There is a moment in every great market dislocation when the facts on the ground diverge so dramatically from the prices on the screen that future historians will marvel at the collective blindness of an entire generation of institutional investors. We are living through such a moment now.
On December 17, 2025, platinum traded at $1,931 per ounce—its highest level since September 2011—while gold breached $4,334. The ratio between these two metals has collapsed to 0.44, meaning one ounce of gold now purchases 2.3 ounces of platinum. This relationship has not been this distorted since 1907. For 115 years, through two world wars, the Great Depression, the abandonment of the gold standard, the rise and fall of the Soviet Union, the birth of the internet, the 2008 financial crisis, and the COVID pandemic, never has platinum been this cheap relative to gold.
And yet, the platinum market is not in surplus. It is experiencing its third consecutive year of structural deficit, with cumulative shortfalls exceeding 2.7 million ounces since 2023. Above-ground stocks have been depleted to levels representing barely four months of demand coverage—the thinnest buffer in modern market history. Physical lease rates spiked above 25% annualized in mid-2025, a stress signal so extreme it typically precedes violent price dislocations. The London market entered backwardation. Sponge platinum began trading at parity with investment-grade bars—an anomaly so rare that veteran traders described it as unprecedented.
This is not a market waiting for a catalyst. This is a market that has already received multiple catalysts and is only beginning to reprice.
The question that should occupy every serious allocator is not whether platinum is mispriced—the evidence for mispricing is overwhelming and verified across multiple independent data sources. The question is why institutional capital has been so slow to recognize what is hiding in plain sight, and what that failure reveals about the structural blindness embedded in modern portfolio construction.
The answer involves a collision of narratives that have blinded the market to physical reality: the death-of-diesel thesis, the electric vehicle inevitability thesis, and the broader assumption that any metal associated with internal combustion engines must be in terminal decline. These narratives contain kernels of truth wrapped in layers of analytical laziness. They have created one of the most asymmetric opportunities in commodity markets—a metal trading at a 115-year discount to its monetary cousin while experiencing persistent physical shortages.
This report is an attempt to see clearly. To separate validated facts from convenient assumptions. To identify where the consensus has confabulated predictions into certainties. And to map the causal chains that connect today’s market structure to the wealth transfers that will define the coming decade.
Part I: The Supply Architecture Nobody Understands
The platinum supply chain is the most geographically concentrated of any major commodity in the world. This is not hyperbole. It is a measurable fact that places platinum in a category of vulnerability shared by no other metal of comparable economic importance.
South Africa produces approximately 72% of global platinum supply from the Bushveld Igneous Complex, a geological formation that contains the world’s largest known platinum reserves. Russia contributes another 12%, primarily as a byproduct of Norilsk Nickel’s operations. Zimbabwe accounts for roughly 8%. Together, these three nations produce 92% of the world’s platinum. The Herfindahl-Hirschman Index for platinum-producing nations stands at 5,355—classified as “extremely high concentration” by any standard measure of supply chain risk.
To understand why this matters, consider what it means operationally. There is no swing producer capable of ramping output in response to price signals. There is no strategic petroleum reserve equivalent that governments can release to calm markets. There is no readily available substitute that can be deployed at scale within any reasonable timeframe. When platinum supply is constrained, there are no good options—only varying degrees of bad ones.
And supply is constrained.
South African platinum production has entered what can only be described as structural decline. The 2025 figures tell the story: mine supply is projected to fall 6.4% to approximately 3,869,000 ounces—the lowest level in roughly 25 years, excluding the anomalous disruptions of the 2014 strike and the COVID pandemic. This is not a temporary setback. It is the mathematical consequence of geological depletion, infrastructure decay, and a capital investment drought that began years ago and whose effects are only now becoming visible in production data.
The geological story is particularly instructive for understanding why supply cannot respond to price signals in any conventional timeframe. Analysis of Anglo American Platinum’s reserve grades from 2007 to 2020 reveals a weighted average decline of 30% across the Merensky, UG2, and Platreef reefs. The industry is systematically transitioning from higher-grade platinum-rich ore to lower-grade formations that are more palladium and rhodium-weighted. This means more tonnes must be mined, hoisted, crushed, and processed to yield the same volume of platinum metal. Energy intensity per ounce is rising. Labor intensity per ounce is rising. Capital intensity per ounce is rising.
