Gold’s appeal rests partly on a physical constraint: obtaining more of it is difficult. What happens if technology changes that constraint?
Lab-grown diamonds provide a striking comparison. Manufacturing expanded access to a once-scarce luxury material, challenged natural producers and drove down synthetic prices. But translating that experience into a prediction that gold will collapse skips the essential questions: what has actually been demonstrated, what does a saleable ounce cost, and how much new supply could reach buyers?
The evidence supports a serious research question. It does not yet support a dated forecast for cheap, industrial-scale manufactured gold.
Yes, making gold is possible
Lab-grown diamonds are carbon atoms arranged into a diamond crystal. GIA describes laboratory and natural diamonds as having essentially the same chemical, physical and optical properties, while specialised testing can distinguish their growth origins. Producing gold from another element requires changing an atomic nucleus. Ordinary chemistry cannot do that. Refining, recycling, plating and growing gold particles from gold-bearing solutions use existing gold rather than create new gold atoms. [1] [2]
CERN’s ALICE collaboration measured lead-to-gold transmutation at the Large Hadron Collider. Across four experiments during 2015–2018, approximately 86 billion gold nuclei were produced—equivalent to just 29 picograms, or 0.000000000029 grams. Those energetic nuclei subsequently fragmented against accelerator components. This was an important physics measurement, not a stockpile of recoverable bullion. CERN also makes clear that artificial gold production predates this experiment. [2]
The newer commercial proposal comes from Marathon Fusion. Its 2025 preprint models fusion neutrons converting mercury-198 into mercury-197, which then becomes stable gold-197 through electron capture. The simulated yield is approximately two tonnes per gigawatt of thermal fusion power per year. This is a model result, not verified commercial output. [3]
There is now an additional claim to assess. In a company announcement reported on 21 October 2026, Marathon said it had produced stable gold from mercury using fusion neutrons, describing its result as ten times a CERN record. The statement reviewed does not establish recovered mass, a fully costed price per ounce or independently replicated commercial performance. Its headline comparison should not be treated as a like-for-like bullion-production benchmark. [4]
| Evidence | What it establishes | What remains open |
|---|---|---|
| Established nuclear transmutation | Gold nuclei can be created from other elements. | An economical manufacturing route. |
| Marathon’s October 2026 company claim | A reported experimental advance. | Independent verification, recoverable quantity and repeatable throughput. |
| Published fusion simulation | A proposed route to substantial output. | Plant engineering, costs, product clearance and actual deliveries. |
These are distinct stages. A successful experiment would strengthen the case for further development without proving that a competitive gold factory exists.
What actually happened to diamonds?
Edahn Golan’s published lab-grown diamond wholesale price index fell 92% from July 2018 to Q3 2026. Yet the latest quarter also brought an approximately 2% increase in the tracked round-diamond categories; the overall index remained down 10% year over year. The market was showing uneven signs of stabilisation after a dramatic repricing. These are wholesale measures for loose laboratory-grown stones, not a 92% decline in natural diamonds, finished jewellery or every owner’s resale value. The index’s coverage has evolved over time, which also matters when interpreting the long comparison. [5]
The consequences extended through the industry. In May 2025, De Beers announced plans to close its Lightbox laboratory-grown jewellery business and focus its Element Six synthetic-diamond activity on industrial applications. [6]
Natural producers also suffered. De Beers reported a 12% decline in its average rough-diamond price index for 2025 and a $511 million underlying EBITDA loss. Its discussion identified consumer substitution towards laboratory-grown diamonds alongside surplus supply, weak Chinese demand and tariff disruption. Those results do not isolate how much damage came from synthetics alone. [7]
By the first half of 2026, De Beers still reported pressure on smaller, lower-quality stones, while larger, higher-quality natural diamonds were more resilient. Its underlying EBITDA loss narrowed to $113 million. This was an uneven industry adjustment, not the disappearance of the natural-diamond market. [8]
The economic lesson is broader than a falling price chart. Customers can gain larger or more affordable products while miners, manufacturers and retailers lose some pricing power. Retailers may sell more units but earn fewer dollars per item. Inventory bought at yesterday’s prices can become a liability. Even successful manufacturers face competition that reduces the scarcity premium they initially hoped to capture.
