Global Atomic

Phase A — Understand the business

Company Overview

Global Atomic is not one company — it is a pre-production uranium developer bolted onto a small, cash-generating zinc-recycling annuity, and you cannot value it correctly without holding both in view at once.

Arm 1 — Uranium (the story, ~95% of the equity value). Global Atomic owns 80% of the Dasa Project in Niger, held through the operating company SOMIDA S.A. (Société des Mines de Dasa); the Republic of Niger holds the other 20% — 10% free-carried and 10% fully-paid (obligated to fund its share of capexCapital expenditureMoney spent on long-lived things — buildings, machines, servers — rather than on running costs. and operating deficits). Dasa is a sandstone-hosted, high-grade underground uranium deposit — per the company, the highest-grade uranium deposit in Africa, with grades bettered globally only by Canada's Athabasca Basin. As of July 2026 it is not yet in production: it is a construction project chasing its final funding tranche.

Arm 2 — Zinc (the ballast, the reason it isn't dead). Global Atomic holds 49% of the Befesa Silvermet Turkey (BST) JV, a modern zinc-recycling plant in Iskenderun, Türkiye that recovers zinc from electric-arc-furnace dust (EAFD) from local steel mills and sells high-grade zinc-oxide concentrate to smelters. Befesa Zinc S.A.U. owns 51% and operates it. In 2025 the JV retired its plant debt and resumed dividends (December 2025); it is now debt-free and expected to keep paying. This arm is the thing that lets Global Atomic fund corporate G&A and part of its equity contribution without diluting to zero.

How it actually makes money (today vs. tomorrow). Today: it makes almost nothing at the corporate line — it books its 49% equity share of BST and collects a small management fee, and otherwise burns cash on Dasa development. Tomorrow (if Dasa is built): it becomes a ~4–4.4 Mlb/yr U3O8 producer selling into a mix of term offtake contracts (blended spot+term pricing with escalation) and spot. It has four offtake agreements with US and European utilities — ~8.8 million lbs over the first seven operating years, ~90% to US utilities. Contract structure is the good kind (term utilities, escalators) but the volume committed is a fraction of the 68 Mlb reserve, leaving most output exposed to the spot cycle.

Contract/payment structure verdict: the zinc side is a clean, take-what-the-smelter-pays commodity flow with a JV dividend now switched back on; the uranium side is pre-revenue with a partially-contracted book — the payment terms are fine, but there is no payment yet, and there won't be until 2028 on the current schedule (see Lens 5).

Supply Chain

Two entirely separate chains — name every node:

Uranium chain (Dasa → nuclear fuel):

• Upstream inputs / build phase: mining equipment and processing plant (the mill is the long pole — see Lens 5), diesel/logistics into a landlocked country (Niger). Logistics cost inflation is an explicitly-cited reason capex rose. Historic export route ran north through Arlit; the company has flagged a new Niger–Algeria cooperation corridor to the Mediterranean as a potential alternative outlet — meaningful because the traditional French-linked logistics chain is politically compromised post-coup.
• The asset: SOMIDA (80% GLO / 20% Niger State) mines and mills to yellowcake (U3O8) on site.
• Midstream (the part GLO does NOT own): yellowcake must be converted (UF6) → enriched → fabricated before it reaches a reactor. Global Atomic sells yellowcake; it captures none of the conversion/enrichment margin. Conversion/enrichment is itself a Western bottleneck (Orano, Urenco, Cameco, and — the geopolitics — Rosatom/Tenex), which is bullish for uranium price but irrelevant to GLO's own take.
• End customers (named): the offtake counterparties are US and European nuclear utilities (the company names them as "American and European nuclear power utilities"; individual utility names not disclosed in the public releases — n/a — not individually disclosed). ~90% US-weighted.
• Chokepoint / single-source dependency: the entire uranium arm is one mine, in one country, with one government as its 20% partner and mining-code counterparty. There is no diversification. This is the whole risk.

Zinc chain (steel dust → zinc oxide):

• Upstream: EAFD feedstock from local Turkish steel mills (the recycler is paid to take a hazardous waste, then sells the recovered metal — a structurally attractive two-sided position).
• The asset: BST plant, Iskenderun (Befesa operates).
• Downstream (named node type): zinc smelters worldwide buy the zinc-oxide concentrate.
• Chokepoint: GLO is a non-operating 49% minority — it does not control the plant, the feedstock contracts, or the dividend policy; Befesa does.

