As a consequence, silver might be in the process of making its last gasp pop up.
Minor league stuff compared to the politicization of the Fed and the inane continued allegiance to pushing a political agenda by the bumbling bureaucrats still remaining who are spoiled leftovers from previous administrations. The continued tariff tariff tariff cries sound as ridiculous as Marcia Marcia Marcia on The Brady Bunch.
I will probably wait to see some evidence of either an AI or silver boom bust move back down before getting short rather than trying to pick a top and one way of doing it might be to get short or play an ETF in either or both. This is an interesting assessment of the correlation between the two, from Grok:
"Understanding the Correlation Between Silver Prices and the AI Industry
The "booming AI industry" refers to the rapid expansion of artificial intelligence technologies, driven by advancements in semiconductors, data centers, and high-performance computing. This growth has a tangible, structural link to silver prices through industrial demand. Silver, the most electrically conductive metal, is essential in AI hardware—particularly in semiconductors (e.g., GPUs and TPUs for internal connections and packaging), sensors, connectors, and high-bandwidth memory (HBM) systems. Additionally, AI's massive energy needs are accelerating demand for silver-intensive solar panels and renewable energy infrastructure to power data centers.
While silver prices are influenced by multiple factors (e.g., investment demand, monetary policy, and broader commodities like gold), the AI boom has emerged as a key driver since around 2023, coinciding with the explosive growth of AI adoption. Below, I'll break down the nature of this correlation, its strength, historical context, and future implications.
The Mechanism: How AI Drives Silver Demand
Semiconductors and Chips: AI relies on advanced chips where silver enables low-resistance interconnects in nodes below 5nm. A typical AI server cluster requires 2–3 times more silver than traditional data centers due to higher power density, complex cooling, and interconnections. The global semiconductor market, fueled by AI, is projected to reach $688 billion in 2024 (up 13% YoY), with AI applications driving increased silver consumption.
Data Centers: Hundreds of new hyperscale data centers are under construction globally to support AI workloads. Each requires significant silver for networking hardware, MLCCs (multilayer ceramic capacitors), and sensors. Goldman Sachs estimates AI will account for 19% of data center power demand by 2028, equivalent to ~200 TWh annually—necessitating silver-heavy infrastructure.
Renewable Energy Tie-In: Tech giants (e.g., Google, Microsoft) are committing to net-zero goals, boosting solar PV installations. A single 500MW solar array for one data center uses ~300 metric tons of silver in photovoltaic cells. Solar alone consumed 17% of global silver demand in 2024, with projections for 3,500 GW of capacity by 2028.
The Silver Institute reports industrial demand hit 1.2 billion ounces in 2024 (up 18% over four years), with electronics and batteries accounting for ~232 million ounces—AI is a fast-growing subset. Supply constraints (e.g., mining output lagging, reserves potentially depleted 85–98% by 2050 from solar/AI growth) amplify price sensitivity.
Strength of the Correlation
This is primarily a supply-demand correlation rather than a direct financial one (e.g., silver prices tracking AI stocks tick-for-tick). It's moderately strong and positive, especially since 2023, as AI investments translate to physical hardware needs. Key evidence:
Price Impact: Silver has more than doubled in 2025 (from ~$30/oz to over $60/oz by December), far outpacing gold's 60% gain. Analysts attribute 20–30% of this rally to AI-related demand, with forecasts for 62% further upside by 2027 if deficits persist.
Causality Over Coincidence: Unlike speculative correlations, this is structural—AI growth (36.6% CAGR through 2030) directly consumes silver. The Silver Institute and Oxford Economics note AI as a "critical" demand driver in digital applications.
Quantitative Proxy: While exact correlation coefficients aren't widely published, silver's industrial use (59% of demand in 2025) ties it to tech cycles. For context, silver prices spiked 332% during past rate-cut cycles with industrial booms; AI is layering on top.
