ATCX — H1 2026 Financial Report Analysis

Overview

Atlas Critical Minerals is an exploration company with mineral rights across Brazil for rare earths, graphite, titanium, uranium and other minerals. It also owns two small side businesses: an iron ore property that a third party mines in exchange for a per-tonne payment, and a quartzite quarry that is currently shut. It listed on Nasdaq on January 9, 2026, at the same time as an offering of 1,380,000 shares at $8.00 each, about $11.0 million before fees. As a foreign private issuer it reports every six months, not every quarter, so this first-half filing (six months to June 30, 2026) is its only 2026 financial report so far.

The offering is what this report is about. Its net loss nearly doubled to $5.37 million. Most of the increase came from head-office costs and stock-based pay. Only $0.24 million was spent on exploration, the activity the offering was raised to fund. Cash went from $30,220 at the end of 2025 to $5.54 million at June 30, so the company is funded for now. At the first half's rate of spending, though, that cash lasts into roughly the first half of 2027.

At a glance

• Net loss of $5.37 million, up from $2.80 million. General and administrative costs rose from $1.32M to $2.93M, and stock-based compensation rose from $1.36M to $2.37M. Together those two lines are 96% of operating expenses.
• $3.22 million of cash used by operations in six months, up from $1.21M. With $5.54M in the bank at June 30 and no debt, that is about 10 months of operating and capital spending at the first-half pace.
• Revenue of $74,386, all of it from one customer: the lessee mining the Rio Piracicaba iron ore property. The quartzite quarry earned nothing, and its restart has moved back to the first quarter of 2027.

Results

n/m = not meaningful. The YoY percentage is left out where both years are losses or the base is tiny. Revenue rose 31% on paper, but the two figures come from different businesses: quartzite sales in 2025 and an iron ore lease payment in 2026. Revenue is about 1% of operating costs, so an operating margin percentage (operating profit as a share of revenue) would tell the reader nothing and isn't shown.

Where the money went

The company spent $5.55 million on operating expenses in the half, against $2.70 million a year earlier:

• General and administrative: $2.93M (was $1.32M). These are the costs of running the company rather than exploring: salaries, offices, professional fees, and the extra costs of being a Nasdaq-listed company. The filing gives no breakdown of the increase.
• Stock-based compensation: $2.37M (was $1.36M). This is pay in shares rather than cash. It isn't cash out the door, but it dilutes existing shareholders. In the half, 228,364 shares were issued to executives and consultants, and 90,207 new restricted stock units (RSUs, promises of shares that vest over time) were granted with a grant-date value of $935,033.
• Exploration: $0.245M (was zero). This is 4.4% of operating expenses. The company also capitalised $185,884 to buy 1,563 hectares of graphite mineral rights at Malacacheta in Minas Gerais. Capitalised means it was recorded as an asset on the balance sheet rather than as an expense.

The 20-F annual report said the offering money would go to "exploration, including drilling and assessment of deposits and reserves," plus capital spending and working capital. Six months in, the split leans heavily toward overhead.