Competitive Advantages (moats)

The moat is the rock, and it is real — up to the mine gate.

• Grade is the durable edge. Dasa reserve grade ~4,113 ppm (early-mine-plan feed ~5,109 ppm) drives a cash cost of ~US$30.73/lb and AISC ~US$35.70/lb per the 2024 FS. Against a spot ~US$86/lb and term ~US$93/lb, that is a lower-half-of-the-cost-curve asset. High grade = lower tonnes moved per pound = structurally lower cost and lower execution surface area. That advantage is geological and cannot be competed away.
• Scarcity of near-term new supply. The uranium bull case is a supply-deficit case; greenfield mines that can actually be financed and built are rare, and Dasa is one of the few in active construction. That scarcity is a genuine relative advantage among developers.
• Where the moat evaporates: everything downstream of the orebody. GLO has no pricing power (it's a price-taker selling yellowcake), no bargaining power over its host government (the government is its partner and its regulator and has just demonstrated, via Orano, that it will nationalise), and no bargaining power over Befesa on the zinc side (51% operator controls the JV). Its bargaining power over equipment suppliers is that of a small single-project buyer — weak.
• Net: a first-quartile-cost orebody is a strong moat for a producer. Global Atomic is not a producer yet, and the moat does not protect it from the two things that can actually kill it — funding and sovereignty. A great rock in a bad jurisdiction with no cash is a moat you can't yet monetise.

Segments

No our figures exists , so this is /.

Read: annualising Q1-2026, the zinc JV is throwing off on the order of ~US$8M/yr of net income and ~US$14M/yr of EBITDA to GLO's 49%. That is a rounding error against US$290M of remaining Dasa capex — useful as G&A/interim ballast, immaterial as a funding source. The segment story is binary: ~100% of the equity's future value is one pre-revenue segment, and the profitable segment is too small to move the needle.

---

Phase B — Measure performance

This name straddles the operating/developer line. It has no P&L for its flagship asset, so Phase B blends the operating battery (zinc has real numbers) with a developer's milestone/funding read (uranium's "performance" is capex, schedule, and the financing gap). I run it that way explicitly.

Earnings Result → Development & funding status (the real "print")

For a developer, the print is capex vs. schedule vs. cash, not EPS. The latest disclosed state (Q1-2026, released May-2026):

• Total Dasa project cost: US$501.3M (up from earlier estimates — "increased logistics costs, general cost inflation and the passage of time").
• Incurred to March 31, 2026: US$210.9M. Remaining: US$290.4M.
• Cash (Mar 31, 2026): C$59.1M — then topped up by a C$72.5M raise on Feb 2, 2026 at C$0.88/unit. (Ordering note: the C$72.5M raise is dated Feb-2026 and the C$59.1M is the Mar-31 balance — treat pro-forma cash as roughly the balance after the raise, i.e. ~C$59M already reflects it..)
• Zinc JV (49%): EBITDA share US$3.6M, net income share US$2.1M, mgmt fee US$0.3M.
• Offtake: 4 agreements, ~8.8 Mlb / first 7 yrs, ~90% US utilities.

The one fact that dominates everything — the schedule slipped ~2 years:

• A year ago the narrative was "production Q1-2026, commissioning H2-2026".
• The Q1-2026 results release now states process-plant commissioning Q4-2027 and first yellowcake shipments H1-2028.
• The "2025 Summary / 2026 Outlook" even carries an "H2-2027" commissioning line.

That is a material, self-disclosed slippage of roughly 18–24 months on first revenue, and it is the single most important number in this dossier — because a single-asset developer's equity value is a function of time-to-cash-flow discounted for the risk that it never arrives. Every month of slip compounds dilutionDilutionIssuing new shares, so each existing share owns a smaller slice of the same company. risk (see Lens 11) and political risk (more time for the junta to change its mind).

Balance-sheet / going-concern flags: the releases I could reach do not carry explicit going-concern language, and the C$72.5M raise + zinc dividends buy runwayRunwayHow long the cash lasts at the current rate of spending. It shortens the moment spending rises, which is why a figure taken from a quiet quarter flatters. — but with US$290M still to fund and the DFC loan unapproved, the company is structurally dependent on financing it does not yet control. That is the flag.

