 shipments of 10 to 12 Mt and nameplate in FY2028; total capital of approximately US$3.9 billion). Grange Resources Limited (ASX: GRR). Quarterly reports and announcements (Port Latta pellet production of more than 2 Mtpa; average realised price of approximately US$126/t in the March 2026 quarter; North Pit Underground Definitive Feasibility Study, February 2025; Southdown Magnetite Project). Iron Bear Resources Ltd (ASX: IBR), formerly Cyclone Metals Limited (ASX: CLE); the company changed its name and ticker in January 2026. Iron Bear scoping study, August 2025, and Vale agreement disclosures, February 2025 (mineral resource of 13.6 billion tonnes at 30.03% Fe, including 4.5 Bt indicated at 29% Fe, per the company's June 2026 project update, earlier reported at 16.7 Bt; 25 Mtpa concept producing about 16 Mt of roughly 70% Fe BF concentrate and 9 Mt of 71% Fe DR pellets at about 1.1% silica; pre-production capital of approximately US$4.6 billion; post-tax NPV8 (net present value at an 8% discount rate) of approximately US$9.8 billion; staged Vale funding of up to approximately US$138 million to earn up to 75% of the project; decision to mine targeted around 2028). Magnetite Mines Limited (ASX: MGT). Company disclosures (Razorback DR-grade concentrate target of approximately 68.5% Fe; 3.8 Bt project resource; proximity to rail, grid and Spencer Gulf ports; Federal Major Project Status effective 11 February 2026). Hawsons Iron Limited (ASX: HIO). Company disclosures (target concentrate of approximately 69% to 70% Fe; completed pre-feasibility study; bankable feasibility study paused in late 2022 amid cost escalation and subsequently re-scoped). Bloom Lake ownership history, public record (Cliffs Natural Resources' acquisition of Consolidated Thompson Iron Mines for approximately C$4.9 billion including net debt, announced January 2011; Bloom Lake suspension and creditor protection, 2015; acquisition of the Bloom Lake mine and rail assets by Champion Iron's subsidiary Quebec Iron Ore for C$10.5 million cash plus approximately C$42.8 million in assumed liabilities, completed April 2016; acquisition of the remaining 36.8% minority interest in Bloom Lake for C$211 million, completed around the turn of 2026). Market capitalisations, exchange data, mid-July 2026, approximate and to be refreshed on publication day (Champion Iron approximately A$2.4 billion; Iron Bear Resources approximately A$83 million at about A$0.073 per share). The illustrative valuation arithmetic in Section 7 combines these figures with the Iron Bear scoping study outputs and an assumed AUD/USD exchange rate of approximately 0.66; all inputs are approximate and to be refreshed on publication day. Historical capital outcomes, public record (CITIC Sino Iron: initial estimate of roughly US$2.5 billion against a final cost reported in the order of US$10 to 12 billion; Karara: capital in excess of approximately A$2.5 billion and the effective loss of Gindalbie Metals shareholder value prior to Ansteel consolidation). Figures approximate; readers should consult primary filings. Spot pricing, mid-July 2026 (62% Fe CFR China trading at approximately US$100 to US$105/t across published indices, including a KORE 62% Fe Qingdao print of US$102.73/t on 14 July 2026, after a late-June dip below US$100/t). Reproduced company imagery: iron ore pellet and Port Latta plant photographs (FIG. 03 and 10) from Grange Resources Limited's corporate website, grangeresources.com.au, tonal levels adjusted for print; Vale briquette plant photograph (FIG. 12) by Rafael Coelho / Vale, from Vale's public news releases on the Tubarão briquette plants, December 2023; Iron Bridge photograph (FIG. 13) from Fortescue's public news release "Fortescue's Iron Bridge commences high grade magnetite production", April 2023; Bloom Lake photograph (FIG. 14) from Champion Iron Limited's corporate website, championiron.com; Iron Bear location map (FIG. 15) from the scoping study materials of Iron Bear Resources Ltd, published under its former name Cyclone Metals, as republished by Stockhead, 12 August 2025; Razorback project map (FIG. 16) from Magnetite Mines Limited's website (project map, October 2025); Hawsons location map (FIG. 17) from Hawsons Iron Limited's website, hawsons.com.au. All reproduced with attribution; rights remain with the respective companies and photographers. Natural Earth. Free vector map data, public domain (coastlines used in FIG. 01). The Samso Way – Seek the Research Here at Samso, we pride ourselves on delivering content for investors that is independent and informed by over three decades of experience in the industry. Our content is well-researched and is only created if I see merit in discussing the company's story. Our mission is simple: cut through the noise and spotlight what matters—genuine stories, grounded insights, and real opportunity. Our content is well-researched and is only created if the team sees merit in discussing the company or concept. Investors can explore our three core platforms: Coffee with Samso Samso Insights Samso News There may be numerous paths to success in investing, but the common thread among successful individuals is that they remain committed to making informed decisions. Equip yourself with the right knowledge and tools, and you will be well on your way to achieving your financial goals. Most importantly, investors need to be absolutely diligent in understanding their own risk-reward tolerance and capabilities. Never bite off more than you can chew. As they say, Rome wasn’t built in a day, and the Great Wall stood because it took centuries to complete. The Samso Philosophy: Stay curious. Stay sharp. And remember—digging deeper always uncovers the real value. In Life, there is no such thing as a Free Lunch. Never bite off more than you can chew is my parting comment. Happy Investing, and the only four-letter word you need to know is DYOR. To support our independent work, please head over to our Support Page and give us a helping hand in any of the ways listed. This is a new initiative for the Samso Platform, and it was always the concept of Samso when we started this journey in 2018. Disclaimer The information or opinions provided herein do not constitute investment advice, an offer, or solicitation to subscribe for, purchase, or sell the investment product(s) mentioned herein. It does not take into consideration, nor have any regard to your specific investment objectives, financial situation, risk profile, tax position and particular, or unique needs and constraints. Read full Disclaimer. About Samso Samso is a trusted platform that equips dedicated investors with up-to-date industry knowledge and insights from top CEOs and thought leaders. By staying informed on business advancements and market trends, investors can enhance their financial decisions through a combination of expert guidance and their own research. Samso Insights | www.samso.com.au | An Investor Lens on ASX-Listed Companies

The company is consolidating 226 claims across seven project areas in Nevada OD6 Metals Limited (ASX: OD6) has completed the acquisition of the Quinn Fluorspar Project in Nevada, United States, through its wholly owned subsidiary US Fluorspar LLC. The acquisition follows the exercise of an exclusive option announced on 9 June 2026 and shareholder approval at an Extraordinary General Meeting held on 20 July 2026. The project sits roughly 220km north of Las Vegas, in what the company describes as a highly prospective epithermal fluorite district. Epithermal means the mineralisation formed from hot fluids circulating at relatively shallow depths in the Earth's crust, which is why so much of this system is visible at the surface. Fluorite is the mineral. Fluorspar is the commercial name for the rock when it is mined and sold. The acquisition itself covered 48 State of Nevada mining claims. Since March 2026, OD6 has staked a further 178 claims around them. The company now holds 226 claims covering approximately 1,890 hectares, or 4,670 acres, across seven separate project areas. Those areas are Mammoth, Horseshoe, Dress Circle, Bonanza, Bruno, El Cortez and Blue Bell (Figure 1). Figure 1: Quinn Fluorspar Project, district claim position and project areas, Nevada USA. (Source: OD6 ASX Release) A quick sense check on that ground position is useful. A standard United States lode mining claim is 600 feet by 1,500 feet, which works out to about 8.4 hectares. Multiply that by 226 claims, and you get roughly 1,890 hectares. The Vocabulary, If you are new to fluorspar, these are the terms that will keep appearing. Fluorspar (CaF₂) is the commercial name for the mineral fluorite, calcium fluoride. It is the feedstock for hydrofluoric acid, which in turn feeds refrigerants, semiconductor etching chemicals, battery electrolytes, uranium enrichment and a long list of defence applications. Grade is expressed as a percentage of CaF₂ in the rock. Producing fluorspar projects are generally considered viable above about 20% CaF₂, and some operators cite figures as low as 8% CaF₂. Acidspar is the premium product and requires more than 97% CaF₂. It is what the chemical industry buys. Metspar is the lower-grade product, typically in the 60% to 85% CaF₂ range, used as a flux in steelmaking. Beneficiation is the processing step that upgrades run-of-mine rock into a saleable concentrate. Optical sorting is one method, using cameras to separate mineralised rock from waste before any chemistry is applied. Flotation is another, using chemical reagents and air bubbles to float the target mineral away from waste. Channel sampling means cutting a continuous groove across an exposed rock face and assaying the material. It gives a grade over a measured width, which is more informative than a single rock chip, but it still only samples the surface. A mining claim in the United States is a right to the minerals on a defined piece of federal land, maintained by