Johnson Matthey's 2026 PGM Market Report confirms that platinum demand will again exceed supply this year, extending a deficit streak that began in late 2024. The autocatalyst sector — which accounts for approximately 40% of global platinum demand — remains the anchor, but the narrative has shifted. Platinum is no longer a purely automotive metal. The hydrogen economy, though still a fraction of total demand, is growing at a trajectory that will make it a meaningful demand driver within five years.
The autocatalyst story is one of substitution, not volume growth. Automakers have been progressively replacing palladium with platinum in gasoline catalytic converters since 2022, driven by a persistent palladium price premium that peaked at nearly $2,000/oz in 2024. That premium has narrowed to approximately $300-400/oz as palladium prices have corrected, but the substitution cycle is long, with automakers locking in current formulations for 3-5 year model cycles. The CME Group's analysis of substitution dynamics estimates that the platinum-for-palladium substitution rate in gasoline catalysts has reached approximately 30% of addressable applications, with potential to reach 50% by 2028.
Mine supply constraints are the second structural pillar of the deficit. South Africa supplies 72% of global platinum, and the operating environment there has deteriorated on multiple fronts. Eskom's load-shedding schedule, while improved from 2024's crisis levels, still imposes 4-6 hours of daily power curtailments on deep-level mining operations. Labor costs have risen 8% year-over-year following the 2026 wage round. And ore grades continue their long-term decline: the average platinum grade at Anglo American Platinum's Mogalakwena mine has fallen from 3.2 g/t in 2015 to 2.4 g/t in 2026, requiring 25% more ore to produce each ounce.
Russian supply — approximately 12% of global primary platinum — faces its own constraints, though for geopolitical rather than geological reasons. Western sanctions on PGM exports remain in place, and while physical supply continues to flow through non-Western channels, the premium for London Good Delivery-certified bars has widened. The WPIC notes that a full Russian export disruption would remove approximately 200,000 oz from annual supply — equivalent to roughly 3% of global production — which alone would tip the market from deficit to acute shortage.
The hydrogen story is the asymmetric upside. Current platinum demand from proton exchange membrane (PEM) electrolysers and fuel cell vehicles is approximately 90,000 ounces per year, according to the WPIC. The pipeline of announced PEM electrolyser projects totals 180 GW globally, though only 12 GW is currently under construction or financed. Even under the WPIC's conservative base case — which assumes 25% of announced projects reach final investment decision — hydrogen-related platinum demand reaches 400,000 oz by 2030. The bull case, which assumes the US 45V clean hydrogen tax credit survives the legislative environment and the EU RED III mandates accelerate, takes demand past 900,000 oz by 2030.
But the hydrogen thesis has a near-term price problem. Green hydrogen production costs currently range from $5-10 per kilogram, making it economically uncompetitive in most applications without subsidies. The Innovation News Network notes that at $8-10/kg, green hydrogen is 3-4 times the cost of grey hydrogen from natural gas. The 45V credit in the US provides up to $3/kg for the cleanest production, which helps but does not fully bridge the gap. Until production costs fall below $3/kg — which requires electrolyser costs to halve from current levels and power prices to average below $30/MWh — hydrogen demand for platinum will grow from a low base rather than explode.
Platinum buyers should be watching the autocatalyst substitution cycle more closely than the hydrogen narrative for 2026-2027. The near-term price driver remains the palladium-platinum spread and automaker formulation decisions, not electrolyzer factory announcements. For physical offtake, the South African supply risk is real but manageable — maintain at least 60 days of inventory buffer against Eskom-related disruptions. The hydrogen thesis is a 2028-2030 story, not a 2026-2027 procurement consideration. However, the forward curve structure (backwardation) suggests that locking in 2027-2028 delivery at current spot levels is attractive if the hydrogen buildout accelerates faster than the base case. The WPIC's deficit projections imply that above-ground inventories are being drawn down at approximately 330,000 oz per year — visible inventory is at its lowest since 2019. Any supply disruption in South Africa or Russia would have an outsized price impact in this environment. Consider allocating 50% of 2027 coverage at current levels and indexing the remainder to the WPIC quarterly balance report to avoid over-paying if the deficit narrows.