Earnings Calls (sentiment trend)

No transcripts on the shelf ; read from the cadence of releases:

• The recurring phrase that keeps getting repeated — and keeps not resolving — is "DFC funding is advancing." Late-2024: "loan in early 2025". Early-2025: "next steps over the next two months / forwarded to Investment Committee". Dec-2025: "internal process nearing completion… year-end approval timeline". Q1-2026: DFC now described merely as "active engagement" with no terms or timeline confirmed.
• The tone has quietly degraded from "imminent" to "engaged." That downgrade in specificity — from committee dates to "active engagement" — is the sentiment tell. Management has stopped giving a close date, which is either honesty or a bad sign; on a name with a live class action about over-promising the DFC, it is likely forced honesty.
• What they now lean on instead: Niger government support ("strong government support," CEO site visits, SOMIDA board optics, the Algeria corridor) and the zinc dividend restart. When a developer pivots its narrative from financing certainty to political reassurance, the market should mark down the financing certainty — which it has (see Lens 8).

Comps

Read: the peer set trades ~0.8–1.0× NAV; GLO's spot-implied P/NAV of ~0.3× is not a bargain-hunter's mispricing — it is the market correctly pricing a jurisdiction/funding discount. The comp that should keep a bull up at night is Bannerman: a same-cohort African developer that just closed a US$321.5M strategic-lender package in Namibia, while GLO's US$295M DFC package remains unapproved after >18 months. The discount is earned; the question (Lens 12) is whether it's over-earned.

Stock-Price Catalysts (what moves it >5%)

The tape tells you exactly what this stock is: a Niger political-risk and DFC-financing option, not a uranium-price play. Moves >5% cluster on:

• Niger coup, July 2023 → the defining event. Shares fell from ~C$5 to ~C$0.50 over the coup and its aftermath — a ~90% drawdown. Nothing else in the chart matters as much.
• US troop expulsion from Niger → "Global Atomic stock plunges as Niger's junta expels US troops" — because it directly threatens the US DFC loan.
• DFC milestone headlines (Credit Committee → Investment Committee → "year-end") → repeated pops and fades; each unfulfilled date bleeds credibility.
• "Niger reassurance" / no-nationalisation statements → sharp relief rallies ("shares jump on Niger project reassurance," "no plans to nationalise Dasa, minister says").
• Orano's SOMAÏR nationalisation (2025) → sector read-through fear for Dasa.
• The class-action alert (Jan-2026) → shares -3.4% on the day.

What the market reacts to: Niger sovereignty risk and US financing risk, in that order. It is remarkably insensitive to the uranium price — spot at an 18-year high has not re-rated the stock, because the market is (correctly) saying "grade doesn't matter until it's funded and out of the ground." That is the whole investment debate in one observation.

---

Phase C — Judge people & books

Management

Founder-led, deeply experienced, and carrying a specific, relevant scar.

• Stephen G. Roman — Founder, Chairman, President & CEO. 40+ years in mining/energy finance and development. The résumé headline: founder/Co-Chairman of Gold Eagle Mines, acquired by Goldcorp for ~C$1.5B in 2008 — a genuine, large, realised win. PDAC "Bill Dennis Award" for Canadian mineral discovery (2016). Prolific promoter/creator (Gabriel Resources, Belo Sun, Silvermet, Polar Star). This is a capital-markets operator who can find rock and raise money — exactly the skill a developer needs, and exactly the skill that can also over-promise.
• Skin in the game — genuine and recently added. Roman directly owns ~3.22% (~C$9.4M) and insiders ~10% (~C$9.0M). He bought more on the way down — a 3.8% increase, ~C$270k at ~C$0.60. Insider buying into a beaten-down developer is a real positive signal; it is not a small token.
• Capital-allocation history — mixed, and the recent chapter is bad. The decisive red flag is Harte Gold. Roman chaired Harte Gold, which rushed its Sugar Zone mine into production in 2019 without a feasibility study, blew production costs out to >US$2,000/oz (above the gold price), and filed for CCAA in December 2021 — common shareholders were wiped out, including major institutions (Appian, Goodman, Mackenzie). That is not a distant footnote: it is the same CEO, running another single-asset development story, where the core risk is rushing/underfunding a build. The one mitigant: Dasa, unlike Harte, does have a completed 2024 NI 43-101 feasibility study — Roman may be running the anti-Harte playbook deliberately. But the pattern (promote → build single asset → cost/finance stress) is on the record.
• Founder vs. professional manager: unambiguously founder/promoter archetype. For a discovery-and-finance stage that is an asset; for the disciplined-execution phase Dasa is now in, it is exactly the archetype whose last execution ended in a wipeout. Watch the execution, not the pitch.