Three other markets offer useful lessons
| Market | What changed | What it suggests for gold |
|---|---|---|
| Aluminium | Commercial electrochemical extraction made an abundant but difficult-to-refine element affordable. ACS records ingot prices falling from $4.86 per pound in 1888 to $0.78 in 1893. [9] | An approximately 84% price fall can accompany expanding industrial usefulness. This was cheaper extraction, not creation of a new element. |
| Synthetic ruby and sapphire | Laboratory crystal production created supplies for jewellery and industrial applications while natural material remained a separate category. [10] | Provenance can retain value where buyers care about origin. That does not automatically protect a standardised bullion price. |
| Industrial diamonds | Synthetic material accounts for more than 99% of global industrial-diamond production and consumption, according to USGS’s 2026 summary. [11] | In performance-driven applications, reliable manufactured material can become dominant. This statistic is separate from the gem-jewellery market. |
Technology can therefore produce several outcomes: cheaper mass-market goods, a surviving premium niche, or near-complete substitution in a particular use. Which outcome emerges depends on what buyers value and how production scales.
Gold has a different market structure
Gold is a jewellery material, an industrial input and a financial asset. The World Gold Council estimates approximately 222,600 tonnes above ground at the end of Q2 2026, including roughly 39,000 tonnes held by central banks. [12]
That accumulated stock changes the analysis. A small new production stream would enter a market dominated by existing holdings. However, existing gold is not all available for sale, and stock size does not immunise the price against changes in investors’ willingness to hold it.
There is also a potential weakness in the diamond analogy. Diamond buyers can pay for an identifiable natural origin. If purified manufactured gold eventually meets the same commercial requirements as mined gold, our inference is that standard bullion would face more direct substitution. A gold atom does not acquire a different elemental identity because of its production route. Historical coins, exceptional nuggets and provenance-based jewellery could retain premiums without insulating ordinary bullion.
Chemical identity alone would not guarantee market access. LBMA’s London Good Delivery framework includes requirements for approved refiners, bar specifications and vault acceptance; minimum gold fineness is 995 parts per thousand. Whether a particular manufactured product satisfies applicable radiological, sourcing and market requirements would need to be established. We found no basis to declare blanket acceptance or blanket rejection of future transmuted gold. [13]
The real hurdle is the cost of a deliverable ounce
A compelling revenue projection is only one part of an investment case. A credible cost model must include construction and financing, energy and fuel, feedstock preparation, separation and recovery, maintenance, environmental controls, waste management, assay, storage and insurance. It must also account for downtime and the time between production and payment.
Product purity can impose a significant delay. Gold-197 is stable, but other isotopes in a production mixture can be radioactive. The 2025 study calculates a 17.7-year cooling period for one conservative activity comparison under its assumed isotope mix; feedstock purification changes that result. This is a model-specific illustration, not a universal waiting period or a regulatory declaration that the product is safe for consumers. [3]
For investors, delayed saleability means additional capital tied up in inventory. Financing that delay can materially change apparent production economics.
Revenue from electricity or other isotopes could help support a plant. Conversely, a business plan that relies on selling gold at a high price becomes less attractive if its own success pushes that price down. This feedback could slow expansion. A profitable specialist isotope business would also not, by itself, establish economical bulk gold production: the products have different prices, volumes and requirements.
Nor can an environmental advantage simply be assumed. Replacing some mining might reduce certain impacts, while reactor construction, energy use, mercury handling and waste controls create others. A meaningful comparison requires a complete assessment per saleable ounce.
How much supply could matter?