In contrast, gold's AI link is weaker (more monetary hedge), with electronics demand recovering modestly to 249t in 2023 but plateauing due to substitution.
Historical Context
Pre-2023: Silver's industrial demand shifted from photography (down 90% since 2000) to electronics/solar, but AI wasn't a factor. Prices averaged $20–25/oz from 2015–2022, correlating loosely with broader tech (e.g., 5G, EVs).
2023–2025 Boom: NVIDIA's stock (a proxy for AI health) surged ~800% since early 2023, mirroring silver's takeoff. Silver futures showed tight price correlation with AI chip demand signals, entering a fourth year of supply deficits. ETF inflows hit 95 million ounces in H1 2025, reflecting bets on AI-driven growth.
Longer-Term Trends: Over 20 years, silver's price volatility (beta ~1.5 vs. S&P 500) amplifies during tech expansions, but AI marks a "supercycle" shift, blending digitalization with green energy.
No direct historical data ties silver to "AI market cap" (estimated at $200–300B in 2023, now trillions via proxies like NVIDIA's $4.47T cap), but the parallel rallies since ChatGPT's 2022 launch underscore the link.
Future Outlook and Risks
If AI sustains its trajectory (e.g., $400B in accelerator spending by 2027), silver demand could rise 5–10% annually, pushing prices toward $70–100/oz by 2030. However:
Risks: Economic slowdowns could curb data center builds; substitution (e.g., copper for silver) or recycling might ease deficits. Geopolitical tariffs on critical minerals add volatility.
Opportunities: Silver's gold-to-silver ratio (now 90:1 vs. historical 50:1) suggests undervaluation, positioning it as an "AI trade" with defensive qualities.
In summary, the correlation is strong and causal through industrial channels, with AI acting as a multiplier on silver's ~50% industrial exposure. This isn't hype—it's physics meeting economics. For investors, silver offers leveraged upside to the AI boom but with commodity risks. Track Silver Institute reports and NVIDIA earnings for signals."
As we have discussed in the past, Econ 101 says Powell is 100% correct. Negotiation 101 says he is wrong. What I dislike about the tariff regime is that Trump is trying to say the US is "making so much money" from it. Actually, it's just a redistribution.
Really good breakdown of why the balance sheet expansion matters more than the rate cut itself. The CDS liquidity angle is somethign most coverage totally misses, but thats where the real systemic risk lives if things sieze up. 2008 showed what happens when banks stop trusting each other to trade. Makes sense the market rallyed hard on that news once peopel actually read past the headline.
Here’s the real issue: the macro data and the lived experience have completely diverged. Yes—the headline numbers look good: low unemployment, cooling inflation, steady GDP. But families don’t feel any of that because the prices never came back down.
Inflation slowed, but prices stayed 20–25% higher. Groceries didn’t get cheaper. Housing is the most unaffordable it’s been in decades. Insurance premiums, childcare, healthcare—everything essential is rising faster than paychecks. Even with a strong labor market, people feel squeezed.
Economists look at trends; households look at receipts.
So you can point to monetary policy, Fed decisions, or balance sheet mechanics all day. But none of that changes the basic truth: most Americans are experiencing a permanently more expensive cost structure while wage growth is slowing. That’s why the economy “doesn’t feel good,” even if the charts say it’s healthy.
Until affordability improves, the data and the day-to-day reality will continue to tell two very different stories.
Read the link I put in two comments and you might have a different view of it. The problem is the news media uses accounting to transmit economic information. It's a long, and sometimes difficult piece to get through-especially if you don't understand basic economics, but it's worth a read. If I translate your comment; relative prices of some goods are higher, even though salaries are higher. Hence, people are still spending a lot relative to the utility they want.
When I said people feel squeezed, I didn’t mean it in a technical “utility” sense. I was speaking in practical cost-of-living terms: essentials cost more, wages haven’t fully kept up, and even with cooling inflation, day-to-day life still feels more expensive. My point was about affordability, not Hicksian utility.
