K92 Mining Ltd.
Combined average of Moat (AI Resilience), Growth, and Valuation scores.
Moat Score
World-class high-grade orebody; Stage 3 execution proves operational quality. Strong PNG regulatory lock-in and still-high mill grades provide durable structural advantages, partially offset by commodity pricing and single-mine concentration risk.
In mining, the "moat" is the Quality of the Orebody:
- Exceptional Grades: Q2 mill feed was 6.2 g/t gold (6.7 g/t AuEq) as Stage 3 ramped tonnes 59% quarter-on-quarter — down from Q1's 10.2 g/t, but still well above typical industry grades. High-grade Kora/Judd veins and Arakompa's AR1 zone (9.47 g/t AuEq over 4.32 m true width) remain the margin cushion that justifies the PNG risk premium.
- Proven Execution: Record Q2 physicals: 225,965 tonnes processed, 3,326 metres of development, and 426,012 tonnes total material mined. Gold recoveries of 93.8% beat the Updated DFS 92.6% parameter for a ninth consecutive quarter. July development of 1,220 metres already exceeds the Stage 4 requirement of 1,200 metres per month — ahead of remaining Q3 enabler projects.
- Exploration Engine: Arakompa maiden resource is targeted for H2 2026 after 100 holes; all holes hit mineralization, with AR1 now defined over ~400 m of strike. Record $31–35M exploration budget for 2026, with up to 16 rigs once a second additional surface rig finishes commissioning. Blue Lake Porphyry remains a longer-dated company-defining option.
- Strong Balance Sheet: Record cash of US$349.4M and net cash of US$310.0M as of June 30, 2026 (up from ~US$183M entering the year). Q2 operating cash flow of US$105.1M and EBITDA of US$140.7M fund Stage 4 without dilutive equity raises; $60M remains undrawn on credit facilities.
Moat Verdict
Six of twelve moats are scored. The durable ones are regulatory lock-in (strong — PNG Special Mining Lease through 2034), proprietary data (intact — geological ore body models), and talent scarcity (intact — underground PNG mining expertise). Business logic is not applicable: mine-planning software is industry-wide, the Cameco precedent. Transaction embedding is weakened: gold doré/concentrate offtake is switchable, so it is not a source of pricing power. Learned interfaces, public data, bundling, network effects, and system of record do not apply to a single-mine commodity producer. AI cannot disrupt the ore body or mining license. The computed moat score (67) sits near mining peer FCX (60): both are concentration-risk commodity producers selling fungible output. The score sits well below software/platform compounders because gold is a price-taking commodity.
66.9 resilient · 65.0 vulnerable · 80/20 = 66.5 · = 67
Open a moat to read its note.
Underground gold mining has no customer-facing interface or learned workflow; this moat does not apply.
K92 does not own proprietary extraction or processing logic; mine-planning and grade-prediction tools are available industry-wide and do not create differentiation versus other high-grade producers. Mining know-how is not encoded business logic, so the pillar is not applicable rather than weakened, the Cameco and GE Vernova precedent. Previously weakened.
K92 does not control access to a public dataset; this moat does not apply to a commodity gold miner.
Underground mining engineers, Papua New Guinea operational expertise, and high-grade ore processing specialists remain genuinely scarce.
gold mining is a commodity operation; the product is fungible gold, not a bundled software or service offering.
Kainantu mine geological survey data, ore body 3D models, and processing optimization data are proprietary operational assets.
PNG Special Mining Lease (ML150, due for renewal June 13, 2034), government royalty agreements, environmental permits, and community agreements are near-impossible barriers to replicate.
gold mining is a commodity business with no network effects; output is priced by global spot markets regardless of volume.
K92 sells gold doré at spot and a copper-gold-silver flotation concentrate under offtake/sales agreements with smelters and traders. As with FCX, output is fungible and counterparties are switchable, so embedding is limited — but not entirely absent.
K92 is not a system of record for any business function; value is entirely in the ore body and mining license, not information systems.
High-grade Kainantu ore keeps all-in sustaining cost low, but a single-mine producer has no portfolio scale.
Sells a fungible product or capacity: buyers pay the market price, not a brand premium.
Combined average of Moat (AI Resilience), Growth, and Valuation scores.