Forensic Red Flags

our figures is empty , so statement-level forensics are limited to what releases disclose. The material risks:

1. Serial equity dilution — the defining forensic feature. Track the raises: C$15M (Jan-2024) → C$35–37M (Oct-2025, C$0.62 units + C$0.80 warrants to 2028) → C$72.5M (Feb-2026, C$0.88), plus a US$50M ATM program. Share count is ~490M. Every DFC delay is funded by selling equity into a depressed price — the shareholder is the lender of last resort, and the warrant overhang (C$0.80 strike) caps upside. This is the mechanism by which a "great orebody" can still lose common holders money.
2. Capitalised development costs are the whole balance sheet. US$210.9M is sitting in Dasa construction-in-progress. If the project is delayed indefinitely or the jurisdiction turns, that carrying value faces impairment — there is no operating cash flow to defend it. (No impairment taken to date.)
3. Equity-method opacity on the profit line. The only earnings are the 49% equity pick-up from BST, a JV operated by Befesa — GLO takes the number Befesa's plant produces; it does not control the accounting or the dividend. Not fraud-flavoured, but it means the sole positive line is a minority interest GLO can't steer.
4. Timeline restatements. The commissioning date moving from H2-2026 → H2-2027/Q4-2027 → first-pour-H1-2028 across successive disclosures is the operational analogue of an earnings restatement — the number the market underwrote keeps changing.

Regulatory findings (required sub-section).

• SEC (EDGAR EFTS — LR + AAER): regulatory/regulatory-findings.md (fetched 2026-07-07) confirms Global Atomic has no CIK and is not an SEC filer — zero EDGAR enforcement results, and none possible.
• Securities litigation (Ontario) — MATERIAL, ACTIVE: Berger Montague filed a statement of claim in the Ontario Superior Court of Justice alleging misrepresentations regarding the US$295M DFC facility — specifically that Global Atomic "failed to disclose certain pre-conditions of the loan and developments that could affect the financing process, such as deteriorating US-Niger relationships." Class period: November 10, 2023 – January 23, 2025. The claim does not yet have leave under s.138.8 of Ontario's Securities Act and must be certified to proceed; no damages amount specified; shares fell ~3.4% on the alert. This directly attacks the credibility of the central catalyst (the DFC loan) and of management's disclosure — it is the most important red flag in this dossier alongside the Harte precedent.
• Non-SEC / host-country: no FTC/DOJ/FDA-type action found (not a US-operating consumer company). The dominant host-country regulatory risk is Niger's mining-code overhaul and nationalisation posture (Orano's SOMAÏR nationalised 2025; junta "overhauling mining codes and renegotiating contracts") — this is a sovereign/expropriation risk, covered in Lenses 12/13.
• 10-K Item 3: n/a — no SEC filer; Canadian AIF/MD&A legal-proceedings disclosure not machine-read here.
• Summary: One active securities class-action claim (pre-certification) directly on the DFC disclosure, no SEC enforcement (none possible), and a severe standing host-country expropriation risk. Not a clean bill.

---

Phase D — Project & stress-test

Forward Projection

A conventional 3-year EPS ramp is the wrong model — Global Atomic has no earnings and, on its own current schedule, first yellowcake is H1-2028, so FY2026 and FY2027 are build years with no uranium revenue. The honest projection is a milestone/value-realisation path, not an EPS line. (No our model create in this unattended watchlist run, per skill rules.)

Base / Bull / Bear — framed on the two variables that actually decide it (funding + Niger), with the FS as the terminal value.

Anchors (all unless noted): FS after-tax NPV8 US$917M @ US$75/lb (IRR 57%, payback 2.2 yr); US$1.62B @ US$105/lb; 68.1 Mlb over ~23.75 yr; cash cost US$30.73/lb, AISC US$35.70/lb; US$290.4M remaining capex; DFC US$295M (60% debt) unapproved; spot ~US$86/lb, term ~US$93/lb; mkt cap ~US$225M; ~490M shares.

