CNL — Q2 2026 Financial Report Analysis

Drilling doubled the quarterly loss, and land deals added $38 million of instalment debt

Collective Mining is an exploration company: it has no mine and no revenue. It is drilling the Apollo gold-silver-copper-tungsten deposit on its Guayabales Project in Caldas, Colombia, and pays for that from cash raised by selling shares. So the useful questions for any quarter are how fast it spent money, what it got for the spending, and how long the cash lasts. In the second quarter of 2026 (April–June), the net loss rose to $20.1 million from $8.5 million a year earlier. Exploration spending doubled to $14.8 million: the company drilled 23,649 metres, 31% more than a year earlier, and paid higher rates per metre under new drilling contracts. Cash fell to $93.7 million from $129.6 million at the end of 2025. The part of this filing that is easy to miss is on the balance sheet. During the first half the company bought about $40 million of land around Apollo and San Antonio, mostly paying in multi-year instalments. Those unpaid instalments sit on the balance sheet as $37.8 million of obligations, up from $4.7 million in December.

All figures are in US dollars, which is the company's reporting currency. The company files with the SEC as a Canadian foreign private issuer, so this quarter's results came as a Form 6-K with interim financial statements and MD&A attached rather than a 10-Q. Its shares moved from NYSE American to Nasdaq on August 11, 2026, still under the ticker CNL. They also trade on the TSX.

At a glance

• $13.9 million of operating cash out in Q2, against $7.6 million a year earlier. This is the company's real spending rate before land purchases. Add $5.5 million of investing cash (land and mineral rights) and Q2 cash burn comes to about $19.5 million.
• $37.8 million of land and mineral-rights instalments owed, of which $11.6 million is due within a year. Subtract all of it from the $93.7 million cash pile and roughly $56 million is left that isn't already committed to these payments.
• No new shares sold in 2026 so far. 92.7 million shares were outstanding at June 30, almost the same as the 92.5 million in December, because the company is still spending money raised in 2025 (C$125.4 million in an October bought deal plus two strategic placements). Weighted-average shares were still 17.8% higher than in Q2 2025, so loss per share understates how much the total loss grew.

The numbers

An exploration company's "income statement" is really a list of costs. Under the company's accounting, exploration spending is expensed as it happens rather than added to the balance sheet as an asset. A bigger drilling program therefore shows up directly as a bigger loss.

Balance sheet at June 30, 2026 vs December 31, 2025: cash $93.7M vs $129.6M; working capital (current assets minus current liabilities) $71.7M vs $122.6M; property, plant and equipment $53.0M vs $11.2M; total long-term obligations including the current portion $37.8M vs $4.7M; equity $116.1M vs $145.1M.

For the first half (January–June), the net loss was $32.5 million versus $25.4 million, operating cash outflow was $24.4 million versus $13.0 million, and the company drilled 35,874 metres versus 30,251.

Where the money went

Almost all of the exploration spending went into one project. Guayabales took $14.7 million of the $14.8 million spent in Q2, and the San Antonio Project took about $0.1 million. The line items show the program moving from finding mineralization toward preparing for a mine permit:

• Drilling services: $6.2 million, up from $3.0 million. Dividing drilling cost by metres drilled gives roughly $260 per metre this quarter against about $166 a year ago (our calculation; the cost line also includes some non-metre charges). Most of the cost increase came from higher rates per metre, not from drilling more metres. The MD&A attributes it to "increased drilling costs per metre following updated drilling contract rates" as well as the higher metreage.
• Studies and technical evaluation: $1.2 million, up from nothing. These are engineering, environmental and geotechnical studies "to support future economic studies and environmental permitting activities." This spending did not exist a year ago. It is a sign that Apollo is being prepared for a permit application.
• Security costs rose from $0.09 million to $0.51 million, and community costs from $0.23 million to $0.50 million. These are small amounts, but in Colombia they are recurring costs that grow as the site gets busier.

General and administrative costs rose $1.7 million. About $0.7 million of that was share-based compensation (stock options granted to staff, a non-cash cost) and about $0.5 million was salaries for a larger corporate team.

Compared with the previous quarter, the loss rose from $12.4 million in Q1 to $20.1 million in Q2. Metres drilled almost doubled over the same period, from 12,225 to 23,649.