Moat Score
World-class high-grade orebody; Stage 3 execution proves operational quality. Strong PNG regulatory lock-in and still-high mill grades provide durable structural advantages, partially offset by commodity pricing and single-mine concentration risk.
Growth Score
Q2 2026 produced 46,093 oz AuEq (H1 92,836 oz) on record mill throughput of 225,965 tonnes (+73% YoY, +59% QoQ), with gold recoveries of 93.8%. Head grade fell to 6.2 g/t gold as Stage 3 tonnes ramped, so ounces were roughly flat versus Q1's 46,743 oz. 2026 guidance of 190–225k oz was reiterated, with Q4 expected to be the strongest quarter as the third mining front (Judd) comes online in Q3 and remaining enablers (pastefill, ventilation, 60-tonne river crossings) complete. Stage 3 growth capital is 98% spent or committed; July development already exceeds the Stage 4 rate.
Valuation Score
At ~$29.80 CAD, the stock has rallied ~23% since the June review and now sits 22% below the $38 CAD base case. Q2 realized gold of $4,493/oz against $1,376/oz AISC still implies ~$3,100/oz margins, but AISC is running above the $1,250–$1,350 2026 guide as Stage 3 tonnes ramp. Analyst consensus is ~$34.50 CAD. Stage 4 execution, the H2-weighted production step-up, and the H2 Arakompa maiden resource are the primary re-rating catalysts.
The Mining Moat (Asset Quality)
In mining, the "moat" is the Quality of the Orebody:
- Exceptional Grades: Q2 mill feed was 6.2 g/t gold (6.7 g/t AuEq) as Stage 3 ramped tonnes 59% quarter-on-quarter — down from Q1's 10.2 g/t, but still well above typical industry grades. High-grade Kora/Judd veins and Arakompa's AR1 zone (9.47 g/t AuEq over 4.32 m true width) remain the margin cushion that justifies the PNG risk premium.
- Proven Execution: Record Q2 physicals: 225,965 tonnes processed, 3,326 metres of development, and 426,012 tonnes total material mined. Gold recoveries of 93.8% beat the Updated DFS 92.6% parameter for a ninth consecutive quarter. July development of 1,220 metres already exceeds the Stage 4 requirement of 1,200 metres per month — ahead of remaining Q3 enabler projects.
- Exploration Engine: Arakompa maiden resource is targeted for H2 2026 after 100 holes; all holes hit mineralization, with AR1 now defined over ~400 m of strike. Record $31–35M exploration budget for 2026, with up to 16 rigs once a second additional surface rig finishes commissioning. Blue Lake Porphyry remains a longer-dated company-defining option.
- Strong Balance Sheet: Record cash of US$349.4M and net cash of US$310.0M as of June 30, 2026 (up from ~US$183M entering the year). Q2 operating cash flow of US$105.1M and EBITDA of US$140.7M fund Stage 4 without dilutive equity raises; $60M remains undrawn on credit facilities.
Moat Verdict
Six of twelve moats are scored. The durable ones are regulatory lock-in (strong — PNG Special Mining Lease through 2034), proprietary data (intact — geological ore body models), and talent scarcity (intact — underground PNG mining expertise). Business logic is not applicable: mine-planning software is industry-wide, the Cameco precedent. Transaction embedding is weakened: gold doré/concentrate offtake is switchable, so it is not a source of pricing power. Learned interfaces, public data, bundling, network effects, and system of record do not apply to a single-mine commodity producer. AI cannot disrupt the ore body or mining license. The computed moat score (67) sits near mining peer FCX (60): both are concentration-risk commodity producers selling fungible output. The score sits well below software/platform compounders because gold is a price-taking commodity.
66.9 resilient · 65.0 vulnerable · 80/20 = 66.5 · = 67
Open a moat to read its note.
Underground gold mining has no customer-facing interface or learned workflow; this moat does not apply.
K92 does not own proprietary extraction or processing logic; mine-planning and grade-prediction tools are available industry-wide and do not create differentiation versus other high-grade producers. Mining know-how is not encoded business logic, so the pillar is not applicable rather than weakened, the Cameco and GE Vernova precedent. Previously weakened.
K92 does not control access to a public dataset; this moat does not apply to a commodity gold miner.
Underground mining engineers, Papua New Guinea operational expertise, and high-grade ore processing specialists remain genuinely scarce.
gold mining is a commodity operation; the product is fungible gold, not a bundled software or service offering.
