What Falling Palladium Supply Means for Investors in 2026

Roughly 80% of the world’s palladium comes from just two countries, and output from both is falling.

That is the problem. Not a risk that might materialise, not a scenario analysts are modelling for the future. It is happening in the production data right now, across every major dataset that tracks it.

Here is why that matters at this moment. Primary mine supply is in a multi-year downtrend with no meaningful wave of new projects coming online, while demand from hybrid vehicles, electronics, and AI-linked infrastructure is holding up better than most investors assume. When you think about critical minerals, you probably think about lithium, cobalt, and copper. Palladium rarely enters the conversation, and that gap is exactly where the interesting questions live.

This piece gives you what the data actually shows about where palladium supply is heading, and whether the structural story is as compelling as it first appears. You will get the supply-crunch case in full, but you will also get the counter-case that serious analysts make against it, because the honest version of this story is not a one-way bet.

Where palladium actually comes from, and why that is a problem

Hold a simple map in your head. Palladium production concentrates in Russia and South Africa, and almost nowhere else matters at scale. Canada, the United States, and Zimbabwe contribute meaningful volumes, but they sit far behind the two dominant hubs.

According to the Critical Materials Atlas, Russia supplies approximately 44% of global unwrought palladium exports and South Africa approximately 34%. Together that is roughly 78% of tradable supply controlled by two countries.

Roughly 78% of the world’s tradable palladium supply is controlled by just two countries. When their output moves, the global market moves with it.

Now look at what that output is doing. The USGS Mineral Commodity Summaries 2025 estimated Russia at 75,000 kg and South Africa at 72,000 kg for 2024, against a world total of around 189,000-190,000 kg. TrueSource Metals, in its September 2026 dataset, put the same two countries higher, at 89,000 kg and 82,600 kg respectively for 2024, with a world total of 217,000 kg.

The two sources disagree on the absolute figures, likely because of differing methodologies and reporting lags. What they agree on is the direction, and the direction is down.

Norilsk Nickel’s output trajectory is the single most consequential variable in the Russia supply picture, because the company accounts for the overwhelming majority of Russian palladium production and its operational decisions propagate directly into the global supply balance.

The USGS put Russia at 92,000 kg for 2023, so its own figures describe a fall of roughly a fifth in a single year. TrueSource’s 2025 estimates of 84,000 kg for Russia and 70,000 kg for South Africa continue the same slide.

This is where the concentration cuts deeper than most battery-metal stories. If you understand supply risk through lithium or cobalt, you probably assume concentration is a known factor that markets have priced in. With palladium, the concentration is more extreme, and the pathway back to higher supply is far less visible.

No pipeline, no relief

The absence of new projects here is structural, not cyclical. Demand uncertainty from electric vehicles, prices well below the 2018-2022 peaks, and the long lead times of greenfield mine development all suppress the appetite to build.

That is the part investors underweight. Declining output is not just a data point; it reflects the near-absence of a recovery mechanism, because there is no queue of large-scale projects ready to fill the gap.

Chalice Mining’s Gonneville deposit in Western Australia stands out as one of very few advanced-stage primary palladium projects anywhere in the world. It is worth holding as an example of what genuinely new primary supply looks like, and we return to it later, because its timeline tells you something the production tables cannot.

Why palladium demand is not collapsing as fast as the EV narrative suggests

You have probably absorbed a tidy assumption: electric vehicles are killing palladium demand. It is true in direction, but the timing is far more textured than a headline EV adoption rate implies, and the hybrid vehicle is the reason.

Hybrids (both HEVs and PHEVs) keep a full gasoline internal-combustion engine alongside their electric drivetrain. That engine still needs a three-way catalytic converter, and palladium is a primary active component inside it, driving the chemical reactions that clean the exhaust.

Because hybrids switch their engines on and off constantly and must control emissions across a wide range of conditions, their catalyst systems are engineered to demanding standards. Palladium loadings are often comparable to, or higher than, those in a conventional petrol car. Every hybrid sold keeps generating real palladium demand.

China’s BEV transition is the single largest variable in the demand erosion timeline, because China accounts for the largest share of new vehicle sales globally, and the pace at which Chinese consumers shift away from petrol engines sets the floor under hybrid-era palladium consumption.

The numbers show the effect. Johnson Matthey’s PGM Market Report 2024 put primary palladium supply at 6.547 Moz for 2023 and forecast 6.454 Moz for 2024, with the market remaining in deficit. Metals Focus estimated that palladium automotive use fell approximately 7% in 2024, pulling total demand down to around 9.73 Moz and shrinking the deficit from roughly 1.02 Moz to 0.358 Moz.

Automotive palladium demand fell about 7% in 2024, yet the market stayed in deficit at 0.358 Moz. The shortfall shrank because demand fell, not because supply grew.

Read that carefully, because it is the pivot of the whole near-term story. The deficit narrowed not because mines produced more, but because demand fell faster than mine output, and the hybrid bridge is what kept demand from falling faster still.

There is a second, quieter demand channel worth knowing. Palladium appears throughout electronics, and AI-linked infrastructure is expanding the places it turns up:

• Multilayer ceramic capacitors (MLCCs) in smartphones, PCs, and servers
• Connector and contact plating across networking equipment
• Certain semiconductor packaging applications
• Data centre servers powering AI workloads
• Networking hardware supporting compute build-outs

Be honest about the limits of this thesis, though. Current sources treat AI infrastructure as an incremental demand driver, not a transformational one, and no quantified AI-specific palladium forecast exists in the reviewed data.

So here is what the deficit really tells you. Primary supply is falling fast enough that demand does not need to hold up well for the market to stay tight in the near term. The EV transition is real, but its bite on palladium is being absorbed gradually by the hybrid stage, and that stage is running at the same time as a shrinking supply base. Anyone assuming a clean petrol-to-battery switch is missing the layered timing underneath it.

