Three Mining Capex Cycles Point to the Same Conclusion at Rio Tinto

The 2007-2012 cycle saw capex grow from $5.0 billion in 2006 to $17.6 billion in 2012, a 250% expansion driven by the Pilbara expansion, the Oyu Tolgoi initial build, and the Alcan integration. Free cash flow troughed in 2012 at approximately $4.0 billion. The equity de-rated from a 12x earnings multiple to 7x through the trough. The subsequent 2013-2014 cash flow recovery and capex moderation produced a 35% total return as the multiple expanded back to 9-10x.

The 2014-2018 cycle was different in shape but identical in pattern. Capex compressed from $13 billion to $4.5 billion as management prioritised debt reduction and dividend defence. Free cash flow expanded from $4.0 billion to $13.5 billion. The equity re-rated from 7x to 12x through the recovery as iron ore prices stabilised in the $60-80 range. The total return through this period was approximately 80%.

The 2020-2025 cycle is the third and current cycle. Capex has expanded from $6.2 billion in 2020 to $12.4 billion in 2025, with the spending concentrated in the Simandou iron ore project, the Oyu Tolgoi underground expansion, and the lithium investments at Rincon and Jadar. Free cash flow has compressed from $17.7 billion in 2021 to $4.8 billion in 2025. The equity multiple has compressed from 11x trailing earnings to the current 16x on depressed earnings; on a normalised EPS basis the multiple discount is the largest of the three cycles.

We count three completed cycles. The pattern is the same in all three. The cash flow recovery follows the capex peak by 12-24 months. The equity multiple re-rates over the recovery window. The historical median total return through the recovery has been 50-80%. The setup at the current cycle bottom looks visually similar to the 2012 and 2018 troughs.

The current capex cycle is anchored on three flagship programs. Simandou, the high-grade iron ore project in Guinea, has $11.6 billion of remaining capex to first ore expected in 2026. Oyu Tolgoi underground continues its ramp with the next $4-5 billion of incremental capex spread across the next three years. The lithium projects at Rincon (Argentina) and Jadar (Serbia, currently navigating regulatory and community engagement constraints) carry approximately $4-6 billion of capex commitments through 2027.

We are looking at the same data the bear case is looking at and reading it differently. The bear case sees Simandou as a supply-side disaster that floods the seaborne iron ore market. The Research Desk reads Simandou as a structural addition to high-grade pellet feed supply that displaces lower-grade alternatives in the global cost curve and sustains the high-grade premium that Rio Tinto's Pilbara product earns. The bear case sees Oyu Tolgoi as a perpetual capex sink. We see Oyu Tolgoi as a long-life copper asset where the capex spend in 2024-2026 is the necessary investment for two decades of high-margin production. The bear case sees the lithium investments as opportunistic and overpriced. We see them as positioning for a multi-decade decarbonisation theme where the capex outlay buys optionality on a structurally undersupplied commodity.

The one piece of the bear case that holds water is the Jadar regulatory overhang. The Serbian project has been in suspended status for several years and the path back to construction is not yet visible. We have written down the embedded value of Jadar to a low-confidence base case in our internal models. Even with that conservative treatment, the cash flow recovery profile remains intact.

We are buyers of Rio Tinto on the cycle pattern, not on a single-quarter view. The 50-day moving average sits at $94.46 with the 200-day at $77.19. The crossover is constructive. The 52-week range of $52.85 to $101.33 captures the volatility, but the volume profile shows institutional accumulation through the second half of 2025.

The Pilbara iron ore system is the unglamorous workhorse that produces the cash flow floor through every cycle phase. The four major mining hubs (Greater Tom Price, Brockman, Hope Downs, and the West Angelas system) collectively generate roughly 320-340 million tonnes of iron ore annually at C1 cash costs in the high $20s per tonne. The unit economics have improved materially over the past decade as the autonomous haulage fleet has scaled from pilot to standard practice. Roughly 95% of haulage in the Pilbara is now autonomous; the productivity benefit has compounded into a unit cost reduction of approximately 15% over the 2018-2025 window.

The Pilbara cost position is what makes the cycle pattern work. At $80 iron ore, the Pilbara generates approximately 65-70% cash margins. Even at $60 iron ore, which is the bear-case scenario, the cash margin holds in the 45-50% range. Across three iron ore cycles, the Pilbara has never lost money on a cash basis. The cost moat is the structural feature that protects the cash flow floor through the cycle troughs and powers the cash flow recovery as prices stabilise.

The Western Range development, which adds 25-30 million tonnes of replacement capacity over the 2026-2028 window, sustains the Pilbara production base without requiring meaningful additional capex once Simandou clears. The supply replacement timeline is well-managed and the production volume guide remains in the 320-340 mtpa range through the next decade. That visibility is itself a value driver that the bear case routinely under-prices.

Rio Tinto's capital allocation framework through the current trough has been disciplined. The dividend policy targets a 40-60% payout ratio of underlying earnings; the 2025 payout sat at the upper end of that range, signalling management confidence in the recovery. The buyback execution has slowed during the heavy capex phase, which is the correct sequence; preserving balance sheet flexibility through the capex peak is more valuable than buyback execution at multiples that may compress further.

Net debt sits at approximately $11 billion against EBITDA of roughly $20 billion, a 0.55x leverage ratio. The investment grade ratings have not been pressured by the heavy capex cycle, in part because the asset disposals (notably the Iron Ore Company of Canada minority interest sale and the divestment of certain non-core assets) have offset some of the capex outflow. The balance sheet has the room to flex; if commodity prices weakened materially, the dividend coverage could be temporarily relaxed without tripping covenants or rating downgrades.

The Capital Desk view from a Research Desk perspective is that Rio's capital allocation through this trough has been more disciplined than the prior 2008-2012 cycle, when the company over-built capacity and ended up with the disastrous Riversdale acquisition write-down. The lessons from the prior cycle have been internalised. Management is investing where the cost curve position justifies the expense (Simandou high-grade, Oyu Tolgoi long-life copper) and avoiding the empire-building acquisitions that destroyed value in the prior cycle. That discipline supports the cash flow recovery thesis.

The copper book at Rio Tinto deserves a separate analytical treatment because the cycle dynamics are different. Oyu Tolgoi is now in underground production ramp, and the Mongolian asset is on track to produce roughly 500,000 tonnes of copper annually at full ramp, with by-product gold credits that materially lower the cash cost per pound. Kennecott in Utah, the smaller copper asset, contributes another 200,000-220,000 tonnes annually with similar by-product economics. The combined copper book has a multi-decade reserve life and a cost position in the lower half of the global cost curve.

Apply a peer copper multiple to the consolidated copper book. At $9,500 per tonne copper and a 35-40% EBITDA margin, the implied EBITDA from the Rio copper segment runs at roughly $1.8-2.2 billion annualised once Oyu Tolgoi is at full ramp. Apply a 7x EV/EBITDA, conservative against the pure-play copper peer average, and the embedded copper business is worth somewhere between $13 and $16 billion. The current enterprise value of $179 billion divides across the iron ore franchise, the aluminium business, the lithium and minerals book, and the copper book. The copper book is increasingly a meaningful contributor and the bear case has been slow to give it credit.

The long-cycle thesis on copper is well-established. Electrification, renewable infrastructure, and grid expansion all require copper at a pace that the existing supply pipeline cannot easily accommodate. Rio's copper book is positioned to benefit from the secular tailwind through the 2030s. The investment cycle in copper is being executed at the right point in the curve. The historical pattern is that diversified miners with growing copper exposure have re-rated meaningfully through the prior secular copper cycles. The set-up here is consistent with that pattern.