At the same time, the infrastructure required to execute this more difficult extraction is aging and underfunded. Mine shaft depths at operations like Impala Rustenburg approach one kilometer. Sibanye-Stillwater’s K4 project targets 1,287 meters. At these depths, continuous ventilation and refrigeration are not optional—they are requirements for human survival. The power demands are enormous, the capital requirements are staggering, and the lead times for new capacity are measured in decades, not years.
The Eskom narrative requires careful recalibration, however. The existing analysis overstates the current electricity crisis while understating its lasting structural damage. As of December 2025, South Africa has experienced 210 consecutive days without load-shedding—a dramatic improvement from the 200-plus days of rolling blackouts that characterized 2023 and early 2024. The Energy Availability Factor has risen 8.33% year-over-year to 66.52%, and official projections show no scheduled load-shedding through March 2026.
This improvement is real but does not reverse the damage already inflicted. Anglo American Platinum alone lost 105,000 PGM ounces in 2022 due to power disruptions—verified against company disclosures. Impala Platinum held 200,000 ounces of unsold metal worth R7.2 billion at fiscal year-end 2023 due to processing bottlenecks. More importantly, the years of unreliable power supply have fundamentally altered investment calculations. No rational mining executive will approve a decade-long capital project predicated on Eskom reliability, regardless of current performance. The “Eskom discount” is now permanent, baked into every expansion decision and every financing model.
The logistics story is arguably worse than the power story, and receives far less attention. Transnet, the state-owned rail and port operator, has become what one National Treasury analysis described as a greater constraint on economic activity than Eskom itself. The quantified damage is staggering: R411 billion in economic losses in 2023 alone, according to official government estimates. R98 billion in lost mineral exports over 2021-2023. One billion rand in daily economic damage at the peak of the crisis.
The structural response is underway—eleven private rail operators received 10-year licenses in August 2025, with official targets to reach 250 million tons of annual freight capacity by 2029—but the timeline for meaningful improvement stretches years into the future. The damage to platinum supply in the interim is not recoverable. Ore that was not mined in 2024 because it could not be transported will never be mined. That supply deficit is permanent.
The recycling story completes the supply picture, and it contradicts nearly everything the market assumed about secondary supply elasticity. Total platinum recycling fell to 1,486,000 ounces in 2024—the lowest level since record-keeping began in 2013. The 2025 recovery to approximately 1,600,000 ounces represents a modest improvement but remains roughly 25% below pre-COVID levels.
The conventional economic assumption—that higher prices stimulate higher recycling flows—has been empirically falsified by the platinum market. The reasons are structural rather than cyclical. Consumer behavior has shifted toward extended vehicle ownership, reducing the flow of end-of-life catalysts into the scrap stream. Collection networks were decimated during the low-price years of 2023-2024, and have not recovered. Regulatory crackdowns on catalyst theft added compliance costs that squeezed margins for smaller recyclers, forcing consolidation and reducing overall collection capacity.
Most importantly, the recycling system operates on lags that are measured in years, not months. A catalytic converter installed in a vehicle today will not enter the recycling stream for 12-15 years on average. The scrap flowing into recycling today reflects vehicle sales decisions made during the 2008-2012 period. The “recycling cliff” hypothesis—that secondary supply could fall even further as the vehicle parc ages and catalyst substrate volumes decline—remains a credible risk scenario, though its timing is uncertain.
The synthesis of these supply factors creates a picture that institutional models have systematically underweighted. Primary supply is in structural decline. Secondary supply has collapsed despite rising prices. Geographic concentration creates geopolitical vulnerabilities that cannot be hedged. Infrastructure constraints in the dominant producing nation are structural rather than cyclical. Capital investment has been cut to preserve cash flow, guaranteeing production shortfalls 18-36 months forward. And there are no new major projects in the pipeline capable of altering this trajectory within any reasonable investment horizon.
Part II: The Demand Narrative That Deceived a Generation of Analysts
The platinum demand story has been systematically misunderstood because it has been filtered through the lens of a single narrative: electric vehicles will destroy autocatalyst demand, and therefore platinum is a secular short.
This narrative is not entirely wrong. Electric vehicles do not require catalytic converters. Global battery electric vehicle penetration is rising. European diesel market share has collapsed. These facts are accurate. What is inaccurate is the conclusion that follows—that platinum demand is in terminal decline and that current market deficits are merely transitory.