annual fees. It is roughly the American equivalent of an Australian exploration licence, but claims are much smaller and are staked individually. A JORC Mineral Resource is a formally estimated tonnage and grade, classified as Measured, Indicated or Inferred according to how confident the geologist is. Quinn does not have one. TREO stands for Total Rare Earth Oxide, the measure used for OD6's Splinter Rock rare earth project in Western Australia. VWAP is the volume-weighted average price, an average share price over a period weighted by how many shares traded at each price. It is commonly used to set an issue price so that a single day's trading cannot distort the value. FAST-41 is a United States federal permitting framework that gives eligible infrastructure and critical minerals projects a coordinated timetable across agencies. Eligibility is applied for, not automatic. What Was Actually Paid OD6 states a total consideration of A$200,000, being A$100,000 in cash and A$100,000 in shares. On completion, OD6 paid the A$100,000 cash to the vendors and issued 700,706 fully paid ordinary shares to them at a deemed value of A$100,000. That works out to approximately A$0.143 per share, based on the 10-day VWAP prior to the option exercise date. OD6 also issued 1,500,000 fully paid ordinary shares to Sapphire Beginnings Capital Pty Ltd, or its nominees, described as the final tranche of facilitation consideration. Facilitation consideration is a fee paid to whoever introduced or brokered the transaction. Applying the same A$0.143 per share reference used for the vendor shares, those 1,500,000 facilitation shares carry an implied value of about A$214,500. That is Samso's arithmetic, not company guidance, and the company has not put a value on those shares in this release. Two observations follow from that. First, the fee to get the deal done was larger in value terms than the consideration paid to the people selling the asset. Second, the all-in cost at completion is closer to A$415,000 than to A$200,000, once the facilitation shares are counted. Neither of those is a criticism. Facilitation fees are normal, and the amounts involved are small in the context of any exploration program. But investors reading the A$200,000 headline should know that it is the vendor consideration, not the total cost of the transaction. From 48 Claims to 226 OD6 acquired 48 claims. It has staked 178 more, taking the holding to 226. Managing Director Brett Hazelden describes the project as having expanded almost fivefold during the option period, and by claim count that description holds up. Alongside the expansion, the company reports more than 20 fluorspar occurrences across the enlarged position, which it says supports a future hub-and-spoke development strategy. Hub-and-spoke means building one central processing plant and trucking ore to it from several smaller deposits, rather than building separate infrastructure at each one. It is a sensible model for a district with many small high-grade occurrences, and it is also a model that depends entirely on there being enough tonnes in aggregate to justify the hub. The work completed during the option period, as listed by the company, covers verification of historical drill and channel results, confirmation of an 8km mineralised corridor, definition of multiple drill targets from mapping, fluorine geochemistry and structural interpretation, collection of metallurgical samples, commencement of environmental baseline and archaeological studies, and advancement of a permitting strategy targeting FAST-41 eligibility. The grades reported through that period, as titled in the company's own ASX releases listed in this announcement, include up to 53.2% CaF₂ at Mammoth on 7 April 2026, a continuous 12m at 40.8% CaF₂ on 9 April 2026, and channel sampling up to 75% CaF₂ at Horseshoe on 15 April 2026. A further Horseshoe release on 8 July 2026 reported continuing grade and width. Set against a general economic threshold of around 20% CaF₂, these are high numbers. The company also notes low impurity levels, which is the quiet technical point that carries a lot of weight. Fluorspar buyers care about silica, calcite and sulphur content, because impurities determine whether a concentrate can reach Acidspar specification of more than 97% CaF₂ or is limited to Metspar. Low impurities widen the product options. They do not guarantee them, because that still depends on the beneficiation testwork now underway. Management Commentary Brett Hazelden framed the completion as a transition point rather than an endpoint. He described the project as having evolved from an acquisition opportunity into what the company believes is a district-scale critical minerals project, pointing to the expansion of the ground, the identification of multiple mineralised systems, and the progress on permitting and metallurgy. He also made a claim about position, stating that OD6 now controls one of the largest consolidated fluorspar exploration positions in the United States. That is a claim Samso has not independently verified, and it is worth noting that the United States fluorspar exploration sector is small, which makes the bar for such a statement lower than it would be in a larger commodity. "What began as an acquisition opportunity has rapidly evolved into what we believe is a district-scale critical minerals project. During the option period we not only confirmed exceptional historical grades but expanded the project almost fivefold, identified multiple new mineralised systems, advanced permitting and metallurgy, and generated numerous high-priority drill targets." — Brett Hazelden, Managing Director, OD6 Metals Limited, ASX Release, 4 August 2026 The Market Backdrop Fluorspar is designated a critical mineral in the United States. The announcement states that the country is 100% import reliant, with approximately 62% of global supply sourced from China, citing the United States Geological Survey 2024 data. Figure 2: Global Fluorspar production by country That figure is worth pausing on, because it moves around. The highlights section of this announcement says approximately 63%. The About OD6 section of the same document says more than 60%. Samso's own May 2026 coverage of the Big Jim rediscovery used approximately 68%. The differences are not material to the argument, and the argument holds under all three versions. China supplies the majority of the world's fluorspar, and the United States supplies none of its own. Nevada is also a genuinely favourable place to be doing this. The company cites the Fraser Institute 2025 Mining Attractiveness Index, which ranked Nevada second globally. Samso has not independently checked that ranking against the source publication. What Samso Is Watching, and the Case Against Quinn has never been drilled by OD6. Every grade published to date comes from surface work, historical records or historical drilling by others. Channel samples and rock chips tell you what is at the surface. They do not tell you whether mineralisation continues at depth, over what width, or with what continuity. Drilling is the test, and the maiden drilling program is still listed under next steps. There is no JORC Mineral Resource. Without one, there is no tonnage figure, no grade estimate at a stated confidence level, and no basis for any economic study. Everything downstream of that, including hub-and-spoke development, depends on a resource that does not yet exist. The historical results are historical. Grades of 94.6%, 96% and 98.6% CaF₂ from Big Jim come from a 1947 report. They are consistent with modern surface observations, but they were not generated under JORC and cannot be relied upon as current. Permitting is a real timeline. The company is preparing drilling and bulk sample approvals and pursuing FAST-41 eligibility. FAST-41 eligibility is applied for, not granted by default. Federal land permitting in the United States takes as long as it takes, and it is outside the company's control. Beneficiation is unproven at Quinn. Reaching Acidspar specification of more than 97% CaF₂ requires the metallurgical flowsheet to work on this specific ore. Optical sorting and flotation testwork is described as advancing, which means it is not finished. Fluorspar is a small and opaque market. It does not trade on a transparent exchange the way copper or gold does. Pricing is negotiated, and a new producer needs an offtake agreement with a real buyer. None has been announced. OD6 is running three projects at once. Splinter Rock rare earths in Western Australia, Gulf Creek copper-zinc in New South Wales, and now Quinn. Each one wants capital. A junior explorer with three assets and one balance sheet has to choose, and the choice usually shows up as dilution. The gold is early. The announcement flags potential gold upside from soil sampling across Mammoth, Horseshoe and Dress Circle, with detail in the 29 July 2026 release. Soil anomalies and rock chips are the very first step in a gold exploration sequence, and most of them do not become deposits. Near-Term Milestones to Watch The company's own list of next steps gives a clear sequence to track. Receipt and interpretation of pending assay results Completion of geological modelling and drill targeting Ongoing detailed geological and structural mapping Expanded field programs Integration of results into permitting activities Progress on metallurgical testwork, including optical sorting and beneficiation Lodgement of bulk sample permit applications Preparation of a maiden drilling program Resource definition Development studies Of those, the two that will move the story furthest are the bulk sample permit and the maiden drilling program. A bulk sample gives real metallurgical data on a representative quantity of ore rather than a few kilograms. Drilling gives depth. Everything else is preparation for those two. Samso Concluding Comments The 4 August 2026 announcement