FWIW - Hicksian utility is an interesting theoretical tool, but it’s not how actual people experience rising prices. Real consumers don’t hold utility constant. They feel the affordability squeeze directly. The behavioral reality matters more than the theoretical framing.
I did read the long article you shared, and I actually think it reinforces my point. It makes the case that our price indexes are imperfect approximations of the true cost of maintaining the same standard of living - especially when relative prices shift and when key costs (like housing, insurance, or future consumption) aren’t fully captured. That’s exactly why there’s such a large gap right now between what the data says and what households feel. The measurement tools say one thing; lived experience says another.
So I wasn’t making a technical argument - just noting that for most people, the “price level” they experience in their budget hasn’t really come down.
You are 100% right on. Inflation is caused by the Government spending. Pumping money into non producing recipients and initiatives does not work well for a healthy economy.
I'm not a fan of the Fed buying anything, so I was a little concerned about the announcement that it would buy short-term Ts. Enough QE. Let the markets work in an unfettered manner. Also, what gives with the record high gold and silver prices? What's that pointing to?
got to make sure the CDS market is liquid. that's why they are buying it. My guess is gold/silver will drop a little on the news and when they figure out that there is less risk in the market
https://www.economicforces.xyz/p/price-theory-and-the-price-level?utm_source=post-email-title&publication_id=86578&post_id=181116682&utm_campaign=email-post-title&isFreemail=true&r=ixt4&triedRedirect=true&utm_medium=email Great discussion of price indexes, and inflation here
As a consequence, silver might be in the process of making its last gasp pop up.
Minor league stuff compared to the politicization of the Fed and the inane continued allegiance to pushing a political agenda by the bumbling bureaucrats still remaining who are spoiled leftovers from previous administrations. The continued tariff tariff tariff cries sound as ridiculous as Marcia Marcia Marcia on The Brady Bunch.
The FED should have cut a half point months ago.
I so want to be short silver......
I will probably wait to see some evidence of either an AI or silver boom bust move back down before getting short rather than trying to pick a top and one way of doing it might be to get short or play an ETF in either or both. This is an interesting assessment of the correlation between the two, from Grok:
"Understanding the Correlation Between Silver Prices and the AI Industry
The "booming AI industry" refers to the rapid expansion of artificial intelligence technologies, driven by advancements in semiconductors, data centers, and high-performance computing. This growth has a tangible, structural link to silver prices through industrial demand. Silver, the most electrically conductive metal, is essential in AI hardware—particularly in semiconductors (e.g., GPUs and TPUs for internal connections and packaging), sensors, connectors, and high-bandwidth memory (HBM) systems. Additionally, AI's massive energy needs are accelerating demand for silver-intensive solar panels and renewable energy infrastructure to power data centers.
While silver prices are influenced by multiple factors (e.g., investment demand, monetary policy, and broader commodities like gold), the AI boom has emerged as a key driver since around 2023, coinciding with the explosive growth of AI adoption. Below, I'll break down the nature of this correlation, its strength, historical context, and future implications.
The Mechanism: How AI Drives Silver Demand
Semiconductors and Chips: AI relies on advanced chips where silver enables low-resistance interconnects in nodes below 5nm. A typical AI server cluster requires 2–3 times more silver than traditional data centers due to higher power density, complex cooling, and interconnections. The global semiconductor market, fueled by AI, is projected to reach $688 billion in 2024 (up 13% YoY), with AI applications driving increased silver consumption.
Data Centers: Hundreds of new hyperscale data centers are under construction globally to support AI workloads. Each requires significant silver for networking hardware, MLCCs (multilayer ceramic capacitors), and sensors. Goldman Sachs estimates AI will account for 19% of data center power demand by 2028, equivalent to ~200 TWh annually—necessitating silver-heavy infrastructure.
