Saving in Gold Versus Speculating in It
There is a fundamental distinction between saving in gold and trading or speculating in it, and that distinction shapes how a person should feel about price movements. I am a saver in gold, which means I am fairly price insensitive. A speculator — someone who is leveraged long on futures contracts, for example — might feel very differently about a pullback, but a saver welcomes lower prices.
For that reason, the recent pullback in metals prices is heaven-sent. There is no price anywhere close to the current price that would make me a seller. The only price action I am interested in is lower, because I would like to own more. This stands the usual public reaction on its head. The ordinary public watched the run-up in spot prices and thought it was fabulous; now that there is a pullback, they become nervous and start asking whether gold is even investable. The honest answer is that you have to understand what you are actually looking at.
Consider the logic of the case. If the U.S. dollar loses 75% of its purchasing power while gold maintains its purchasing power, there is no reason that people who can afford to do so would not want to be buyers. In that sense, savers are actually served by lower prices, not harmed by them.
Spot Price Is a Reference, Not the Asset Itself
A great deal of confusion comes from treating the spot price as though it were the true value of the metal. It is not. Spot — and particularly the London reference spot — is arbitrary. Most investors cannot even buy at spot anyway. They typically buy physical gold at spot plus five or spot plus five and a half, and when they sell back to the dealer they receive spot minus three or spot minus three and a half. There are products through which the spreads can be narrowed — certificated products such as a gold trust where one can buy without the wide retail premiums — but for most people purchasing physical gold, spot is a reference price at best.
It is emphatically not a reflection of the true value of either gold or silver in the physical world. Price declines therefore only matter if you believe that the spot quote is the asset, rather than merely the reference mechanism that markets organize themselves around. The deeper point is that most physical buyers never transact at a quoted spot price at all, which means that public sentiment is reacting emotionally to a number that many investors never actually touch. For wealth preservers, the lower price becomes an accumulation window precisely while leveraged participants are being pushed out of the market.
A useful way to frame this is that markets celebrate rising gold right up until they are forced to ask why the paper quote and physical reality keep drifting apart. The institutional pricing conventions quietly depend on investors confusing liquidity with value.
Nominal Versus Real: The Lying Scoreboard
One of the most important disciplines an investor can develop is the habit of looking at outcomes in real terms rather than nominal terms. The stock market makes new highs seemingly every day — a new high, a new high, a new high — but when you adjust those numbers for inflation, they are a whole lot lower. The fastest way to lose purchasing power is to celebrate nominal gains while ignoring what those gains can still actually buy. If a portfolio rises 20% while the cost of living compounds faster, the scoreboard is lying to you.
I do not really care about my holdings in dollar terms. What matters is the maintenance of purchasing power. Quoted performance and real purchasing power are simply not the same measurement, and a recurring institutional advantage comes from keeping investors focused on dollar highs instead of inflation-adjusted outcomes.
This is why I favor long-term strategies over trading. They let you sleep at night. You set a focus, you make sure you know what is going on, and you do not whipsaw yourself reacting to every swing. Long-term positioning matters more than reacting to every price move.
Question raised and answered: Can gold and silver be put into perspective against inflation? Gold in particular has done well over time when people are concerned about maintaining their purchasing power in fiat-denominated instruments. The correlation with silver, however, has historically been less precise. Over roughly fifty years in the precious metals markets, the pattern I have observed is consistent on the gold side and looser on the silver side.
Why Silver Follows Gold
I am not certain exactly why silver runs the way it does, but I have noticed a repeatable sequence. Momentum has to be established first by gold. When the momentum established by gold draws the generalist into the metals market, silver tends to outperform. My suspicion is that this happens because of silver's reputation for volatility and also because of its lower unit price, which makes it more accessible to smaller participants.
There is a recent data point that fits this pattern. About ten days ago, I saw data indicating that physical silver importation into India is at multi-year highs. That tends to suggest that the Indian peasantry is once again concerned about the maintenance of their purchasing power in the rupee — and for good reason. There are probably 300 million people in India who cannot afford to save in gold, so they save in silver instead. I am candid that I am partly speculating about the precise mechanism; I do not know it to be definitively true. But I have observed in prior markets that momentum is established by gold, and when the generalist enters the space — and here I am referring broadly to the Indian peasantry as the generalists — the leadership changes from gold to silver.