is an administrative milestone and a reminder of the exploration results over the preceding five months. The key here is ownership. OD6 now holds the ground outright rather than under option, which removes a condition that has sat over the story since March. This marks the "start" of the project, and markets behave differently on ground that is owned outright. Market perception is what this business is all about, and for investors, the risk that sat over an optioned asset, the risk that the option lapses before it delivers anything, is now negated. As OD6 moves forward, that ownership also gives investors some assurance that the spending on drilling, permitting and metallurgy carries real intent. The money going into the ground now has more meaning, because investors can take part knowing the project is actually going forward. The next twelve months should therefore look different in pace to the last five. On the price, my view is that A$200,000, or A$415,000 including the facilitation shares, is a small number for what has been assembled. It is also a number that tells you something. Nobody sells a proven high-grade critical minerals district for A$200,000. What OD6 bought was a well-documented historical district with excellent surface grades and no modern drilling. The vendors were selling optionality, and OD6 was buying the right to spend money finding out. What that all means in the scheme of things is that the vendors and OD6 are partners and what the end result will look like is kind of anyone's guess, but this is the business of finding an economically viable mineral deposit. The grades are the reason to keep watching. Surface channel results in the 40% to 75% CaF₂ range are several times the threshold at which fluorspar projects are generally considered economic. If those grades persist at depth over meaningful widths, the arithmetic of a small high-grade domestic supplier becomes interesting quickly, particularly with a customer base that currently imports every tonne it uses. The other upside we have not talked about, though it has been mentioned, is that the project may not be a single commodity. Management have mentioned the potential of a mineral-rich epithermal system. Should shareholders and potential investors be excited? This is still a pre-drilling exploration project, and the single most important data set has not been collected, but as investors, this is what will drive value creation. With a market capitalisation of A$35M (as of 5 August 2026), there is room for improvement. As usual, DYOR. About OD6 Metals Limited OD6 Metals Ltd (ASX: OD6) is an Australian critical minerals exploration and development company with projects across fluorspar, rare earth elements and copper in the United States and Australia. The Quinn Fluorspar Project in Nevada now comprises 226 mining claims over approximately 1,890 hectares, hosting the Horseshoe, Mammoth and Big Jim deposits among more than 20 fluorspar occurrences. Figure 3: Location of OD6 Metals' Quinn Fluorspar Project (Source: ASX Announcement) The Splinter Rock Rare Earth Project in Western Australia is 100% owned and hosts a Mineral Resource Estimate of 119Mt at 1,632ppm TREO in the Indicated category and 563Mt at 1,275ppm TREO in the Inferred category. Indicated means the geologist has enough data to be reasonably confident in tonnage and grade. Inferred means the estimate rests on limited sampling and has lower confidence. OD6 is advancing a processing flowsheet using heap leaching, nanofiltration and ion exchange, targeting approximately 75% overall recovery of neodymium and praseodymium and a mixed rare earth carbonate or hydroxide product of approximately 56% to 59% TREO. The Gulf Creek Copper-Zinc VMS Project in New South Wales sits in a historically high-grade copper district, with recent drilling and geophysics identifying targets along more than 10km of prospective strike. The Samso Way – Seek the Research Here at Samso, we pride ourselves on delivering content for investors that is independent and informed by over three decades of experience in the industry. Our content is well-researched and is only created if I see merit in discussing the company's story. Our mission is simple: cut through the noise and spotlight what matters—genuine stories, grounded insights, and real opportunity. Our content is well-researched and is only created if the team sees merit in discussing the company or concept. Investors can explore our three core platforms: Coffee with Samso Samso Insights Samso News There may be numerous paths to success in investing, but the common thread among successful individuals is that they remain committed to making informed decisions. Equip yourself with the right knowledge and tools, and you will be well on your way to achieving your financial goals. Most importantly, investors need to be absolutely diligent in understanding their own risk-reward tolerance and capabilities. Never bite off more than you can chew. As they say, Rome wasn’t built in a day, and the Great Wall stood because it took centuries to complete. The Samso Philosophy: Stay curious. Stay sharp. And remember—digging deeper always uncovers the real value. In Life, there is no such thing as a Free Lunch. Never bite off more than you can chew is my parting comment. Happy Investing, and the only four-letter word you need to know is DYOR. To support our independent nature of our work, please head over to our Support Page and give us a helping hand in any of the ways listed. This is a new initiate for the Samso Platform, and it was always the concept of Samso when we started this journey in 2018. Disclaimer The information or opinions provided herein do not constitute investment advice, an offer, or solicitation to subscribe for, purchase, or sell the investment product(s) mentioned herein. It does not take into consideration, nor have any regard to your specific investment objectives, financial situation, risk profile, tax position and particular, or unique needs and constraints. Read full Disclaimer. If you find this article informative and useful, please help me share the information. I try to write about topics that are interesting and have the potential to be of investment value. It is not easy to find stories that fit those parameters. If you or your organisation sees the benefit of what Samso is trying to achieve and has a need to share your journey, please contact me at noel.ong@samso.com.au. About Samso Samso is a trusted platform that equips dedicated investors with up-to-date industry knowledge and insigh0ts from top CEOs and thought leaders. By staying informed on business advancements and market trends, investors can enhance their financial decisions through a combination of expert guidance and their own research. Samso News | www.samso.com.au | An Investor Lens on ASX-Listed Companies

Washington has set a hard deadline of 1 January 2027. Australian producers are on the right side of it, but only if they can prove where every gram came from. Australian critical minerals producers have been handed their clearest opening yet into the United States defence supply chain, after President Trump signed an executive order closing the loophole that let American contractors keep buying Chinese magnets, tungsten and tantalum. Executive Order 14415, signed on 20 July 2026 and published in the Federal Register three days later, does not create a new ban. It enforces an existing one. US law already bars the defence department from buying certain materials produced in a “covered nation”, a list containing only China, Russia, North Korea and Iran. Australia is not on it. The order is unusually candid about why it was needed. Its opening section states that despite the longstanding prohibition, defence contractors have historically under-prioritised domestic production and resilience. That is an admission that the existing law has not been working. How the escape hatch worked, and how it closes The statute has two release valves. The first lets the Secretary waive the prohibition for a specific end item when a specific covered material of satisfactory quality and quantity cannot be obtained when needed at a reasonable price. Each such waiver runs for a maximum of 36 months. The second is a national security waiver, allowing delivery of an item containing covered material from a covered nation where the Secretary determines in writing that accepting it is necessary to United States national security interests. Contractors have leaned on the first valve heavily, because for most of these materials no tested, approved, available Western alternative existed. Qualifying a new supplier into a defence programme is slow, expensive and thankless work, and when a compliant Chinese supplier sits there at a fraction of the price, nobody volunteers to do it. From 1 January 2027, the order shuts the first valve and narrows the second. Routine non-availability waivers stop. A contractor still wanting one must submit a formal mitigation plan, accepted by the Secretary, that identifies the non-compliant source, documents evidence of exhaustive efforts to find compliant material, describes the steps to remove the non-compliant material from the supply chain, and sets a strict projected timeline for doing so. National security waivers survive, but they now require a request from the Secretary or a service Secretary to the Assistant to the President for National Security Affairs. Routing that decision through the National Security Council is a meaningful change of altitude. A waiver that was previously an acquisition matter becomes a White House matter. The sharpest clause is Section 2(c). A contractor's failure to qualify a domestic source no longer constitutes non-availability, unless that contractor can demonstrate active, adequately funded and ongoing efforts to qualify one. In plain terms, “we couldn't find anyone else” stops being an acceptable answer if you