For scale, the World Gold Council’s January 2026 report estimated 2025 mine output at 3,671.6 tonnes and total annual supply at 5,002.3 tonnes, including recycling and net producer hedging. These are historical estimates subject to revision, not forecasts for 2026. [14]
The following is APEX arithmetic using the published simulation’s approximate yield. It is not a deployment forecast. “GW thermal” measures fusion heat output, not electricity supplied to the grid. [3]
| Hypothetical operating capacity | Model-implied gold per year | Relative to 2025 mine output | Relative to Q2 2026 above-ground stock |
|---|---|---|---|
| 1 GW thermal | 2 tonnes | 0.05% | 0.0009% |
| 10 GW thermal | 20 tonnes | 0.54% | 0.0090% |
| 100 GW thermal | 200 tonnes | 5.45% | 0.0898% |
| 500 GW thermal | 1,000 tonnes | 27.24% | 0.4492% |
Assumptions: capacity operates at the modelled annual yield. These figures do not separately deduct operational shortfalls or recovery losses, and are not immediate saleable supply; cooling or clearance could delay deliveries. Stock percentages compare one year’s hypothetical production with a fixed historical stock. They are not cumulative impacts.
This illustrates why an experiment and a market disruption are far apart. A large fleet could create a meaningful addition to annual supply, even while that addition remains small relative to accumulated holdings. But a 5% supply addition does not translate mechanically into a 5% price decline. Demand responses, recycling, mine closures, inventory decisions and expectations all matter.
Three possible paths for the gold price
Scientific progress with limited commercial output. If costs remain high or throughput small, manufactured gold would contribute little physical supply. Announcements could still create short-lived volatility, while interest rates, currencies, central-bank activity and investment demand continue to shape prices.
Competitive production with constrained expansion. Verified low costs and a credible factory pipeline could reduce expectations of future scarcity before much metal arrives. High-cost mines and new projects would face tougher investment decisions. Lower mine output or reduced recycling could offset part of the new supply.
Cheap production that scales repeatedly. If manufactured gold becomes broadly accepted and capacity can expand economically, the long-term scarcity premium could weaken substantially. Industrial demand might rise at lower prices, but whether that would compensate for reduced investment demand is unknown. The equilibrium would depend on production costs, capacity and the value people continue to place on holding gold.
These are conditional paths, not assigned probabilities or price targets. The sources reviewed do not justify importing the diamond index’s 92% decline into a gold forecast, setting a collapse date, or attributing a particular gold-market move to the latest company announcement.
The effects would spread beyond bullion
The following are APEX’s conditional market inferences, assuming competitive manufactured supply becomes credible and eventually reaches buyers.
| Market or stakeholder | Potential consequence | What determines the result |
|---|---|---|
| Bullion, physically backed funds and gold-backed tokens | Lower metal prices would reduce the value of gold exposure. Tokenisation would not remove that exposure. | Actual backing, fees and custody still matter; token holders retain issuer and redemption risks. |
| Gold futures and options | New information about future supply could change prices and volatility before deliveries begin. | Credibility, timing, positioning and leverage; no immediate trade direction follows from the science alone. |
| Miners, developers and royalty businesses | Lower prices could compress margins, reduce project values and delay investment. | Costs, debt, hedging, mine life and contract terms determine vulnerability. |
| Gold-producing countries and communities | Export receipts, tax revenue and mining employment could weaken. | Exposure varies with economic diversification, ownership and policy responses. |
| Central banks and reserve managers | Gold holdings could lose market value, prompting reassessment of reserve composition. | Policy objectives differ; a valuation loss does not mechanically force sales or cause a currency crisis. |
| Jewellers, refiners and industrial buyers | Cheaper inputs could expand demand and applications, while existing inventories lose value. | Certification, consumer acceptance, inventory turnover and competitive pricing. |
| Fusion and isotope suppliers | Demand could grow for engineering, separation, materials and processing services. | Technical execution and sustainable margins; gold-related publicity does not guarantee profitable businesses. |
Mining equities can amplify a metal-price move. Consider a purely illustrative operation selling gold for $4,000 per ounce against $2,500 in relevant costs. Its $1,500 operating spread falls to $500 if the gold price drops to $3,000 and costs stay unchanged. That is a 25% metal-price decline and a 67% spread decline. It is arithmetic, not a forecast for any company; actual earnings and share prices involve additional factors.