What the headline numbers hide

• Loss per share understates how fast the loss grew. The net loss rose 92%, but loss per share rose only 15%, from $0.94 to $1.08. The difference is that the average share count went up 66% after the offering and the share awards. Each shareholder owns a smaller slice of a larger loss.
• Cash burn is lower than the loss, and the gap is almost all share-based pay. Operating cash outflow was $3.22M against the $5.37M net loss. Of the $2.14M difference, $2.37M is non-cash stock compensation, partly offset by paying down $0.23M of payables. Excluding stock pay, the cash cost of running the company was roughly $3M for the half.
• About $1 million of the offering repaid the former parent. Financing outflows included $1.07M to related parties. That was mainly the $1,005,049 of intercompany loans owed to Atlas Lithium (which owns 20.16% of the common stock), settled in Q1 2026. It was a one-off use of the new cash. It isn't in the operating burn, but the money is gone.
• Runway estimate. Operating cash outflow ($3.22M) plus capital spending ($0.19M) came to $3.42M in six months, or about $0.57M a month. On $5.54M of cash, that pace lasts roughly ten months from June 30, into about the second quarter of 2027. Any larger drilling programme would shorten that.
• Going concern. The audit opinion on the 2025 financial statements (in the 20-F) flagged "substantial doubt" about the company's ability to keep operating. That is a formal warning that it might not have enough money to continue. Management said the January offering removed that doubt and that it had enough cash for the next 12 months. The interim statements don't repeat any going-concern language. The 12-month horizon set in February 2026 runs out in early 2027, which matches the runway estimate above.
• Small one-offs in other income. Net finance income was $0.10M, compared with a $0.02M cost a year ago. It includes a $42,158 gain from the company's currency hedges (contracts that lock in a US dollar–Brazilian real exchange rate, $1.5M notional, settling July–December 2026). There is also an accounting change: from January 1, 2026, the Brazilian subsidiaries report in US dollars instead of reais, so the currency-translation line in comprehensive loss disappeared (it was +$0.18M in H1 2025). None of these items changes the picture.
• Revenue depends on one customer. All revenue came from the single lessee at Rio Piracicaba, where iron ore mining began on November 28, 2025. The quartzite quarry was stopped "to address certain operational issues." The 20-F (filed February 2026) said quartzite would resume in the second half of 2026. The interim filing now says the first quarter of 2027, a slip of at least one quarter.
• Governance. CEO Marc Fogassa holds the single Series A preferred share, which carries 51% of all shareholder votes regardless of how much common stock is outstanding. Common shareholders cannot outvote management on any matter. The CFO resigned for personal reasons on April 1, 2026, and Virgilio Santiago, formerly CFO of a Brazilian nickel project, replaced him on April 7.

An $8 million option that cash can't cover

In December 2024 the company paid $500,000 in stock to Atlas Lithium for an option to buy Brazil Mineral Resources Corporation, which holds 60 more Brazilian mineral rights (copper, gold, graphite, nickel, rare earths, titanium). Exercising the option costs $8.0 million, payable in cash, in shares valued at $7.519 each, or a mix. Atlas Lithium chooses the form of payment, not the buyer. All-stock would mean about 1.06 million new shares, roughly 21% of the 5.08 million outstanding at June 30. All-cash would be more than the company holds. Under the 20-F terms, the option lapses 12 months after the F-1 registration filing of September 15, 2025, so around mid-September 2026, unless both sides extend it. The June 30 balance sheet still carries the $500,000 as a current asset. The filings we reviewed through September 29, 2026 do not say whether the option was exercised, extended or allowed to lapse. Readers should watch for that disclosure.

Takeaway: The January IPO turned a company with $30,000 in the bank into one with $5.5 million. In the first six months, though, overhead and stock pay ($5.3M) outran exploration spending ($0.24M) by more than 20 to 1. At the current pace the money lasts into roughly mid-2027. Before then the company needs either drilling results that justify raising more money on better terms, or another dilutive share sale.

What's next

The first real exploration result arrived after the period. On September 29, 2026 the company filed a technical report summary, prepared by SGS Canada, with a maiden resource estimate for the Malacacheta graphite project. A resource estimate measures how much mineralised rock is likely there. It is not a reserve, which would mean extraction has been shown to be economic. The estimate, at a 2.0% graphitic-carbon cut-off grade (the minimum concentration counted):

"Indicated" means enough drilling to estimate tonnage and grade with reasonable confidence. "Inferred" is lower-confidence and cannot be used in an economic study. The report states there are no mineral reserves, so economic viability has not been demonstrated. Test work produced a concentrate at 93.95% total graphitic carbon, "a little bit shy" of the 94% industry specification. The independent geologist recommends more drilling, since the deposit is open along strike and at depth, plus a metallurgical study. Both cost money the current runway has to cover.

Management's guidance in the filing is limited to project timing: the quartzite quarry restarting in Q1 2027. The company gives no spending or revenue guidance.

Our view: Going from a graphite land purchase in H1 to an SEC-compliant resource estimate by September is real progress, and it is the company's first quantified asset. The financial trajectory points the other way. A cost base of about $0.5–0.6M a month, mostly corporate rather than field spending, against $5.5M of cash means 2027 almost certainly brings another capital raise. Shareholders will want to know the outcome of the BMR option, whether H2 spending shifts toward drilling and metallurgy rather than G&A, and whether the quartzite restart actually happens in Q1 2027 after the second delay. The full-year 2026 results (Form 20-F, due by April 2027) will show the year-end cash balance, which is the number that sets how much more stock has to be sold.

This is the first Atlas Critical Minerals report published on this site, so there is no earlier outlook to check against.