History, in other words, often shows that gold opens the door and silver arrives only when attention turns speculative. Momentum in precious metals tends to begin as a purchasing-power defense and only later becomes a broad participation event. The timing gap matters because retail typically enters after institutional conviction has already formed, and that gap determines whether volatility becomes opportunity or regret.
For my own part, I save in gold. I do not invest in gold or speculate in gold — I save in it. Silver occupies a different bucket entirely.
Speculating in Silver: Hated and Unhated Assets
I have, in the past, speculated in silver, and the speculation was built on a specific thesis. For a long time silver was a hated asset class. My belief was that gold would run, and when gold ran, silver would eventually run too and become unhated. That is exactly what occurred. As a result, I sold 80% of the silver in my speculative bucket.
Question raised and answered: Would I start buying silver again? Yes. If at some future conference we notice in the attendee notes that people hate silver again — if it becomes a hated asset once more — I would buy.
This connects to a discipline I practiced as the owner of a brokerage firm. At the end of every trading day I would go into the cage and review the tickets. On any day when buy tickets overwhelmed sell tickets four to one or better, I made myself sell something the next day. On any day when sell tickets overwhelmed buy tickets four to one, I made myself buy something the next day. I am not a trader by temperament at all, but what I learned is that even within a reasonably high-volume retail brokerage firm you accumulate enough data to trade against overbought and oversold conditions.
That is essentially how I speculated when silver entered its melt-up phase earlier this year, in January. What I have learned — and any experienced stock broker learns the same lesson — is that those hyperbolic upward moves, what the Canadians call "hockey stick" charts, resolve themselves unpleasantly for the longs. The same is true in reverse: hyperbolic declines also resolve themselves. That contrarian, extreme-reading approach is the only kind of trading I am able to do. You have to give me something very, very obvious for me to make money as a trader.
The lesson embedded here is that the crowd tends to think momentum confirms safety when it more often signals exhaustion. Hated assets and loved assets create opposite opportunities, because emotion compresses judgment at both ends. The overlooked signal is not price alone but participation intensity — when everyone agrees, the asymmetry disappears. Long-term investors survive by separating savings behavior from speculation rather than blending the two into a single decision.
The Circle of Competence and the AI Capital Boom
Question raised and answered: With hyperscalers issuing huge amounts of equity and debt, where are the trillions of dollars funding them coming from, and what historical period does this resemble? The honest answer is that I do not know. What I have learned to do is stick to what I do. When people ask me whether, say, Nvidia is overpriced, I have to say I don't know. I can price natural resource companies, and I can price conventional financial-services businesses — banks, wealth managers, asset managers, insurance companies — but I do not know how to price technology companies. So I genuinely cannot tell you whether their valuations are reasonable or not, and to be honest, I don't care.
I am reminded of my mentor Peter Cundill, who said there is always something to do somewhere if you stick within your own circle of competence. Warren Buffett has said the same thing. Buffett has no fear of missing out, because over time he has done well doing what he knows how to do. This is the shift from prediction to discipline: refusing to price what falls outside your circle of competence. Institutions often win not because they know more, but because they ignore more. Avoiding forced opinions can protect more capital than chasing the hottest narrative in the market.
That said, I do have suspicions, which I share with the explicit hope that nobody loses money acting on them. The biggest bubbles rarely begin with irrationality; they begin with capital becoming too available.
The Hidden Energy Assumption Beneath AI
My first suspicion concerns the enormous sums being aimed at AI and data centers. If as much capital as appears likely actually gets thrown at AI and data centers, then we will probably find a way to build that capacity more cheaply. The capital-input estimates, I suspect, will come down as we build these things at scale, and we will also find ways to operate them in a far more energy-efficient fashion. We will be pushed toward that efficiency out of necessity, because if all of the current projections come true, we simply do not have that much energy.
This loops directly back to the question of energy security. I am no expert in technology, but my suspicion is that if there is as much genuine utility in AI as people claim, and if as much money gets thrown at it as is being suggested, then that technology probably will solve roughly half of the energy problem on its own.
Interestingly, even that does not derail the broader energy investment case. There are about a billion people on Earth — not AI hyperscalers, but poor people — who have no access to primary electricity at all. I believe we will solve that problem over the next 20 years, which means energy demand is going to take care of itself regardless of what AI does.
