Contrarian Codex — Newsletter #120: 1H Done, Onto 2H

Cosa is a tiny exploration company drilling for uranium in Canada's Athabasca Basin — the highest-grade uranium district on earth. Their winter drill results came back showing a 5-metre interval with respectable uranium grade (0.55% U3O8), but the more geologically telling detail is that the rock also contains nickel and cobalt alongside the uranium. That polymetallic fingerprint is the same geochemical signature you find in the best Athabasca deposits ever discovered — Cigar Lake, Key Lake, and the nearby Hurricane deposit — because it marks the specific kind of fluid interaction that creates these high-grade ore bodies.

Why is the stock flat? The Athabasca Basin sets an extremely high bar: nearby deposits run at 30–50% grade, so 0.55% is solid early-stage work but not the kind of bonanza number that makes speculators pile in. The next test is the largest drill program Cosa has ever run, launching mid-June, targeting 1.2km of open strike on the Cyclone trend. If they can show comparable grades on multiple sections the equity re-rates meaningfully. If the trend turns out to be patchy, the story stalls for now — but Mart is backing the team that found Hurricane to take another run at finding a second one.

Minera Alamos runs the Pan gold mine in Nevada and just put up its best quarter ever — record revenue (~$39m), record mine earnings (~$19.5m), costs below the low end of guidance. But the company was never mainly about Pan; the real story is the Copperstone mine they are about to build in Arizona. The pre-feasibility study dropped this issue and the numbers are striking: at today's gold price (~$4,500/oz), Copperstone has an after-tax NPV of ~$537 million, an IRR above 150%, and pays back in under 10 months. The mine is fully permitted, the board has formally voted to build it, and first gold is targeted for mid-2027.

Here is the striking math: that $537m NPV is essentially equal to Minera's entire market cap. So the market is valuing the Pan mine (which is producing right now), the Copperstone construction project, and two other growth-stage assets at close to zero on top of it. Mart is a believer — he calls Minera one of his main gold-producer winners for the year. The caveat he acknowledges: funding a new mine, sustaining the old one, and running two other projects simultaneously is tight math if gold wobbles or Pan stumbles. But given his conviction that gold heads higher, he is inclined to let it ride.

Global Atomic is building the Dasa uranium mine in Niger — one of the higher-grade undeveloped uranium projects in Africa, in a country that has historically been a significant producer. The investment case is simple: a quality uranium project in an environment where uranium prices are rising and new mine supply is badly needed. This issue they scored a genuinely positive development: a high-profile visit to Niamey produced a formal written endorsement of Dasa from Niger's president, and a new cooperation agreement with Algeria opens a potential export corridor for the finished uranium all the way to the Mediterranean — a useful alternative to the southern route through Benin that has been on and off since the 2023 coup.

The honest catch is that political support does not pay for the mine. The company ended Q1 with roughly C$59m in cash and is spending C$25–28m per quarter, which puts cash exhaustion roughly 2–2.5 quarters away if nothing changes. They still need about US$265m to actually finish the plant, and there is still no financing announcement — no debt facility, no JV partner, no concrete deal. Mart holds the position and sees the political de-risking as a genuine positive, but the clock is ticking loudly and a financing announcement before the summer ends is now the single most important event for this stock.

Westwater is building the Kellyton plant in Alabama, which will process graphite into battery-grade anode material — the stuff that goes into EV, industrial, and defense batteries. They also own the Coosa graphite deposit in Alabama that will feed the plant. The numbers are remarkable on paper: add up the pre-tax NPV of both phases of Kellyton plus the Coosa mine and you get over $1.6 billion, set against a market cap of around $80 million. That is roughly a 20x gap between what the company says its assets are worth and what the market is paying for them.

The reason for the gap is a simple but stubborn one: the financing to actually build Phase 1 of Kellyton has been "almost closed" for 18 months and is still not closed. An offtake agreement with a major South Korean battery maker was terminated this quarter. The company is now pivoting to government funding programs (DC is emphasizing domestic critical minerals), which is the right move in this policy environment, but government timelines are slow. The qualification line is actually running and shipping samples to customers, so the technology works — the obstacle is pure financing. Mart holds it for the asymmetric upside, acknowledges the frustration openly, and says the back half of 2026 needs to deliver something concrete.

Aldebaran owns the Altar copper-gold porphyry in Argentina's central Andean belt — a massive bulk-tonnage deposit already measured at 2.4 billion tonnes in the measured and indicated resource categories, containing roughly 22 billion pounds of copper plus 5 million ounces of gold. Porphyry copper deposits are graded differently from other mining projects: world-class examples like the ones right next door to Altar (Los Pelambres, El Pachón) all run at 0.4–0.6% copper over enormous tonnage, so that is the yardstick, not the high-grade vein thinking you might apply elsewhere.