Other assets would not have guaranteed outcomes. Bitcoin might attract renewed scarcity comparisons, but a gold-supply breakthrough would not establish a reliable rotation into crypto. Silver and platinum-group metals have different production pathways and demand balances; a gold result cannot be generalised into cheap production of every precious metal. Nor would cheaper gold automatically create broad consumer-price inflation or make government bonds rise: those markets have additional drivers.
What would change the assessment?
Five developments would make this a stronger market story:
- Independent experimental validation: published production rates, recovered quantities and isotope composition.
- Auditable unit economics: the full cost of a saleable ounce, including capital, financing and delays.
- Product acceptance: a demonstrated route through relevant safety, refining and wholesale-market requirements.
- Repeatable operating capacity: financed facilities that sustain output beyond a one-off experiment.
- Actual deliveries: annual volumes large enough to affect market balances, alongside evidence of buyers accepting the metal.
Gold’s supply constraint is not an immutable law of nature. Nuclear physics allows transmutation. Economics determines whether it becomes a meaningful source of bullion.
Diamonds show that manufactured abundance can redistribute value from scarcity owners towards consumers and industrial users. For gold, that outcome remains conditional on a much more demanding chain of proof. The issue deserves monitoring—and better evidence than a headline announcing that alchemy has arrived.
Research method and limitations. Information reviewed through 23 October 2026. Scientific findings, company statements, simulations and APEX’s conditional analysis are identified separately. De Beers and the World Gold Council are industry sources; their figures are attributed rather than treated as independent causal studies. No audited commercial gold-production cost, independently replicated market-volume output or reliable gold-price elasticity was established from the sources reviewed. This article is educational research, not a personalised investment recommendation.
Sources
- GIA: Is There a Difference Between Natural and Laboratory-Grown Diamonds?, 31 May 2019. Material properties and identification.
- CERN: ALICE detects the conversion of lead into gold at the LHC, 8 May 2025. Experimental quantities and fate of the nuclei.
- Rutkowski, Harter and Parisi: Scalable Chrysopoeia via (n, 2n) Reactions Driven by Deuterium-Tritium Fusion Neutrons, preprint, revised 3 August 2025. Yield model; PDF, section 4.1.2 for product activity. Modelled results are not commercial validation.
- Marathon Fusion: Company announcement on X, reported 21 October 2026. Direct X retrieval was unavailable; statement inspected through a thread mirror, with announcement timing corroborated by Business Today, 22 October 2026. Cited as a company claim, not independent scientific verification.
- Edahn Golan Diamond Research & Data: Was the Bottom Reached? Lab-Grown Wholesale Prices Rise in Q3 2026, 7 October 2026. Published wholesale index, category changes and methodology.
- De Beers: Intention to close Lightbox, 8 May 2025.
- De Beers: Preliminary financial results for 2025, 20 February 2026. Rough-price index, earnings and market conditions.
- De Beers: Interim financial results for 2026, 30 July 2026. Segment differences and earnings.
- American Chemical Society: Production of Aluminum: The Hall–Héroult Process. Commercialisation and historical ingot prices.
- GIA: An Introduction to Synthetic Gem Materials. Synthetic crystals, relative rarity and industrial applications.
- US Geological Survey: Mineral Commodity Summaries 2026: Industrial Diamond. Industrial synthetic share; retrieved through indexed text when direct PDF access was unavailable.
- World Gold Council: Above-ground stock, 18 August 2026, covering end-Q2 2026. Stock and central-bank estimates.
- LBMA: London Good Delivery—Gold and Silver. Current wholesale framework, not a ruling on transmuted gold.
- World Gold Council: Gold Demand Trends: Full Year 2025—Supply, 29 January 2026. Fixed historical baseline for scenario arithmetic.