never seriously looked. There is a penalty tail. Where the Secretary determines a contractor has engaged in fraud or deliberately misled the government in its mitigation plan, or has knowingly failed to implement it, the order directs him to exercise all appropriate contractual remedies and permits referral to the Attorney General for investigation and possible prosecution. Section 4(b) is blunter still: failure to qualify an alternative source becomes grounds to suspend or terminate task orders, decline to exercise contract options, and terminate the existing contract. That is the real weapon. Not a fine, but the threat of losing the work. Why 1 January 2027 The date is not arbitrary, and this is the detail most commentary has missed. Legislation passed in 2021 amends the same statute on exactly that day. From 1 January 2027 the department may not enter into a contract for any covered material mined, refined, or separated in a covered nation, rather than only material melted or produced there. The same amendment narrows a widely used carve-out. The exception for commercially available off-the-shelf items currently excludes items that are 50 per cent or more tungsten by weight. From January 2027, that exclusion widens from tungsten to every covered material. So the waiver door and the traceability net move together. A contractor relying on a Chinese input buried three tiers down will find, on the same morning, that the input is more clearly caught, the off-the-shelf shortcut is narrower, and the exemption is gone. Figure 1. The two changes that matter land on the same day. Everything else is process around them. The list is shorter than “critical minerals” suggests The covered materials are narrower than most investors assume. “Critical minerals,” as a phrase, now encompasses roughly 60 commodities. The statute this order enforces does not. As it currently stands, the list is samarium-cobalt magnets, neodymium-iron-boron magnets, tungsten metal powder, tungsten heavy alloy and components containing it, tantalum metals and alloys, and molybdenum. Molybdenum was added in December 2025. Germanium and gallium join on 18 December 2027 under the same legislation. Two of the six current entries are tungsten. One is tantalum. Only two relate to rare earths. That matters for a market that has spent three years treating “critical minerals exposure” as a synonym for rare earths. Figure 2. The statutory list is specific and short. Broad “critical minerals” exposure is not the same as exposure to this order. Australia's Tungsten Paradox Tungsten has the highest melting point of any metal, which is why it goes into armour-piercing rounds, cutting tools and anything that runs hot. China dominates production. Australia ranks second in the world for tungsten resources, according to Geoscience Australia and peer-reviewed assessments of the national inventory, yet supplies only around one per cent of global output. Figure 3. Resource endowment without production capacity. The order raises the value of closing that gap. Tantalum, tungsten or molybdenum produced from recycled material is exempt where the contractor can show the recycled material was produced outside a covered nation and the melting and further processing takes place in the United States or in the country of a qualifying foreign government. Australia appears on the qualifying country list used to implement that concept in United States defence acquisition regulation, alongside Canada, Japan, the United Kingdom and around two dozen others. From 18 December 2027 the same exception extends to gallium and germanium. The practical reading is that Australian downstream processing of recycled tungsten, tantalum and molybdenum sits inside a statutory exemption, not merely outside a prohibition. For a country with substantial tungsten resources and an existing hard-metal recycling industry, that is a specific and under-discussed opening. It would be worth confirming the current qualifying country schedule before building a business case on it, since these lists are amended. Two other exceptions matter less to Australian producers but tell you where the pressure is heading. Off-the-shelf items are exempt, on the narrowing basis described above. Electronic devices are exempt unless the Secretary determines domestic availability of a particular device is critical to national security, and Section 2(f) of the order directs the Secretary to review how that electronics exemption is currently being applied. That is the single largest remaining gap in the regime, and the order has put it under review. Figure 4. Compliant supply is not only a domestic question. Four of the routes that stay open run through allied and US-financed sources The catches Being Australian is not sufficient. Section 3 requires contractors to submit a complete indentured Bill of Materials, tracing every component back to the origin of the raw materials across all supplier tiers. It also requires screening every supplier for foreign ownership, control or influence from a covered nation, defined broadly enough to capture influence that exists but is never exercised. Two consequences follow. First, an Australian mine that ships concentrate to China for separation is not producing compliant material once the January 2027 language bites. The premium attaches to the chain, not the deposit. Second, an ASX-listed company with a substantial Chinese shareholder or offtake counterparty can be caught despite mining in Australia. Registers and offtake books are now due-diligence documents. The clause nobody is discussing Section 6 may be the most commercially significant provision for Australian developers. Nothing in the order impairs acquisition of critical minerals produced by a foreign project financed, guaranteed or insured by the US Export-Import Bank or the Development Finance Corporation, or supported by the US departments of State, War, Commerce or Energy. That is a protected lane, written into the order, for foreign projects carrying United States government money. American financing has stopped being merely cheap capital and become a compliance position. It sits alongside Project Vault, the US Strategic Critical Minerals Reserve announced in February 2026 and backed by an EXIM loan of up to US$10 billion plus roughly US$2 billion in private capital, which the order also carves out. Two reality checks Defence volumes are small. Global magnet demand is dominated by electric vehicles, wind turbines and industrial motors, and defence consumption is a modest share of it. This order creates high-value, low-volume, certification-heavy demand. It does not move the bulk price deck. Any company whose investment case rests primarily on defence demand is selling a story the tonnages will not support. Qualification takes years. Getting a material tested and approved into a specific weapons platform is a multi-year exercise. The order concedes as much: Section 4(c) gives the Secretary 90 days simply to produce a strategy for accelerating testing and qualification, including identifying regulations to rescind. When a policy document has to commission a plan to speed up its own approvals process, that tells you how slow the process is. Revenue from this shift lands well after the announcements. What happens next Implementing guidance is due within 180 days, with a further 90 days for the regulations themselves. Those will matter more than the order, because they settle whether “allied” means “preferred” or merely “permitted”. The preamble commits to sourcing “domestically or from allied nations”, but the operative sections build only a mechanism for removing Chinese material, not a preference for Australian material over American. Two cautions. Defence is a small share of global magnet demand, so this creates high-value, low-volume, certification-heavy business rather than a shift in the bulk price deck. And qualifying a material into a weapons platform takes years, which is why Section 4(c) gives officials 90 days simply to devise a strategy for speeding that up. The opening is real. The revenue is not imminent. FOUR QUESTIONS FOR ANY COMPANY CLAIMING EXPOSURE • Is its output on the covered materials list, directly or through a product it feeds? • Can it document chain of custody from pit to delivered product, and where does the trail go cold? • Does its share register or offtake book create a foreign ownership, control or influence problem? • Is it inside the Section 6 financing perimeter, or should it be? The Samso Way – Seek the Research Here at Samso, we pride ourselves on delivering content for investors that is independent and informed by over three decades of experience in the industry. Our content is well-researched and is only created if I see merit in discussing the company's story. Our mission is simple: cut through the noise and spotlight what matters—genuine stories, grounded insights, and real opportunity. Our content is well-researched and is only created if the team sees merit in discussing the company or concept. Investors can explore our three core platforms: Coffee with Samso Samso Insights Samso News There may be numerous paths to success in investing, but the common thread among successful individuals is that they remain committed to making informed decisions. Equip yourself with the right knowledge and tools, and you will be well on your way to achieving your financial goals. Most importantly, investors need to be absolutely diligent in understanding their own risk-reward tolerance and capabilities. Never bite off more than you can chew. As they say, Rome wasn’t built in a day, and the Great Wall stood because it took centuries to complete. The Samso Philosophy: Stay curious. Stay sharp. And remember—digging deeper always uncovers the real value. In Life, there is no such thing as a Free Lunch. Never bite off more than you can chew is my parting comment. Happy Investing, and the only