Renewable Energy Tie-In: Tech giants (e.g., Google, Microsoft) are committing to net-zero goals, boosting solar PV installations. A single 500MW solar array for one data center uses ~300 metric tons of silver in photovoltaic cells. Solar alone consumed 17% of global silver demand in 2024, with projections for 3,500 GW of capacity by 2028.
The Silver Institute reports industrial demand hit 1.2 billion ounces in 2024 (up 18% over four years), with electronics and batteries accounting for ~232 million ounces—AI is a fast-growing subset. Supply constraints (e.g., mining output lagging, reserves potentially depleted 85–98% by 2050 from solar/AI growth) amplify price sensitivity.
Strength of the Correlation
This is primarily a supply-demand correlation rather than a direct financial one (e.g., silver prices tracking AI stocks tick-for-tick). It's moderately strong and positive, especially since 2023, as AI investments translate to physical hardware needs. Key evidence:
Price Impact: Silver has more than doubled in 2025 (from ~$30/oz to over $60/oz by December), far outpacing gold's 60% gain. Analysts attribute 20–30% of this rally to AI-related demand, with forecasts for 62% further upside by 2027 if deficits persist.
Causality Over Coincidence: Unlike speculative correlations, this is structural—AI growth (36.6% CAGR through 2030) directly consumes silver. The Silver Institute and Oxford Economics note AI as a "critical" demand driver in digital applications.
Quantitative Proxy: While exact correlation coefficients aren't widely published, silver's industrial use (59% of demand in 2025) ties it to tech cycles. For context, silver prices spiked 332% during past rate-cut cycles with industrial booms; AI is layering on top.
In contrast, gold's AI link is weaker (more monetary hedge), with electronics demand recovering modestly to 249t in 2023 but plateauing due to substitution.
Historical Context
Pre-2023: Silver's industrial demand shifted from photography (down 90% since 2000) to electronics/solar, but AI wasn't a factor. Prices averaged $20–25/oz from 2015–2022, correlating loosely with broader tech (e.g., 5G, EVs).
2023–2025 Boom: NVIDIA's stock (a proxy for AI health) surged ~800% since early 2023, mirroring silver's takeoff. Silver futures showed tight price correlation with AI chip demand signals, entering a fourth year of supply deficits. ETF inflows hit 95 million ounces in H1 2025, reflecting bets on AI-driven growth.
Longer-Term Trends: Over 20 years, silver's price volatility (beta ~1.5 vs. S&P 500) amplifies during tech expansions, but AI marks a "supercycle" shift, blending digitalization with green energy.
No direct historical data ties silver to "AI market cap" (estimated at $200–300B in 2023, now trillions via proxies like NVIDIA's $4.47T cap), but the parallel rallies since ChatGPT's 2022 launch underscore the link.
Future Outlook and Risks
If AI sustains its trajectory (e.g., $400B in accelerator spending by 2027), silver demand could rise 5–10% annually, pushing prices toward $70–100/oz by 2030. However:
Risks: Economic slowdowns could curb data center builds; substitution (e.g., copper for silver) or recycling might ease deficits. Geopolitical tariffs on critical minerals add volatility.
Opportunities: Silver's gold-to-silver ratio (now 90:1 vs. historical 50:1) suggests undervaluation, positioning it as an "AI trade" with defensive qualities.
In summary, the correlation is strong and causal through industrial channels, with AI acting as a multiplier on silver's ~50% industrial exposure. This isn't hype—it's physics meeting economics. For investors, silver offers leveraged upside to the AI boom but with commodity risks. Track Silver Institute reports and NVIDIA earnings for signals."
Always a great take on markets, thanks.
Powell cracked me up today, bashing tariffs over-and-over (seemed over played, but whatever).
Trump's response on cue, calling him a real dummy.
It will be interesting when the SC chimes in.
As we have discussed in the past, Econ 101 says Powell is 100% correct. Negotiation 101 says he is wrong. What I dislike about the tariff regime is that Trump is trying to say the US is "making so much money" from it. Actually, it's just a redistribution.