Kainantu mine geological survey data, ore body 3D models, and processing optimization data are proprietary operational assets.
PNG Special Mining Lease (ML150, due for renewal June 13, 2034), government royalty agreements, environmental permits, and community agreements are near-impossible barriers to replicate.
gold mining is a commodity business with no network effects; output is priced by global spot markets regardless of volume.
K92 sells gold doré at spot and a copper-gold-silver flotation concentrate under offtake/sales agreements with smelters and traders. As with FCX, output is fungible and counterparties are switchable, so embedding is limited — but not entirely absent.
K92 is not a system of record for any business function; value is entirely in the ore body and mining license, not information systems.
High-grade Kainantu ore keeps all-in sustaining cost low, but a single-mine producer has no portfolio scale.
Sells a fungible product or capacity: buyers pay the market price, not a brand premium.
Growth Analysis
Growth Drivers
Key Risk
Single-asset PNG concentration: a Stage 4 ramp delay (geotechnical, permitting, or community/sovereign disruption) combined with a gold correction to $3,000/oz would cut operating cash flow ~50% over 12-18 months. H1 AISC of ~$1,400/oz is already above the $1,250–$1,350 2026 guide; if mill grade stays near Q2's 6.2 g/t into H2, cost guidance misses and the H2 step-up needed to hit 190–225k oz slips.
Score Derivation
86.0 base + 2.7 trajectory − 10 risk = 79
Base 86 (24% midpoint of 20-28% production CAGR) + 2.7 trajectory (throughput and exploration accelerating; gold/AISC margins stable) + 0 stable margins − 10 high single-asset PNG/Stage 4 risk = 79
Stage 3 Ramp — Stage 4 Development Already at Rate
The Transformation Journey
The 1.2 Mtpa Stage 3 process plant has been fully operational since December 2025; 98% of Stage 3 growth capital is spent or committed. Q2 printed record mill throughput (225,965 t) as the second mining front ramped. Attention is on Stage 4 (400,000+ oz AuEq run-rate): July development of 1,220 metres already exceeds the 1,200 m/month Stage 4 requirement, and 2026 growth capital is guided at $100–108M ($25–28M remaining Stage 3, $75–80M Stage 4).
Production is expected to be strongest in H2 2026, with Q4 the peak quarter. Remaining enablers: surface pastefill commissioning in Q3, underground paste plant in Q4, Phase 4 ventilation electrification in Q3 (airflow ~350 → 600 m³/s), 60-tonne river crossings in Q3, and a third mining front from Judd in Q3. The 15.3 MW power station (Stage 4 power requirement) was completed in May.
Price Scenarios (12–24 Months)
Where We Are vs Targets
Loading live price…
Gold corrects sharply to $3,000/oz and Stage 4 faces significant delays, compressing margins and undermining the production re-rating thesis.
- Gold price correction to $3,000/oz reduces operating cash flow by ~50% from peak levels
- PNG jurisdictional risk leads to tax/permit disruptions, delaying Stage 4 by 18+ months
- Grade stays near Q2's 6.2 g/t and AISC remains above $1,400/oz, missing the $1,250–$1,350 2026 guide
- Multiple compression to 8x cash flow on reduced earnings power brings stock to $17 CAD
Gold holds near $4,500/oz and Stage 4 ramp-up proceeds on schedule, driving a production and cash flow re-rating.
- 2026 production reaches the top half of 190–225k oz guidance as Q4 prints the strongest quarter, with AISC moving toward the $1,250–$1,350 guide
- Stage 4 development stays above 1,200 m/month and remaining Q3/Q4 enablers (pastefill, ventilation, 60-tonne haulage) complete on schedule
- Arakompa maiden resource in H2 2026 supports the high-grade thesis and attracts further analyst target upgrades
Gold surges above $6,000/oz and Stage 4 delivers ahead of schedule, with Arakompa or Blue Lake emerging as a company-defining discovery.
- Gold at $6,000+ with AISC near $1,300/oz generates ~$4,700/oz margins on 225k+ oz production
- Stage 4 commissioning accelerates to late 2027, pushing run-rate above 350k oz
- Arakompa resource or Blue Lake drilling prompts a major gold producer to table a takeover bid at a significant premium