four-letter word you need to know is DYOR. To support our independent nature of our work, please head over to our Support Page and give us a helping hand in any of the ways listed. This is a new initiate for the Samso Platform, and it was always the concept of Samso when we started this journey in 2018. Disclaimer The information or opinions provided herein do not constitute investment advice, an offer, or solicitation to subscribe for, purchase, or sell the investment product(s) mentioned herein. It does not take into consideration, nor have any regard to your specific investment objectives, financial situation, risk profile, tax position and particular, or unique needs and constraints. Read full Disclaimer. If you find this article informative and useful, please help me share the information. I try to write about topics that are interesting and have the potential to be of investment value. It is not easy to find stories that fit those parameters. If you or your organisation sees the benefit of what Samso is trying to achieve and has a need to share your journey, please contact me at noel.ong@samso.com.au. About Samso Samso is a trusted platform that equips dedicated investors with up-to-date industry knowledge and insigh0ts from top CEOs and thought leaders. By staying informed on business advancements and market trends, investors can enhance their financial decisions through a combination of expert guidance and their own research. Samso News | www.samso.com.au | An Investor Lens on ASX-Listed Companies

The company has released its first Pre-Feasibility Study for its namesake magnetite project in Canada's Labrador Trough, together with a maiden Ore Reserve of 3.3 billion tonnes. Iron Bear Resources Limited (ASX: IBR) released the maiden Pre-Feasibility Study (PFS) for the Iron Bear Project in Newfoundland and Labrador, Canada, and with it the project's first Ore Reserve. The headline numbers are large by any standard: A maiden Probable Ore Reserve of 3.3 billion tonnes at 29.1% iron; A post-tax net present value of USD 9.0 billion. A 44-year mine life producing 23 million tonnes per annum (Mtpa) of high-grade iron ore products. A development partner in Vale S.A., one of the world's largest iron ore producers. Table 1: Ore Reserve Summary Plain English first: what is a PFS? A Pre-Feasibility Study is the middle stage of mining project studies. It sits between a Scoping Study (a first-pass concept) and a Bankable Feasibility Study or BFS (the detailed document lenders rely on). Iron Bear's PFS has been completed to what engineers call AACE Class 4 accuracy, meaning the cost estimates carry a stated accuracy range of roughly plus or minus 30%. In other words, the numbers are considered robust enough to decide whether to keep spending money, but not yet precise enough to build from. The maiden ore reserve on a giant mineral resource The distinction between a Mineral Resource and an Ore Reserve matters more here than in most stories. A Mineral Resource is mineralisation that has been measured and modelled with reasonable prospects of eventual economic extraction. An Ore Reserve is the smaller, harder-won subset that has been put through what the JORC Code calls modifying factors: pit designs, schedules, costs, prices and processing recoveries. A reserve is a statement that this material can be mined economically under the stated assumptions. Iron Bear's overall Mineral Resource Estimate stands at 13.6 billion tonnes at 30.03% iron, using a cut-off of 12.5% magnetic iron (MagFe, the portion of the iron sitting in magnetite, the magnetic iron mineral that this style of processing recovers). Of that, 4.5 billion tonnes is in the Indicated category, and 9.1 billion tonnes is Inferred, the lowest-confidence category. Table 2: Iron Bear Mineral Resource Estimate at 12.5% magnetic Fe cut-off grade The new maiden Ore Reserve converts the Indicated material inside the designed pit into 3.3 billion tonnes of Probable Ore Reserve at 29.1% total iron and 19.8% MagFe. The reserve was signed off by TME Mine Consulting, working from the Mineral Resource prepared by Snowden Optiro, and was evaluated at a long-term 62% Fe iron ore price of USD 100 per tonne. Figure 1: Process Feed Quantities and Grade by classification (Source: ASX Announcement) Two things stand out. First, there are no Proved Reserves, only Probable. That is normal at the PFS stage and reflects the confidence level of the underlying resource. Second, and more striking, the entire 44-year mine plan draws on less than 25% of the total Mineral Resource. Table 3: Iron Bear Mineral Resource Estimate at 12.5% magnetic Fe cut-off grade The 3.4 billion tonne mill feed inventory is 95% Probable Ore Reserve and 5% Inferred material. Managing Director Paul Berend makes the implication explicit in the announcement: significantly higher production scenarios can be envisaged in future economic studies. The company is careful, as it must be, about the Inferred component. The Inferred material makes up just 4% of mill feed during the payback period and 9.4% during the first 20 years, peaking at up to 15% in some years between years 16 and 21. The cautionary statement is blunt: there is a low level of geological confidence associated with Inferred Mineral Resources and no certainty they will convert. The company's position is that the 5% inclusion is immaterial to the economics, and the reserve itself is stated to be not reliant on the Inferred material. Why the Labrador Trough location matters The Iron Bear Project sits in the Labrador Trough, a 1,600 km long belt of iron-rich rocks that has supported mining since 1954. The project comprises eleven licences totalling 11,025 hectares across 441 mineral claims in Newfoundland and Labrador, roughly 30 km northwest of the town of Schefferville and about 1,200 km northeast of Montreal. The mineralisation is described as typical Labrador Trough taconite, a hard, fine-grained rock carrying magnetite and hematite. Location is doing a lot of work in this study, in three specific ways. First, rail and port. The study assumes concentrate moves on existing heavy-haul rail infrastructure to the open-access port of Pointe-Noire, near Sept-Iles on the St Lawrence.(Figure 1). Port of Sept-Iles has an annual cargo handling capacity of 100 million tonnes, which it describes as the largest mineral port in North America. Products are planned to be sold into Rotterdam, Corpus Christi or North Africa. For a bulk commodity, not having to build a railway or a port from scratch is a major point of difference against greenfield peers. Figure 2: Connectivity of Iron Bear Project (Source: ASX Announcement) Second, power. The process plant is designed to run entirely on renewable hydropower from Churchill Falls, approximately 350 km from the mine site, delivered by three 315 kV powerlines added in step with each production stage. The highlights page quotes hydropower at a very low USD 0.0315 per kWh, while the financial model applies a delivered levelised cost of energy (LCOE) of USD 0.148 per kWh, a figure that also recovers the cost of the transmission infrastructure, assumed to be built and owned by a third party. Cheap, clean power flows through to both the cost line and the emissions line: Scope 1 emissions are estimated at 29 kg of CO2 per tonne of sales, which the company describes as among the lowest in the industry. Third, policy. The study leans on a supportive Canadian backdrop: high-purity iron ore was added to Canada's Critical Minerals list in June 2024, the Building Canada Act of June 2025 created an accelerated federal approvals pathway for projects of national significance, and in May 2026 the company applied to the Major Projects Office for that recognition. The mine plan, DR pellets and the green steel story The mining method is conventional and large: open pit, truck and shovel, 15 metre benches, electric rope shovels and 360 tonne class trucks, moving up to roughly 150 Mtpa of material at peak. The life-of-mine strip ratio, the amount of waste moved for every tonne of ore, is just 0.44 to 1. The announcement frames this as driving efficient, low-cost open pit mining, and it is genuinely low for the scale involved. Processing is staged. Three identical process trains, each rated at 8.3 Mtpa of concentrate output, are built in sequence: the first over five years, with trains two and three commissioned in years 5 and 9. Full nameplate production of roughly 24.9 Mtpa of concentrate arrives around year 11. The flowsheet is conventional for magnetite: three-stage crushing, grinding to 32 microns, and wet low-intensity magnetic separation, with a reverse flotation step on trains two and three to lift a portion of the concentrate to Direct Reduction grade. Two straight-grate pellet plants at Pointe-Noire, each with a nominal capacity of about 9.3 Mtpa, turn that DR concentrate into pellets. BF concentrate and DR pellets, decoded. Blast Furnace (BF) concentrate is a high-grade magnetite product, here 69.1% Fe with 3.5% combined silica and alumina, sold to conventional steel mills. Direct Reduction (DR) pellets are the premium product: the concentrate is upgraded to 71% Fe and just 1.2% impurities, then rolled and fired into pellets for Direct Reduced Iron (DRI) steelmaking. DRI plants replace metallurgical coal with natural gas (and eventually hydrogen), and the announcement puts their carbon footprint at half that of the blast furnace route. The catch is that DRI plants need very high-purity feed, above 67 to 68% Fe, which the study says is achievable only with beneficiated magnetite concentrate of the type Iron Bear will produce. At steady state, the product split is 4.5 Mtpa of BF concentrate and 18.4 Mtpa of DR pellets. That weighting toward pellets is the commercial heart of the study. The announcement cites independent estimates of global DR pellet demand at about 180 Mtpa in 2024, growing to 447 Mtpa by 2050 as steelmaking