The December 16-17, 2025 announcement from the European Union represents the most significant regulatory development for platinum demand in a decade, and it directly contradicts the EV-extinction thesis. The EU Commission announced it is scrapping the previously planned 2035 ban on internal combustion engine vehicles, proposing instead a 90% CO2 reduction target that explicitly permits continued sales of hybrids, plug-in hybrids, and certain ICE vehicles beyond 2035.
The implications are profound. The 2035 deadline had been priced into automaker production planning, capital allocation, and supply chain decisions. Its reversal extends the runway for autocatalyst demand by at least a decade beyond what markets had assumed. Hybrid vehicles—which the policy now accommodates—require catalytic converters. Many hybrid configurations require higher PGM loadings than conventional ICE vehicles due to their stop-start operating cycles and lower exhaust temperatures.
The substitution story requires similar recalibration. For years, automakers responded to palladium’s price premium over platinum by engineering platinum back into gasoline autocatalysts. This substitution reached critical mass in 2024-2025, adding approximately 700,000-742,000 ounces of platinum demand annually. Once an automaker certifies a vehicle platform with a specific catalyst formulation, that formulation is locked in for the platform’s 7-year lifecycle. The engineering investment has been made. The regulatory certification has been obtained. The supply chains have been configured. Reverse substitution is not impossible, but it is slow, expensive, and subject to the same multi-year certification cycles.
The market’s assumption was that substitution would peak at approximately 1 million ounces annually in 2025. The actual figure appears closer to 845,000-877,000 ounces—revised downward because the platinum-palladium price differential narrowed faster than expected. This is the correct analytical response to changing data, and it illustrates why the substitution story requires nuance rather than simple extrapolation.
What has not been adequately appreciated is the asymmetry of the substitution mechanism. The 700,000+ ounces of annual substitution that occurred between 2022 and 2025 represents demand that will persist for 7-year platform lifecycles regardless of short-term price movements. It is structural rather than tactical. And it has occurred against a backdrop of tightening emissions standards—China VI, Euro 7—that increase PGM loadings per vehicle even as total ICE vehicle counts decline.
The heavy-duty diesel segment deserves particular attention because it contradicts nearly every element of the EV-inevitability narrative. Heavy trucks remain 95% diesel-powered globally. Electric heavy-duty vehicles represent approximately 2% of sales. The Department of Energy projects that zero-emission heavy trucks will not achieve cost competitiveness until 2035 at the earliest. In this segment, platinum’s position is not merely secure—it is strengthening, as tighter emissions regulations drive higher catalyst loadings.
The jewelry demand story has experienced an unexpected renaissance driven by platinum’s extreme discount to gold. Chinese platinum jewelry fabrication recorded its first growth since 2013-2014. The World Platinum Investment Council reports that approximately ten new platinum-orientated wholesale showrooms opened in Shuibei—representing 90% of China’s jewelry manufacturing hub—in early 2025. This effectively tripled platinum’s wholesale presence in China’s largest jewelry production center. Over 40 platinum-dedicated wholesale showrooms have opened since January 2025.
India has emerged as the fastest-growing jewelry market, with demand surging 31% year-over-year. Platinum is positioning itself as the metal of choice for younger consumers seeking alternatives to gold’s inflated prices. Total jewelry demand reached approximately 2 million ounces in 2024, growing 8% annually despite—or perhaps because of—platinum’s undervaluation.
Investment demand represents the swing factor that could either sustain or undermine the bullish thesis. In 2024, investment demand surged 77% to 702,000 ounces, with China accounting for 64% of global bar and coin purchases. This represents a dramatic shift from 2019, when China represented just 11% of investment demand. The launch of China Gold Coin Group’s first 1-kilogram platinum bar signals official recognition of platinum’s strategic importance. The Guangzhou Futures Exchange’s December 2025 launch of physically-settled platinum futures provides the first real transparency into Chinese platinum inventories.
The synthesis of these demand factors creates a picture fundamentally at odds with the terminal-decline narrative. Autocatalyst demand is proving more resilient than projected, supported by hybrid growth, substitution lock-in, and policy reversals. Jewelry demand is growing, driven by platinum’s extreme discount to gold. Investment demand is surging, led by Chinese strategic accumulation. Industrial demand—in glass manufacturing, petroleum refining, chemical processing, and medical applications—provides a stable baseload. And hydrogen-related demand, while currently small at approximately 41,000 ounces, represents optionality that the market has not adequately priced.