This issue they released infill drill results showing holes up to 1.3 kilometres long averaging 0.45% copper equivalent, with a 500-metre higher-grade core at 0.71%. These are excellent numbers for a porphyry, and more importantly they show the system is continuous and predictable at depth — which is what you need to upgrade the resource from inferred to indicated and eventually to reserves. The share price has not ripped because the market already knows Altar is a huge deposit; the question is who will write a big cheque to put it into production. That question gets answered at the Q3 resource update and the 2027 pre-feasibility study — the real value-unlocking events. Today's holes are an incremental step toward both.

Comstock held an extensive live investor tour of their facilities and land assets this issue. Mart followed the entire event closely but describes it as a lot to go through, and he is spending time this weekend carefully analyzing all the new information and data along with commentary and photos from attendees. A comprehensive written report is coming shortly after the weekend — this newsletter entry is effectively a placeholder with the substantive analysis to follow.

LibertyStream is a company working on extracting lithium directly from oil-field brine — essentially recovering lithium as a byproduct of water that comes up alongside oil during production. This is a potentially attractive approach because the infrastructure (wells, pumping equipment) already exists, and the lithium-bearing brine that oil companies currently treat as a waste product could instead be a revenue stream. Direct lithium extraction (DLE) technology has been getting growing attention in the energy sector.

The stock is up over 50% from recent lows. The drivers, per Mart: speculation that an offtake agreement (a commitment from a buyer to purchase the lithium) is close to being announced — he notes this kind of speculation has surfaced before, so he is not counting on it — and a popular article about DLE circulating on a major oil-industry platform that brought new eyes to the space. Mart added more stock in the mid-C$0.80s to what was already an overweight position in his portfolio and says he will hold through whatever near-term volatility comes, up or down. The company is delivering on its operational milestones and he is happy to own it.

Merlin Labs makes the AI "brain" that flies a plane. It is a software platform you install into existing aircraft so it can take off, cruise, and land with a reduced or eventually zero human crew. The genuine differentiator: it can communicate with air traffic control over the radio the way a human pilot does, understanding accents and complex instructions — something almost no competitor has achieved — and every flight it completes adds to a proprietary dataset that makes that skill progressively harder to replicate. It has been flown on 7 different airframes from a small Long-EZ up to the C-130J military transport, and it holds the most stringent aviation-software certification class (DO-178C Level A).

Mart added it at $6 as the first non-commodity name in the Codex portfolio. His valuation work (covered in detail in the separate deep-dive newsletter): bear ~$16 if execution is minimal (and $0 if it fails entirely — so the position is sized for that possibility), base ~$77, bull ~$216 (a 35x from entry). The catalyst calendar: a C-130J Critical Design Review expected in 2H 2026, New Zealand civil aviation certification progress, and likely additional sell-side analyst coverage coming. The bull case is enormous; the honest caveat is that the company is burning through cash, the field has credible competition, and a de-SPAC stigma weighs on sentiment.

ASPI enriches Silicon-28 — a highly purified form of silicon where one of the naturally-occurring isotopes (Si-29) has been removed. Why does that matter? When you build quantum computer chips on natural silicon, the Si-29 atoms act like constant radio static, knocking quantum bits out of their computing state before they can do any useful work. Strip Si-29 out to extremely high purity (above 99.99%) and those qubits can hold their quantum state roughly a thousand times longer — the difference between a laboratory experiment and a processor that can actually run algorithms. There is also a commercial benefit for regular chips: pure Si-28 conducts heat about 50% better than natural silicon, which matters a great deal for next-generation AI accelerator design where heat is the binding constraint.

This issue ASPI announced the first 18 stages of their Silicon-28 enrichment plant in Pretoria have been successfully restarted and have been running at target enrichment levels for over three weeks, with Q3 2026 commercial shipments now guided. The stock moved +25% on the news. Mart's read: positive on the margin — this is real operational progress after roughly a year of setbacks — but he is explicitly not popping the champagne. The original commercial shipment target was Q2 2025; it has now been revised at least six times in 18 months. The Q1 earnings call held just six days before this release was still guiding mid-year, and the new release quietly moved it to Q3 without flagging the change. The inflection that actually matters is when Silicon-28 revenue hits the income statement for the first time. Until then, the company has been promising the same milestone on a rolling basis.