decarbonises, and notes that DR pellets currently command premiums of over USD 70 per tonne above the 62% Fe benchmark. It also points to a regulatory tailwind: the EU's Carbon Border Adjustment Mechanism entered its definitive phase in January 2026, progressively pricing carbon into imported steel through to 2034. One relationship is worth pausing on. Vale, Iron Bear's development partner, is the largest producer of DR pellets globally, controlling approximately 60% of the merchant DR pellet market. The partner funding the studies is also the company that knows this exact product market better than anyone. Iron Bear PFS economics: NPV, IRR and the three scenarios The Base Case financials are built on a discounted cash flow at an 8% nominal weighted average cost of capital (WACC), on an unleveraged, post-tax basis. In plain English: the value of the project's future cash flows, brought back to today's dollars, before any debt structuring, after Canadian taxes, and including tax holidays and credits available in the resource regions of Newfoundland and Labrador and Quebec. The capital and operating cost build-up is presented by stage, and the staging is the point: the study spreads USD 11.2 billion of life-of-mine production capital across three stages over 13 years, so that later trains are funded into an operating business rather than all upfront. Three scenarios, one variable at a time Rather than a single configuration, the PFS models three, each changing exactly one strategic variable so its effect can be read cleanly. Scenario 1 is the Base Case with rail transport and a wet, centreline-raised tailings facility. Scenario 2 replaces rail with a 600 km buried slurry pipeline, including a return-water loop. Scenario 3 keeps rail but swaps to filtered dry-stacked tailings. The slurry pipeline delivers the highest value (post-tax NPV of about USD 9.5 billion in Table 5) and the lowest operating costs, but requires the most capital upfront at USD 6.7 billion pre-production. The dry tailings option is framed as the most environmentally sustainable, removing the reliance on a wet tailings dam, but costs more to build and run, and notably is not covered by the reserve statement because its mine plan and modifying factors are not presented in this announcement. The Base Case is described as the most capital efficient with the lowest capital at risk. Figure 3: PFS Scenarios (Source: ASX Announcements) Who funds the Iron Bear Project? The Vale agreement explained This is the section where the project story and the company story part ways, and readers should hold both in mind at once. The company estimates USD 138 million is needed to reach Decision to Mine (DTM), the point at which construction is committed. Under the binding Development Agreement executed with Vale on 17 February 2025, that entire amount can come from Vale, in two phases. Phase 1 sees Vale contribute USD 18 million to fund the PFS, resource drilling and environmental baseline studies; the announcement notes Vale has contributed A$25.8 million (US$16.7 million) to date. If Vale elects to trigger Phase 2, a joint venture is formed, Vale takes an initial 30% interest, and funds up to a further USD 120 million of development activities including the BFS, environmental impact studies and Impact Benefit Agreements with First Nations. Here is the number that defines the corporate equation: upon conclusion of Phase 2, Vale will have earned a 75% equity stake in the Iron Bear Project, and it is Vale, not IBR, that takes the Decision to Mine. The DTM will not occur until the roughly USD 4.2 billion of pre-production capital has been secured, and the study is explicitly prepared on the assumption that this financing will leverage Vale's financial capacity. For scale, the announcement notes Vale reported capital expenditure of USD 5.5 billion in 2025 and revenues in excess of USD 38 billion. The trade is therefore clear in both directions. IBR shareholders carry a project of this magnitude to DTM with no further study funding required from the company, backed by a partner who is the world's largest DR pellet producer. In exchange, IBR's share of the project reduces to 25% if the agreement runs its full course. The announcement is equally clear about the alternative path: if Vale elects not to proceed to Phase 2, the project remains 100% IBR's, and alternative funding through debt, equity, strategic partnerships or partial divestment would need to be pursued, with no certainty of success. The cautionary statement adds that such strategies could be dilutive or could materially reduce the company's ownership of the project. What happens next: PVI, BFS, permitting and people The development plan targets DTM in four to six years, with the caveat that approvals timelines sit outside the company's control. Before the BFS begins, a Project Value Improvement (PVI) phase will chase specific upside: flowsheet optimisation and further metallurgical variability test work, higher-throughput scenarios that better sweat the rail and pipeline infrastructure, power cost trade-offs, pellet plant location alternatives, and a mine plan that brings more magnetic iron into the early years. An updated Version 2 of the PFS is expected after that work. Permitting is on the critical path. The Environmental and Social Impact Assessment for the mine, concentrator, tailings and power infrastructure is expected to take up to six years from commencement. The company will seek a National Interest designation for the rail component to streamline federal timelines, and permitting for stages 2 and 3 is planned to run in parallel with Stage 1 construction and operations. Just as critical is the social licence work. The study names the five Indigenous groups engaged since October 2024: the Naskapi Nation of Kawawachikamach, Nation Innu Matimekush-Lac John, Innu Takuaikan Uashat mak Mani-utenam, Innu Nation, and NunatuKavut Community Council. Impact Benefit Agreements with these communities are described as a critical component of the environmental approvals, and the company has committed to co-designing the operation with them, including cleaning up legacy drill sites. The PFS risk register identifies 26 project risks, 16 of them rated High, with permitting and social acceptability, groundwater and tailings, processing and infrastructure delivery, and funding among the key categories. What the PFS means for Iron Bear Resources Strip the PFS document back and three facts carry the story. The resource is enormous and now partially converted to a maiden reserve that supports 44 years of production while touching less than a quarter of the deposit. The product strategy is aimed squarely at DR pellets, the fastest-growing and highest-premium corner of the iron ore market, with cheap hydropower underwriting both the cost position and the green-steel credentials. And the funding structure means the project's fate now rests substantially on Vale's Phase 2 election and, ultimately, Vale's Decision to Mine. The company's own cautionary statement deserves the last factual word: the PFS is based on material assumptions that remain subject to refinement, there is no certainty the outcomes will be realised, approximately USD 4.2 billion of funding will likely be required, and investors should not make investment decisions based solely on the results of the study. Summary What did Iron Bear Resources (ASX: IBR) announce? The maiden Pre-Feasibility Study for the Iron Bear Project, including a maiden Probable Ore Reserve of 3.3 billion tonnes at 29.1% Fe, a post-tax NPV of USD 9.0 billion, a 15.2% post-tax IRR and a 44-year mine life producing 23 Mtpa of high-grade iron ore products. How big is the Iron Bear mineral resource? 13.6 billion tonnes at 30.03% Fe under JORC 2012, comprising 4.5 billion tonnes Indicated and 9.1 billion tonnes Inferred at a 12.5% magnetic Fe cut-off. The current mine plan draws on less than 25% of it. How much will the Iron Bear Project cost to build? Approximately USD 4.25 billion in pre-production capital for the Base Case, within USD 11.2 billion of life-of-mine production CAPEX spread across three stages, estimated to AACE Class 4 accuracy of roughly plus or minus 30%. Who is funding the Iron Bear Project? Vale S.A., under a binding Development Agreement signed in February 2025, can fund up to USD 138 million through to Decision to Mine and in doing so earn a 75% stake in the project. The study assumes the USD 4.2 billion construction bill will then leverage Vale's financial capacity. If Vale does not proceed to Phase 2, IBR retains 100% and must find alternative funding. What are DR pellets and why do they matter? Direct Reduction pellets are a very high-purity iron ore product (here made from a 71% Fe concentrate) used in Direct Reduced Iron steelmaking, which the announcement says has half the carbon footprint of the blast furnace route. The study cites forecast demand growth from about 180 Mtpa in 2024 to 447 Mtpa by 2050, with current premiums above USD 70/t over the 62% Fe benchmark. When could Iron Bear reach a Decision to Mine? The plan targets four to six years, subject to approvals. The environmental assessment alone is expected to take up to six years from commencement, and a Project Value Improvement phase plus a Bankable Feasibility Study sit between here and any construction decision. Samso Concluding Comments For shareholders, like me, this is a significant step in realising the potential of the Iron Bear story. This is what the market needs to see and I dont think there are any surprises. Iron Bear management has been telling everyone that this is a major project and the release of the PFS is prove. The startegy of tagging Vale as a partner is the reason why management got that startegy right from the start. You need a big brother to come along this journey. How many mineral resource stories have we heard that their project is Tier-1 but ther eis