• BEAR (real, and not the tail): DFC approval slips again or dies (US-Niger relations, class-action chill, DFC bandwidth). GLO keeps funding on equity + ATM at depressed prices; share count runs well past 550–600M; commissioning slips past 2028. Or the junta applies the Orano playbook to Dasa (renegotiation, higher state take, or worse). Equity re-rates toward option value on a stranded asset; a 50%+ drawdown from here is plausible if either the loan dies or Niger turns.
• BASE: DFC closes in late-2026/H1-2027 (a further slip from the last "year-end" guide is the base, given the track record), the C$72.5M + ATM + zinc dividends bridge the equity portion, first pour H1-2028 roughly on the restated schedule. As financing de-risks and construction completes, the P/NAV discount narrows from ~0.3× toward ~0.5–0.6× — implying a materially higher equity value even at a still-discounted NAV, offset by dilution. Net: meaningful upside, but back-end-loaded and dilution-taxed.
• BULL: DFC closes in 2026; the Rubio-chaired, Trump-administration DFC treats a US-utility-supplying, non-Russian, non-Chinese African uranium mine as strategically desirable (the "friend-shoring critical minerals" thesis); Niger stays supportive (it wants the royalties/jobs — "largest active mining project in Niger"). First pour on schedule into US$90–100+/lb term uranium. NPV re-rates toward the US$1.6B @ US$105/lb case; on GLO's 80% that is ~US$1.3B attributable vs. ~US$225M today — a multi-bagger if every domino falls.

The projection's honest core: this is a binary, path-dependent option on (1) the DFC loan closing and (2) Niger not expropriating — with a first-quartile orebody and a supportive uranium tape as the payoff, and serial dilution + an ~18-month execution slip as the cost of waiting. The base case is up; the distribution is wide and left-skewed.

Bull vs Bear

Bull case (narrative). Dasa is a genuinely tier-1 asset — the highest-grade uranium deposit in Africa, first-quartile cash costs (~US$31/lb) against an 18-year-high term price (~US$93/lb), a 68 Mlb reserve with 51 Mlb of inferred upside, and a 23-year life. The world is short new uranium and shorter on financeable, near-term new mines; Dasa is one of a handful actually in construction. Management has a real M&A win (Gold Eagle → Goldcorp, C$1.5B) and is buying stock personally. The DFC package is championed by a Trump-administration DFC (Rubio as chair) that has every strategic reason to fund a US-utility-supplying, non-Russian, non-Chinese African uranium mine. A zinc JV — now debt-free and paying dividends — funds the lights. Close the loan and the ~0.3× P/NAV discount is absurd; this re-rates hard.

Bear case (2–3 things that permanently impair).

1. Sovereign expropriation. Niger's junta just nationalised Orano's SOMAÏR and is overhauling the mining code. GLO's "strong government support" is the same relationship Orano had before it soured into ten legal proceedings and an ICSID fight. A 20%-partner government that has shown it will seize a mine is a permanent, un-hedgeable risk to a single-asset company.
2. Financing that never closes / dilutes the equity to death. After >18 months the US$295M DFC loan is still "active engagement," not signed. Every delay is funded by selling stock at C$0.60–0.88 into a ~490M share count with warrant overhang. Even if Dasa gets built, common holders can be diluted out of most of the upside — the Harte pattern in a different mineral.
3. Execution. ~US$290M and a mill that was only 30–35% complete in late-2025, in a landlocked post-coup country, run by a CEO whose last build (Harte) blew out costs and wiped out common. First pour has already slipped ~18–24 months.

Pre-mortem (18 months out, thesis broke): the DFC loan lapsed again in H2-2026 (US-Niger relations, or the DFC deprioritised it); GLO funded another year on the ATM at falling prices; the junta opened a "renegotiation" of the SOMIDA convention citing the new mining code; the stock is down another 50%+ and the FS NPV is now theoretical. Nothing exotic is required for this — it is just the base rate for single-asset African development stories that miss their financing window.

Are multiples too high? No — the opposite. At ~0.3× spot-implied P/NAV the equity is cheap on the rock and expensive on the risk. The market isn't over-paying; it's demanding a huge discount for tail risk it has watched materialise next door.

Contrarian view (what the market refuses to see): the market is anchored on the Orano precedent and may be under-weighting that Niger's incentives toward GLO are opposite to its incentives toward Orano. Orano was the French colonial-legacy operator the junta wanted to punish and replace; Global Atomic is a small Canadian company building the "largest active mining project in Niger," offering the junta royalties, jobs, a new Algeria trade corridor, and a chance to show it can partner with non-French Westerners. Niger may want Dasa built and paying far more than it wanted Orano gone. If so, the sovereign discount is overdone — but that is a bet on a junta's rational self-interest, which is exactly the bet Orano's shareholders lost.