Some younger listeners will respond that the new demand will be met by alternative energies — solar, wind, biomass, and the like. To them I would point out a sobering figure: over the last 45 years, humankind has invested about $10 trillion in alternative energy generation, and over that same span we have reduced the market share of fossil fuels from a high of 83% all the way down to only 81%. If the prognostications of the world's big thinkers are even remotely true, we do not have that much energy, and we do not currently have the capacity to produce that much energy.
The broader insight is that everyone talks about demand projections, but almost nobody asks whether the energy actually exists to support them. Scale changes economics, and technology can bend cost curves, but it cannot eliminate physical constraints overnight. The contradiction is that markets price abundance while infrastructure still behaves like scarcity. Investors focused only on software narratives may be badly underestimating the strategic value of energy and materials.
Copper and the Impossibility of the 2050 Timeline
Some of us will make a ton of money trying to meet these energy ambitions — truly a ton of money trying — but the energy-consumption numbers that people are projecting for 2050 simply cannot be achieved. We can't do it. We just can't do it.
To make the point concrete: meeting these targets would require more copper production over the next 15 years than has occurred in all of human history. Given that we have under-invested in the copper business for 30 years, the required volume is impossible within the timelines being discussed. It will be a great deal of fun trying, but we cannot do it.
Copper is one way to play the energy theme, and it is simultaneously a way to play the hyperscalers. The supply math is stark. At Metals Week in London at the end of last year, a paper was published stating that the ten largest copper companies in the world need to invest $250 billion over the next ten years simply to maintain current output. And current output is already in deficit to current consumption — meaning that even fully funding that $250 billion would only sustain an existing deficit, not eliminate it. This is occurring in a market where the lowest estimate of demand growth is 1.5% compounded and the highest estimates run as high as 3.5%. It is also worth stressing that the $250 billion figure was stated in constant 2025 dollars — it is not inflation-adjusted, so the real bill will be higher still.
The conclusion follows inescapably: no matter what we do — with the possible exception of a synchronized global depression — we cannot derail the rationing of these substances by price over time. It is simply too late to correct the supply shortages within the next 5 to 10 years. Shortages will arrive long before the public narrative adjusts to them. Supply constraints become unavoidable when investment cycles lag demand by decades rather than quarters, and markets keep discounting future abundance while current production already struggles to meet existing consumption. The danger for wealth preservation is assuming prices will stay stable while the physical system quietly tightens beneath the surface. Notably, commodity shortages of this kind can emerge even during periods of slowing growth.
Sovereign Debt and Unfunded Promises
When you look at society's ability to fund ever-increasing demands, the capacity is simply not there. We may want to help poor people — and I am all for helping poor people — but increasingly we are going to have to do that privately, because the public balance sheet is overwhelmed.
At the federal level, the United States now owes almost $40 trillion. That alone is a big number. But the deeper problem lies in unfunded promises. The net present value of unfunded entitlement promises — Medicare, Medicaid, Social Security, federal pensions, and military pensions — is roughly $120 trillion. Add that $120 trillion to the $40 trillion of explicit debt and you arrive at something like $160 trillion in total obligations.
Now place that against the asset side. The Internal Revenue Service puts the private net worth of all American citizens at about $172 trillion. That leaves only about $12 trillion of headroom over the combined liabilities. The federal government does hold assets, of course — but those assets are not collateral in any meaningful sense. You cannot imagine a bondholder walking up and saying, "Yes, I'll take Yosemite, please." National parks and the like are not pledgeable security.
The U.S. economy is often said to run on tax receipts, but in practice it runs at least as much on simply printing money — and it is getting easier all the time. Once the system becomes fully digital, the authorities will not even need to use their fingers to do it; they have a far bigger money-creation machine than any private citizen could ever build.
The real risk here is not the debt itself but what policymakers eventually choose to protect from it. Liabilities and asset values can appear manageable right up until confidence in future purchasing power begins to crack. The uncomfortable question is whether fiscal promises will be honored in real terms or quietly inflated away over time. Savers who ignore that distinction may discover that nominal security and real losses are not opposites at all — they can arrive together. That is the thread tying every part of this together: gold, silver, energy, copper, and sovereign solvency are all, ultimately, questions about the future purchasing power of money.