no realistic funding strategy. How many times have we seen great project get stalled because of funding. I think this sets a lot of doubts aside and with the DR pellet potential. the incentive for the "real" players of the iron ore industry to get along the ride should be starting to be real. Time for some serious DYOR in my opinion for those that have not taken positions. With a current market capitalisation of AUD 77M, Iron Bear Resources is way too undervalued in the scheme of things. As I always say, DYOR and do your own thorough check. The Samso Way – Seek the Research Here at Samso, we pride ourselves on delivering content for investors that is independent and informed by over three decades of experience in the industry. Our content is well-researched and is only created if I see merit in discussing the company's story. Our mission is simple: cut through the noise and spotlight what matters—genuine stories, grounded insights, and real opportunity. Our content is well-researched and is only created if the team sees merit in discussing the company or concept. Investors can explore our three core platforms: Coffee with Samso Samso Insights Samso News There may be numerous paths to success in investing, but the common thread among successful individuals is that they remain committed to making informed decisions. Equip yourself with the right knowledge and tools, and you will be well on your way to achieving your financial goals. Most importantly, investors need to be absolutely diligent in understanding their own risk-reward tolerance and capabilities. Never bite off more than you can chew. As they say, Rome wasn’t built in a day, and the Great Wall stood because it took centuries to complete. The Samso Philosophy: Stay curious. Stay sharp. And remember—digging deeper always uncovers the real value. In Life, there is no such thing as a Free Lunch. Never bite off more than you can chew is my parting comment. Happy Investing, and the only four-letter word you need to know is DYOR. To support our independent nature of our work, please head over to our Support Page and give us a helping hand in any of the ways listed. This is a new initiate for the Samso Platform, and it was always the concept of Samso when we started this journey in 2018. Disclaimer The information or opinions provided herein do not constitute investment advice, an offer, or solicitation to subscribe for, purchase, or sell the investment product(s) mentioned herein. It does not take into consideration, nor have any regard to your specific investment objectives, financial situation, risk profile, tax position and particular, or unique needs and constraints. Read full Disclaimer. If you find this article informative and useful, please help me share the information. I try to write about topics that are interesting and have the potential to be of investment value. It is not easy to find stories that fit those parameters. If you or your organisation sees the benefit of what Samso is trying to achieve and has a need to share your journey, please contact me at noel.ong@samso.com.au. About Samso Samso is a trusted platform that equips dedicated investors with up-to-date industry knowledge and insigh0ts from top CEOs and thought leaders. By staying informed on business advancements and market trends, investors can enhance their financial decisions through a combination of expert guidance and their own research. Samso News | www.samso.com.au | An Investor Lens on ASX-Listed Companies

A first-time explorer wants to raise $6 million to hunt for gold and copper across two historic grounds of Western Australia. Gwardar Resources Limited is asking investors for at least $5 million, and up to $6 million, to list on the Australian Securities Exchange and hunt for gold and copper in Western Australia's historic gold fields. The company plans to issue 25 million shares at 20 cents each to raise the minimum, with room to accept another 5 million shares if demand is strong. If it lists, it will trade under the code GRS, with an indicative market value of roughly $5.6 million to $6.6 million on day one. The pitch is straightforward. Gwardar Resources IPO lined up the right to buy, outright, two exploration projects in Western Australia: the Doolgunna Project in the Murchison region, prospective for copper and gold, and the Kurnalpi Project in the Eastern Goldfields near Kalgoorlie, prospective for gold. Both sit in districts that have produced metal for decades. Neither has been drilled by Gwardar yet. In fact, the company was only incorporated on 17 December 2025 and has no revenue or operating history. That combination, established mining postcodes paired with a brand-new company and unproven ground, is the whole story. The rest of this article walks through the numbers, the two projects, where the money goes, who is running it, and the risks the prospectus asks you to weigh. Section 1 What is being offered- Gwardar Resources IPO The core offer is 25 million shares at 20 cents to raise $5 million before costs, which the prospectus calls the Minimum Subscription. The company can also accept oversubscriptions of up to a further 5 million shares to raise an extra $1 million, taking the maximum to $6 million (the Maximum Subscription). Applications start at $2,000, or 10,000 shares, and rise in parcels of 2,500 shares after that. Alongside the main offer sits a smaller, separate “Secondary Offer” of 2 million shares. These are not for the public. They go to the vendors who are selling the two projects to Gwardar, as part of how the company is paying for them, and are covered here so those shares can trade freely once the escrow period ends. The offer is conditional. It proceeds only if three conditions are met: the $5 million minimum is raised, the ASX agrees to admit the company, and the two project purchase agreements become unconditional. If any of these fall over, the offer does not proceed, and the application money is refunded without interest. Table 1: Key offer statistics Section 2 The two projects Everything Gwardar is raising money to do centres on two project areas, both in Western Australia, and both being bought in full from their current owners. The map below shows how far apart they are: Doolgunna sits in the Murchison, well to the north, while Kurnalpi lies out east near Kalgoorlie (Figure 1). Figure 1: Location of Gwardar’s Doolgunna and Kurnalpi projects in Western Australia. Both sit within “greenstone” belts (shown in green), the ancient rock formations that host most of the state’s gold. (Source: Gwardar Resources prospectus) Doolgunna: chasing a DeGrussa-style copper story The Doolgunna Project sits in the Murchison Mineral Field, about 750 kilometres north-east of Perth and 100 kilometres north-east of the town of Meekatharra. It is made up of four exploration licences covering roughly 372 square kilometres, or 120 graticular blocks. (Figure 2). Only one of those four licences has actually been granted; the other three are still applications, which the prospectus flags as a risk. Figure 2. The Doolgunna Project licences (outlined) in relation to the DeGrussa mine and known gold and copper occurrences (yellow and orange dots). (Source: Gwardar Resources prospectus) Geologically, this is copper-and-gold country. The area is known for what the industry calls VHMS deposits, short for volcanic-hosted massive sulphide, a type of ore body rich in copper and gold that forms around ancient underwater volcanic vents. The region’s headline example is DeGrussa, a high-grade copper deposit discovered by Sandfire Resources in 2009 that reshaped exploration across the district. Gwardar’s ground is described as prospective for the same style of mineralisation, and for structurally controlled gold as well. Prospective is the key word: it means the geology looks favourable, not that a deposit has been found. What has past drilling actually found? On the granted licence, historical reverse-circulation drilling at a prospect called Halloween West returned intercepts including 6 metres at 561.7 ppm copper, 4 metres at 0.18 g/t gold, and 6 metres at 1,029 ppm copper. These are modest, early-stage numbers that point to a mineralised system worth testing further, rather than an economic discovery. The prospectus is careful to frame Doolgunna as an early-stage project inside a mature, well-explored region. Figure 3. Regional geology and mineral occurrences around Doolgunna. The district hosts several known copper-gold (VHMS) deposits and numerous fault-controlled gold occurrences. (Source: Gwardar Resources prospectus) Table 2: Doolgunna Project Tenements Kurnalpi: under-explored ground near Kalgoorlie The Kurnalpi Project lies in the Eastern Goldfields, about 90 kilometres north-east of Kalgoorlie, reached by sealed road to Kanowna and then gravel. It covers roughly 90 square kilometres, or 29 graticular blocks, across two granted licences and one pending application. This is squarely gold country: the surrounding district is what the prospectus calls a “world-class” gold and nickel province, with major mines within 50 kilometres. Figure 4. The Kurnalpi Project (outlined) north-east of Kalgoorlie, among the well-known deposits of the Eastern Goldfields. (Source: Gwardar Resources prospectus) Here the target is orogenic gold, the classic Kalgoorlie style of gold that forms in fault and shear zones deep in the earth’s crust. Table 3: Kurnalpi Project Tenements Gwardar argues that the ground has been overlooked: it is buried under a thick blanket of younger surface material, which blunted the older, shallow sampling techniques earlier explorers relied on. In other words, previous crews may have looked, but not looked properly. “The Company considers the Project area is inadequately explored for gold using contemporary techniques.” There is something to build on. Earlier work outlined a gold-anomalous zone about 600 metres long, known as the SW003 or “Cessna 