Devil's Advocate (short-seller)

Dismantling the bull case:

• The moat you're paying for is un-monetisable without two things you don't have. Grade is real, but grade in the ground in Niger with no loan and no plant is not a business — it's a hope. The short thesis doesn't need the orebody to be bad; it needs the path to production to stay blocked, which it has for 18+ months.
• Revenue concentration is total and sovereign. 100% of the flagship's value is one mine whose landlord/partner/regulator has a live, demonstrated nationalisation habit. There is no scenario tree where GLO diversifies out of this in time.
• The most dangerous competitor bulls underestimate is not a company — it's Niger's own government as a counterparty, and the DFC's own inertia/politics as a gatekeeper. Bannerman got its US$321.5M in Namibia; GLO is still waiting on Washington. The comp isn't other orebodies; it's other financings that actually closed.
• Worst capital-allocation / governance facts: a CEO who wiped out common at Harte running the identical single-asset-development risk; a serial-dilution funding model with warrant overhang; and a securities class action alleging he over-sold the very DFC loan the thesis rests on. Any one of these is a yellow flag; together they are the short.
• Assumptions that must hold for today's price: (1) the DFC (or an alternative lender) signs; (2) Niger honours the SOMIDA convention through 2049; (3) the mill gets built near the restated schedule and budget; (4) uranium term price stays >US$60/lb (well-covered today). Break #1 or #2 and the equity is impaired, full stop.
• If growth (here: production) disappoints by 20–30%: it's worse than for an operator — a 20–30% further slip in first-pour, or a 20–30% capex overrun on US$290M (~US$60–90M more) is another dilution round of ~25–40% of the current market cap at these prices.
• The single scenario that permanently impairs, and its plausibility: Niger nationalises or forcibly renegotiates Dasa (Orano playbook). Plausibility: not low — it has literally just happened to the mine next door. That is why the stock is at C$0.63 and not C$3.

Management Questions (ordered by information value)

1. The DFC loan has been "closing" since early-2025 and is now "active engagement" with no date. What are the specific unmet pre-conditions today, and what is the drop-dead date at which you pivot to the alternative (Eastern JV / strategic) lender? (Answer most changes the view — it is the whole catalyst.)
2. Given the Orano SOMAÏR nationalisation, what contractual and practical protections does the SOMIDA convention give you that Orano lacked — and has Niger given any written reaffirmation of the fiscal/ownership terms under the new mining code?
3. What is the current all-in funding plan to close the US$290M gap — how much debt, how much equity, and what share count do you expect at first pour under your base case?
4. Walk me through what is different between the Dasa build and the Harte Gold build that ended in CCAA — specifically on feasibility discipline, contingency, and the trigger points at which you would halt rather than push through.
5. The commissioning date has moved from H2-2026 to Q4-2027 with first pour H1-2028. What is the critical path now, and what is the P50 vs. P90 on first yellowcake?
6. What contingency is in the US$501M budget, and at what uranium price and cost overrun does the after-tax IRR fall below your hurdle?
7. On the class action alleging non-disclosure of DFC pre-conditions — without litigating it here, what has changed in your disclosure practice around the loan since January 2025?
8. Of the 68 Mlb reserve, only ~8.8 Mlb is contracted. What is your term-vs-spot strategy, and how much do you contract before first pour to de-risk financing?
9. What is the realistic export logistics plan post-coup — is the Niger–Algeria/Mediterranean corridor real and costed, or aspirational, and what does it do to your AISC?
10. The Niger State's 10% paid interest obligates it to fund its share of capex/deficits. Is it actually funding, on time, in cash?
11. What is the sensitivity of the DFC's participation to US-Niger diplomatic status (troops expelled, embassy posture)? Is DFC approval formally conditioned on any diplomatic milestone?
12. How much cash runway do you have today (pro-forma the Feb-2026 raise) at current burn, and at what date do you need the next financing regardless of the DFC?
13. The zinc JV is your only earnings. Do you have any path to increase your 49% stake or influence over Befesa's dividend policy, or is it purely passive ballast?
14. What insider buying should we expect from you and the board over the next two quarters — and would you commit to a lock-up through first pour?
15. If the DFC and the Eastern lender both fall through, what is the honest sequence of events — is there a scenario where you sell down or JV the asset rather than dilute common further?

---