Really good breakdown of why the balance sheet expansion matters more than the rate cut itself. The CDS liquidity angle is somethign most coverage totally misses, but thats where the real systemic risk lives if things sieze up. 2008 showed what happens when banks stop trusting each other to trade. Makes sense the market rallyed hard on that news once peopel actually read past the headline.
Keynesian... This isn't exactly that, but the same context. Health care from 100 years ago and how they f'ed it up for us(markets).
https://www.zerohedge.com/political/obamacare-was-not-failure
Here’s the real issue: the macro data and the lived experience have completely diverged. Yes—the headline numbers look good: low unemployment, cooling inflation, steady GDP. But families don’t feel any of that because the prices never came back down.
Inflation slowed, but prices stayed 20–25% higher. Groceries didn’t get cheaper. Housing is the most unaffordable it’s been in decades. Insurance premiums, childcare, healthcare—everything essential is rising faster than paychecks. Even with a strong labor market, people feel squeezed.
Economists look at trends; households look at receipts.
So you can point to monetary policy, Fed decisions, or balance sheet mechanics all day. But none of that changes the basic truth: most Americans are experiencing a permanently more expensive cost structure while wage growth is slowing. That’s why the economy “doesn’t feel good,” even if the charts say it’s healthy.
Until affordability improves, the data and the day-to-day reality will continue to tell two very different stories.
Read the link I put in two comments and you might have a different view of it. The problem is the news media uses accounting to transmit economic information. It's a long, and sometimes difficult piece to get through-especially if you don't understand basic economics, but it's worth a read. If I translate your comment; relative prices of some goods are higher, even though salaries are higher. Hence, people are still spending a lot relative to the utility they want.
TLDR:
When I said people feel squeezed, I didn’t mean it in a technical “utility” sense. I was speaking in practical cost-of-living terms: essentials cost more, wages haven’t fully kept up, and even with cooling inflation, day-to-day life still feels more expensive. My point was about affordability, not Hicksian utility.
FWIW - Hicksian utility is an interesting theoretical tool, but it’s not how actual people experience rising prices. Real consumers don’t hold utility constant. They feel the affordability squeeze directly. The behavioral reality matters more than the theoretical framing.
I did read the long article you shared, and I actually think it reinforces my point. It makes the case that our price indexes are imperfect approximations of the true cost of maintaining the same standard of living - especially when relative prices shift and when key costs (like housing, insurance, or future consumption) aren’t fully captured. That’s exactly why there’s such a large gap right now between what the data says and what households feel. The measurement tools say one thing; lived experience says another.
So I wasn’t making a technical argument - just noting that for most people, the “price level” they experience in their budget hasn’t really come down.
You are 100% right on. Inflation is caused by the Government spending. Pumping money into non producing recipients and initiatives does not work well for a healthy economy.
As in the inaptly titled “Inflation Reduction Act” aka Green New Deal. Wrong headed in every degree.
https://www.economicforces.xyz/p/price-theory-and-the-price-level?utm_source=post-email-title&publication_id=86578&post_id=181116682&utm_campaign=email-post-title&isFreemail=true&r=ixt4&triedRedirect=true&utm_medium=email
I'm not a fan of the Fed buying anything, so I was a little concerned about the announcement that it would buy short-term Ts. Enough QE. Let the markets work in an unfettered manner. Also, what gives with the record high gold and silver prices? What's that pointing to?
got to make sure the CDS market is liquid. that's why they are buying it. My guess is gold/silver will drop a little on the news and when they figure out that there is less risk in the market
Agreed. I had seen some concerning overnights in the last several months.
I think of it as lubricant…too much and the gears can’t mesh, too little and they grind to a halt.
Jeff, above, did you mean Krugman and Stigler or Krugman and Stiglitz?
Stiglitz, will correct...thank you