Part III: The Hydrogen Mirage and What It Actually Means
The hydrogen economy represents platinum’s most compelling long-term demand driver and simultaneously its most overhyped near-term catalyst. Separating these two realities is essential for any serious analysis.
Platinum serves as the critical catalyst in two technologies underpinning the green hydrogen ecosystem: Proton Exchange Membrane water electrolyzers, which produce hydrogen from water using renewable electricity, and PEM fuel cells, which convert hydrogen back into electricity for transportation and stationary power applications. Both technologies require platinum. Neither has a viable platinum-free alternative at scale.
The physics are straightforward. PEM electrolyzers currently require approximately 0.8 grams of platinum per kilowatt of capacity, though this figure varies by manufacturer and technology generation. The Department of Energy has established a target of 0.1 grams per kilowatt by 2026, and several companies are now marketing catalysts that approach this target. Heavy-duty fuel cell trucks require 45-120 grams of platinum per vehicle, with loadings at the higher end for applications demanding the 30,000-hour operational lifetimes required for commercial viability.
The scale projections are where analysis typically goes wrong. The World Platinum Investment Council projects hydrogen-related platinum demand reaching approximately 900,000 ounces by 2030—representing roughly 11-12% of total supply if deficits persist. This projection requires certain assumptions to prove correct: global PEM electrolyzer capacity must scale from approximately 1.4 gigawatts today to over 200 gigawatts cumulative by 2030, representing a compound annual growth rate exceeding 90%.
Is this achievable? The announced pipeline suggests it might be—over 520 gigawatts of electrolyzer capacity has been announced globally. But announcements are not deployments. As of late 2025, only approximately 4% of announced capacity has reached Final Investment Decision. The gap between announcement and execution is measured in years and billions of dollars of capital expenditure.
The fuel cell vehicle story is similarly nuanced. The global FCEV fleet stands at approximately 70,000-90,000 vehicles—somewhat below the frequently cited “100,000” figure but within the same order of magnitude. More importantly, FCEV adoption has stalled and in some segments reversed. Hydrogen refueling infrastructure has collapsed in key markets—Shell closed its California stations, and 75% of South Korean stations closed in 2023. Battery electric vehicles have won the passenger car competition decisively, and the infrastructure network effects are increasingly insurmountable.
The heavy-duty segment tells a different story. Trucks, buses, and maritime applications face range and weight constraints that batteries cannot overcome with current or near-term technology. In these segments, hydrogen fuel cells remain competitive, and several major OEMs are investing accordingly. The timeline is longer than passenger vehicles—commercial-scale heavy-duty fuel cell adoption is more likely a 2030-2035 phenomenon than a 2025-2027 reality.
The iridium constraint deserves mention because it shapes the upper bound of PEM electrolyzer deployment. Current PEM technology requires iridium as an anode catalyst, and global iridium production of approximately 250,000 ounces annually would be fully consumed by just 20 gigawatts of PEM capacity at current loadings. This creates a physical ceiling on PEM deployment that can only be raised through technological innovation—thrifting, recycling, or alternative catalyst development. Progress is occurring—Heraeus has developed catalysts with 50-90% less iridium, and various research programs target 90% reductions—but the timeline for commercial deployment remains uncertain.
The alkaline electrolyzer competition is real and must be acknowledged. Alkaline electrolyzers dominate the current market with approximately 80% share globally, and they require no platinum or iridium. China’s electrolyzer market is approximately 90% alkaline due to cost considerations. If alkaline technology captures the majority of hydrogen production growth, platinum’s hydrogen thesis weakens significantly. The counterargument—that PEM’s ability to handle intermittent renewable power makes it better suited for green hydrogen applications—is technically valid but has not yet proven decisive in market share competition.
The honest assessment is that hydrogen represents material but not transformational upside for platinum over the next five years, with potential for significant impact in the mid-2030s if deployment trajectories accelerate. Modeling hydrogen demand as a 2025-2027 catalyst is analytically unsound. Modeling it as a 2030-2040 structural driver is reasonable, with substantial uncertainty around timing and magnitude.
Part IV: The 115-Year Anomaly and What It Reveals
The platinum-to-gold ratio is the single most important metric for understanding platinum’s current valuation, and it tells a story of extreme dislocation that defies any explanation grounded in fundamentals.
