Will HVC MLE capex estimate exceed CAD 2.4B by end of 2026?

The CAD 2.1-2.4B estimate at sanction already incorporated 15-20% inflation from prior estimates. Management explicitly framed this as incorporating QB2 learnings — adding contingencies, inflation provisions, and tariff risk. If genuine, the top of the range (CAD 2.4B) already includes significant buffer. Engineering was ~70% complete at sanction, which is higher than typical and reduces scope change risk. The project is brownfield (lower complexity than QB) and 100% owned. However, the pattern of capex inflation (QB2, then HVC initial estimate) suggests a structural bias toward underestimation. Canadian construction cost inflation and tariff uncertainty add risk. Below 50% because the wide contingency band may hold, but not much below.

The distinction between HVC MLE and QB2 is important. HVC is brownfield (existing mine infrastructure), 100% owned (no JV complexity), and in a well-understood jurisdiction (BC). QB2 was greenfield in Chile with JV partners and extreme geographic/logistical challenges. These are fundamentally different risk profiles. Management's explicit incorporation of QB2 learnings through larger contingencies should be taken seriously given the institutional trauma of QB2 overruns. The independent construction readiness assessment was positive. The remaining 30% engineering risk is real but manageable for a brownfield expansion. Leaning toward NO on exceeding CAD 2.4B.

US-Canada tariff uncertainty is a material wildcard. The capex estimate already includes 'potential tariffs' but the tariff environment could worsen beyond what was provisioned. Construction material inflation in BC could exceed forecasts if housing and infrastructure spending increase. Indigenous rights dispute (SSN) could introduce delay-related costs. While the project is simpler than QB2, mining capex has a well-documented tendency to exceed estimates across the industry, not just at Teck. The wide range (CAD 2.1-2.4B) suggests management itself sees meaningful uncertainty. If costs land at CAD 2.3B, they'd call it 'within range' — the question is whether they breach CAD 2.4B specifically.

The estimate at sanction was already inflated 15-20% from prior guidance. This is either management genuinely padding estimates after QB2, or the first in another round of increases. Given that HVC is brownfield, 100% owned, and 70% engineered, I lean toward the padding interpretation. The $300M range (CAD 2.1-2.4B) provides flexibility. The project is less complex than QB by any measure. If they exceed CAD 2.4B, it would require a significant cost surprise — not just normal inflation drift. Below 50% but above 30% given the industry pattern.

The consolidation-calibrator classified CAPITAL_DEPLOYMENT as MIXED, partly due to HVC capex inflation. The committee saw this as a pattern with QB2. But patterns can be partially corrected — the question is whether the correction is sufficient. Mining capex overruns are common industry-wide (McKinsey data shows 60-80% of mining projects exceed initial budgets). The sanctioned estimate is already revised up, which should reduce the probability vs. original estimate, but the question is about the revised top end. I estimate ~43% chance of exceeding even the padded CAD 2.4B.

The strongest factor is that management specifically designed the estimate to incorporate QB2 learnings. They added contingencies, inflation provisions, and tariff buffers — the exact factors that drove QB2 overruns. If you've already added provisions for the things that went wrong last time, the probability of the same factors causing overruns is lower. New risks (indigenous disputes, supply chain disruption, unusual weather) could emerge but would need to be significant to breach the padded top end. The BC government is fast-tracking permitting, reducing regulatory delay risk. Brownfield advantage is genuine.

Brownfield, 70% engineered, padded estimate with QB2 learnings. Simpler project than QB by every measure. Below 50% for exceeding CAD 2.4B. But mining capex overruns are common, so not far below 50%.

The estimate is already the REVISED estimate. CAD 2.4B is the top of a range that was itself inflated 15-20%. To exceed it requires costs worse than the already-pessimistic scenario. While possible (tariffs, supply chain, indigenous dispute), the probability is below 40% given the built-in buffers.

Pattern of Teck capex overruns (QB2, then initial HVC estimate increase) creates a prior for further increases. But the padded estimate should account for known risks. Tariff uncertainty is the biggest wildcard. Near 40% reflecting the tension between pattern-based prior and estimate-padding counter-signal.