The case against the supply-crunch thesis: what the bears have right

The tight story is credible. It is also not the whole story, and the counter-case deserves genuine weight rather than a polite nod, because the data behind it is just as real.

Start with substitution. Palladium and platinum are partially interchangeable in catalytic converters, and catalyst formulators shift loadings toward platinum when palladium gets expensive. Metals Focus cited exactly this as a key reason the 2024 average palladium price was expected to fall approximately 23% year-on-year to around US$1,030/oz, even with the market in deficit. Substitution is a demand-side release valve that opens automatically whenever scarcity fear pushes prices up.

The PGM market balance is not uniform across the platinum group: platinum was in structural deficit through 2025-2026 while palladium moved toward surplus on the recycling growth path, and those diverging trajectories matter for how you read any single institution’s supply-demand numbers.

Then there is recycling, and this is where the structural counter-argument gets serious.

Recycling as the structural release valve

The mechanics are straightforward. Automotive scrap from existing petrol and hybrid vehicles is the feedstock, and as more of that fleet reaches end of life, recycled palladium volumes grow.

The World Platinum Investment Council (WPIC), in its September 2025 Essentials report, projects total palladium supply rising from approximately 9,414 koz in 2024 to approximately 9,997 koz in 2029, a gain of around 7%, even as mine supply falls. Recycling grows fast enough to more than offset the decline. On that path, market deficits are expected to become surpluses around 2027 and turn material by 2029.

Sit with what that implies. The supply buffer builds without a single new mine needing to succeed. The relief comes from cars already on the road, not from projects that might or might not clear feasibility.

There is a caveat that cuts both ways. Inventory transparency in palladium is poor, thanks to industrial stockpiles, ETF structures, and opaque holdings, so nobody knows precisely how large the above-ground buffer already is.

Now look at where the institutions land for 2026, because the spread is the whole point.

UBS, in September 2026, raised its December 2026 and March 2027 targets by US$200/oz each, citing supply tighter than previously expected, but framed the move as a medium-term adjustment rather than a long-term bullish call.

The gap between HSBC at US$1,135/oz and Bank of America at US$1,725/oz is not analysts being confused. It reflects a genuinely binary outcome, where the timing of demand erosion against supply decline decides which way the market breaks. The most useful thing you can take from this section is permission to hold that uncertainty rather than resolve it artificially. Both the supply-crunch case and its rebuttal are supported by real data, and the forecasters are not converging.

What genuine new primary supply actually looks like, and why it is rare

To understand how hard it is to add palladium to the world, look at the most advanced credible attempt to do it. Chalice Mining’s Gonneville project in Western Australia is a large polymetallic deposit where precious metals make up most of the net revenue, with palladium the single largest contributor.

The December 2025 Pre-Feasibility Study describes a mine of real scale:

• 220,000 oz of 3E PGMs (palladium, platinum, gold) per year
• 8,000 tonnes of copper annually, around 25% of net revenue
• 7,000 tonnes of nickel annually, around 20% of net revenue
• 700 tonnes of cobalt annually

The project carries “strategic project” status from the Western Australian government and “major project” status from the Australian federal government, and Chalice is reported as funded through to a Final Investment Decision. This is about as strong a starting position as a new palladium project gets.

Now watch the timeline, because this is where the lesson lives.

A pilot plant is scheduled for Q3 2026 as a metallurgical de-risking step ahead of the feasibility study, which is targeted for the second half of 2027. The environmental ministerial decision and Final Investment Decision are both expected in H1 2028, with first production around 2030 on a two-year construction assumption.

Here is what that runway means for you. Even the most advanced non-Russian, non-South African primary palladium project in the world is roughly four years from first production. The market must absorb a stretch of falling primary mine supply before any new primary ounces from a project like this arrive. Gonneville gives you a concrete benchmark: this is what genuine new supply looks like, this is how long it takes, and this is why there is no short-term fix even when a large, advanced project genuinely exists.

Making an informed call in a rate-plateau palladium market

You now have the full picture, and the picture is deliberately two-sided. The structural primary supply case is credible and data-supported. Yet the market has already priced in some of the recycling and substitution dynamics that cap the upside, which is exactly why the 2024 average price fell despite an ongoing deficit.

As of late September 2026, palladium was oscillating in the low-to-mid US$1,290-1,330/oz range, according to Kitco and Trading Economics. That is the level from which any thesis has to prove itself.

Three variables will decide whether the supply-crunch story becomes a genuine price catalyst or gets quietly defused:

• The pace of hybrid vehicle sales, which extends or shortens the demand bridge
• The speed at which automotive recycling volumes scale, which sets the WPIC surplus timeline around 2027
• Whether the EV transition outside China accelerates or stalls

The supply side is the piece that does not easily reverse. There is no meaningful new project pipeline in Russia or South Africa, and Gonneville’s roughly 2030 first-production target shows how long fresh primary supply takes to arrive.

Palladium sits at an unusual intersection, where genuinely constrained primary supply meets growing secondary supply and slow demand erosion. That makes it a story that rewards careful timing and threshold-watching, not a simple long-conviction trade. If you can read the sequencing of these forces, you hold an edge that most critical-mineral narratives never offer cleanly.

For readers wanting a single-source institutional view of how supply, demand, and price interact across all three major PGMs in 2026, our dedicated guide to the 2026 PGM market outlook covers CPM Group’s full forecast framework, including the recycling growth assumptions that underpin the surplus trajectory.

This article is for informational purposes only and should not be considered financial advice. Investors should conduct their own research and consult with financial professionals before making investment decisions. Past performance does not guarantee future results, and financial projections are subject to market conditions and various risk factors.