3” anomaly. A 2012 drilling program by a company called Pioneer put down 97 shallow holes for 4,309 metres, with a best result of 3 metres at 0.91 g/t gold from 15 metres depth. The company sees that anomaly as a priority worth drilling with modern methods. Figure 5. Historical gold-in-soil results and drill intercepts over the Kurnalpi (Cessna) target. Warmer colours show stronger gold-in-soil readings; labels show past drill hits such as 15 m at 0.291 g/t gold. (Source: Gwardar Resources prospectus.) Section 3 How Gwardar is buying the projects Here is a detail that matters: as at the prospectus date, Gwardar did not yet own either project. It had signed binding agreements on 3 June 2026 to acquire 100 per cent of both, but the purchases only complete (“settle”) once conditions are met, including the company successfully listing. So buyers of shares are backing a company that is acquiring its assets at the same time as it floats. The prices are small, and mostly paid in shares rather than cash. For Doolgunna, Gwardar agreed to issue 1 million shares (valued at $200,000 at the offer price) plus up to $50,000 cash, to vendors Cuvier Resources and Tasex Geological Services. For Kurnalpi, it agreed to issue another 1 million shares ($200,000) plus up to $100,000 cash to Australian Nickel Company Limited. Those 2 million vendor shares are the “Secondary Offer” mentioned earlier, and they will be subject to ASX escrow. Table 4: Project Acquisition terms Section 4 Where the money goes If the raise lands at the $5 million minimum, a little under 56 cents in every dollar goes directly into exploring the two projects. Drilling, sampling and geophysics dominate the two-year work plan. The rest covers the cost of the float itself, corporate and administration costs, the small cash portion of the acquisitions, and a working-capital buffer. Table 5: Use of funds over the next two years The exploration budget itself, about $2.86 million at the minimum raise and $3.41 million at the maximum, is spread across data review, land access and heritage work, geochemistry, geophysics and drilling. Drilling is the single biggest line item, which is what you would expect from a company whose whole reason for listing is to test whether these two areas hold something worth mining. Table 6: Two-year exploration budget by activity (minimum raise) Section 5 Who is running it Gwardar is led by a three-person, all-non-executive board, with day-to-day geology handled by a consultant principal geologist. Several of the directors hold current roles at other ASX-listed resource companies, which the prospectus presents as a sign of relevant experience. The three directors are each on $40,000 a year in fees, starting only once the company lists. Each already holds about 100,000 shares from the pre-float seed round, roughly 10 per cent each of the tiny existing share register, which shrinks to a fraction of a per cent once the new public shares are issued. All three have signalled an intention to buy more under the offer: Mr Hardcastle up to $100,000, and Mr Rovira and Mr El Sayed up to $350,000 each, subject to how the offer is allocated. Table 7: Two-year exploration budget by activity (minimum raise) Section 6 The finances, and the risks There is not much of a financial history to examine. Gwardar was incorporated in December 2025 and, to the end of that month, recorded a loss of just $5,491 and held $100 in assets. It has no revenue and does not expect to pay dividends for the foreseeable future. Table 8: Pro Forma Financial Position at Listing On a pro forma basis, that is, adjusted to show the position as if the raise and acquisitions had already happened, the company would hold roughly $5 million in cash at the minimum raise, or about $6 million at the maximum, and net assets of about $4.99 million to $5.93 million. The prospectus does not shy away from risk; it devotes a long section to it and tells readers plainly that the investment is speculative. The most important risks for a would-be investor to understand are these: No track record. The company is months old, has never explored its ground, and has no revenue. It expects to keep making losses until, and unless, it finds something valuable. It does not own the projects yet. The acquisitions must still settle. If a vendor fails to complete, Gwardar may not end up with the project, or may get it on different terms. Several tenements are only applications. Three of the four Doolgunna licences and one Kurnalpi licence are pending. They may not be granted, or only partly granted. The prospectus singles out E27/676 as facing a competing prior claim. Exploration usually fails. Most exploration ground never becomes a mine. Even a promising result may prove uneconomic to dig up. Native title and heritage. The tenements overlap registered native title claims and land-use agreements. Access, approvals and heritage surveys can delay or restrict work. More money will likely be needed. The cash is budgeted for about two years. Beyond that, Gwardar will probably need to raise again, which can dilute existing shareholders, and there is no guarantee funding will be available. Commodity prices and thin trading. Returns depend on gold and copper prices the company cannot control, and small explorers often trade in low volumes, which can make shares hard to sell at a good price. Section 7 The Samso Concluding Comments - The bottom line Strip away the geology and Gwardar Resources is a familiar kind of proposition: a small, first-time explorer raising a modest sum to test two parcels of ground in proven mining districts. The appeal, as the prospectus frames it, is a portfolio of gold and copper ground in established Western Australian provinces, an experienced board, and enough cash to run two years of exploration. The catch, which the prospectus states just as clearly, is that nothing has been found yet, the company does not formally own the projects at the point of listing, several licences are still applications, and the whole thing is expressly “highly speculative.” For a retail investor, the practical takeaway is that this is your typical mineral explorer that is really a listed vehicle for now. Some credentialed names on the Board and Karl Jupp is a experienced campaigner. The common feature of early mienral explorer with no glaring upside is to try and see if you are investing in the future and going for that "punt" or for managemnet to bring in a glory project in the coming future. The issue with the second option is that the IPO price may be "expensive" as time normally dewvalue the IPO price. Waiting and picking something up later at a market discount could be the play. However, as most of us retail players don't have the wisdom of the "purple circle" this is always hard to decide. For now, I think a wait and see is a safer option but if you feel that there is a FOMO happening, take a punt. The good news is that you can always average down if things take a turn for the worst. The Samso Way – Seek the Research Here at Samso, we pride ourselves on delivering content for investors that is independent and informed by over three decades of experience in the industry. Our content is well-researched and is only created if I see merit in discussing the company's story. Our mission is simple: cut through the noise and spotlight what matters—genuine stories, grounded insights, and real opportunity. Our content is well-researched and is only created if the team sees merit in discussing the company or concept. Investors can explore our three core platforms: Coffee with Samso Samso Insights Samso News There may be numerous paths to success in investing, but the common thread among successful individuals is that they remain committed to making informed decisions. Equip yourself with the right knowledge and tools, and you will be well on your way to achieving your financial goals. Most importantly, investors need to be absolutely diligent in understanding their own risk-reward tolerance and capabilities. Never bite off more than you can chew. As they say, Rome wasn’t built in a day, and the Great Wall stood because it took centuries to complete. The Samso Philosophy: Stay curious. Stay sharp. And remember—digging deeper always uncovers the real value. In Life, there is no such thing as a Free Lunch. Never bite off more than you can chew is my parting comment. Happy Investing, and the only four-letter word you need to know is DYOR. To support our independent nature of our work, please head over to our Support Page and give us a helping hand in any of the ways listed. This is a new initiate for the Samso Platform, and it was always the concept of Samso when we started this journey in 2018. Disclaimer The information or opinions provided herein do not constitute investment advice, an offer, or solicitation to subscribe for, purchase, or sell the investment product(s) mentioned herein. It does not take into consideration, nor have any regard to your specific investment objectives, financial situation, risk profile, tax position and particular, or unique needs and constraints. Read full Disclaimer. If you find this article informative and useful, please help me share the information. I try to write about topics that are interesting and have the potential to be of investment value. It is not easy to find stories that fit those parameters. If you or your organisation sees the benefit of what Samso is trying to achieve and has a need to share your journey, please contact me at noel.ong@samso.com.au. About Samso Samso is a trusted platform that equips dedicated investors with up-to-date industry knowledge and insigh0ts from top CEOs and thought leaders. By staying informed on business advancements and market trends, investors can enhance their financial decisions through a combination of expert guidance and their own research. Samso News | www.samso.com.au | An Investor Lens on ASX-Listed Companies