What the headline numbers hide

• The land purchases are debt in all but name. In the first half the company signed instalment deals worth $49.7 million at face value. The largest was a four-year Guayabales land purchase for $33.6 million, and another was a four-year San Antonio land purchase for $10.6 million. It also took on the Trap mining concession application. Each deal is recorded at present value using a 9.5% discount rate, and the payments run to 2030. Most of the property growth from $11.2 million to $53.0 million came from these deals, not from cash spent. The cash flow statement lists $34.9 million of PP&E additions as non-cash. These payments are fixed. The MD&A notes that the option agreements can be cancelled at any time, but these purchases cannot.
• Finance costs jumped because of those instalments. Finance costs were $1.41 million in Q2 against $0.08 million a year earlier, and $1.32 million of that was "interest accretion" on the instalment obligations and leases. That is the 9.5% discount being added back over time, so it costs no cash today, but it is the price of paying for land later. It now cancels out the $0.85 million of interest the company earns on its cash.
• Operating cash burn looks lower than the loss partly because bills went unpaid longer. First-half operating outflow was $24.4 million against a $32.5 million loss. Part of the gap is genuinely non-cash: $2.9 million of share-based pay and $0.9 million of depreciation. But $3.2 million came from working capital, mainly accounts payable rising from $4.8 million to $10.1 million. Those bills will be paid in a later quarter.
• The first-half comparison is distorted by a one-off item in 2025. The H1 2025 loss included a $10.6 million non-cash loss on revaluing warrants. These were share-purchase rights held by Agnico Eagle, which rose in value as the stock price rose. Without it, the H1 2025 loss would have been about $14.9 million. So the underlying first-half loss more than doubled, not the 28% the headline comparison suggests. The Q2-only comparison has no such distortion.
• Currency moved the other way this year. Q2 had a $0.74 million foreign-exchange loss against a $0.53 million gain in Q2 2025. The company holds Canadian-dollar cash and owes Colombian-peso liabilities, so exchange rates move the bottom line by amounts of this size from quarter to quarter.
• Dilution has paused, not stopped. No shares were sold in 2026 and only 204,709 were issued, all from option exercises. Fully diluted shares (including 6.6 million options) are 99.4 million. The next equity raise is the dilution to watch. The company filed a C$500 million shelf prospectus on May 13, 2026, a pre-approval that lets it sell securities quickly when it chooses.

Takeaway: Collective has moved from a drilling budget to a development-scale budget. Q2 burned about $19.5 million in cash, and the company now has $37.8 million of fixed land and mineral-rights payments running to 2030. $93.7 million of cash looks large, but at Q2's pace and after this year's $11.6 million of instalments it covers roughly four quarters. A share sale in 2027, or sooner, is the base case. The company says as much: its going-concern assumption rests on "the anticipation of obtaining additional sources of financing."

What comes next

The company has made several announcements since the quarter ended. These are press releases, not part of the Q2 financial statements.

• Maiden resource (September 8, 2026). Apollo's first formal mineral resource estimate reported 2.58 million ounces of gold-equivalent in the Indicated category (37.0 million tonnes at 2.17 g/t) and 2.83 million ounces Inferred (48.2 million tonnes at 1.83 g/t). Gold-equivalent converts the deposit's silver, copper and tungsten into gold-value terms, and gold makes up about half of the open-pit Indicated grade. "Indicated" means drilling is dense enough to estimate tonnes and grade with reasonable confidence. "Inferred" is less certain. The company states that a resource is not an economic study and does not show that a mine would make money.
• Exploration adit. An adit is a horizontal tunnel used for underground drilling. The 1.8 km access road is finished, construction is scheduled for Q4 2026, and underground drilling at the Trap target is planned for Q1 2027. Building it will add capital spending on top of the drilling budget.
• Permitting. Metallurgical work (tests of how much metal can be recovered from the rock) is feeding a processing flowsheet for a 2027 licensing submission in Colombia.

Our view: the Q2 numbers fit a company that has decided Apollo is worth building toward. That shows in the land consolidation, the permitting studies and the adit. It also means costs are unlikely to fall back to 2025 levels. Three things to watch in the Q3 report, due around mid-November: whether quarterly operating burn holds near $14 million or climbs as the adit starts; how much of the $11.6 million in near-term instalments is paid; and whether the company uses the C$500 million shelf. Beyond Q3, the key event is a first economic study (a PEA or similar). The resource estimate shows the deposit is large. Only an economic study can show whether mining it would be profitable.