Adrian Day opened with a question, not a price target. Would you lend money to the United States government for 30 years at 5.5 percent? He would not. He does not care what the bond paid in some other decade. He cares whether, with inflation where it is and the deficit where it is, that yield is enough. For him it is not.
That refusal is the backdrop. The fact he kept coming back to, in a long talk with David Lin, is simpler. Gold has held up in a tape that should have hurt it. Higher oil. A strong dollar. Rising yields. A Federal Reserve that has said inflation is the priority and that the tool is a higher rate. Day said that mix should have been very negative for gold. Gold should have collapsed. It has not. The people who have been buying it for the last few years are still buying it on the way down.
That resilience is the story. It is not a breakout. Day runs gold separately managed accounts, and the EuroPacific Gold Fund, at Adrian Day Asset Management. In the accounts, where he can hold cash, he is buying slower than he otherwise would. He is still buying. He is not chasing a metal that has refused to die. This piece is his argument. It is not a recommendation to buy or sell any metal, any share, or any bond.
The long bond is not yet a rival to stocks
Lin asked the October question. Thirty-nine years after Black Monday, when the Dow fell 22 percent in a day, some of the old conditions are back in the conversation. Higher rates. A dollar people argue about. A deficit. A stock market a lot of people call expensive, with the crowd packed into the artificial-intelligence names. Are yields high enough to pull money out of shares and into bonds, and then to knock the shares down in a cascade?
Day said no, not yet, and not at the long end. A transfer like that sits on a slope. One buyer might leave stocks for a 10-year at 5 percent. Another wants 5.5. Another wants 5.75. He does not think the slope has been climbed. The 10-year auction the day they spoke was well bid. He noted that. He also noted that the long auctions are smaller than people remember, because so much of the debt of the last six or seven years has been parked at the short end. Bills are a parking lot. A 10-year, a 20-year, and a 30-year are investments. The parking lot can clear. The investment, at a yield he will not accept for 30 years, has not yet pulled the stock market apart.
He does think another quarter-point increase in the policy rate is likely before the year is out. The first quarter-point, he said, mostly showed up in credit-card rates for people who do not pay the balance. Those rates are already in the 20s. Another quarter-point will not, in his view, be the break for the stock market or for commercial property. It will be one more turn of a screw. Each turn puts more people in trouble. That is why he has said, more than once, that the Fed's room to keep raising is very small.
Three groups pay for each quarter-point
The first is the Treasury itself. Day put the gross issuance near $4 trillion a year, because short paper has to be rolled, sometimes more than once. A 10-year sold ten years ago, when that yield was under 2 percent, comes due into a market that charges about twice as much. The same arithmetic waits for the 20-year and the 30-year. The interest bill gets worse even if the deficit does not grow, and even if the Fed does nothing new. Refinancing is the tax that was delayed.
The second group is commercial property and private credit, where stress has already shown. A lot of that credit, he said, not a majority, floats with the rate. A hike arrives in their interest line in the same quarter. Fixed-rate paper waits. Floating paper does not.
The third group is the lower half of American households by income. Day called net worth a euphemism there, because that half, in his telling, has no positive net worth. He said a majority of households could not put their hands on $1,000 in an emergency. Treat that as his claim, not as a census this piece re-ran. The shape of it is enough. If the paycheck already goes to food, rent, a car, and clothes, there is no slack. The first rate that moves when the Fed hikes is not the mortgage. It is the card. People who cannot clear the balance move closer to the day they cannot clear the minimum.
Lin put a University of Michigan line next to that. Buying conditions for durable goods had improved a little, because some households think prices will be higher later, so they would rather buy the washing machine now. Day's answer was debt, or a store's zero-interest offer for six or twelve months. The machine comes home. The bill comes later. He does not think the same logic is what is bidding stocks or houses. Other forces, he said, matter more there.
The AI loan was competition. He thinks it is fading.
The September Fed minutes, out the day they spoke, listed one reason term premiums and Treasury yields had risen. Private borrowers were raising a lot of debt to build artificial-intelligence infrastructure, and that paper was competing with the government. Day said that had been true. Amazon, Microsoft, Meta, and others have been in the market. He added a warning the coupon can hide. A great deal of the build-out debt sits off the balance sheet. A 9 percent yield with a famous name on the term sheet is not the same object as a bond the company has put on its own books.
He also thinks the wave is late. Skepticism is already in the price. He cited a force majeure at Intel, and growing local opposition to data centers, and a clearer public count of how much has already been borrowed. If he is right, the next year of Treasury supply will not be fighting a fresh flood of AI bonds. It will be fighting whatever else wants the same dollars. Good companies, he said, are already paying 6, 7, or 8 percent. A lender then has a plain choice. Five percent from the government, or more from a business. That choice, not a speech, is what a term premium is.
Oil and the 10-year came apart
Lin showed a chart. Oil and the 10-year had moved together until the yield hit about 5 percent. Then yields kept rising while oil fell into the end of September and into that week. The 10-year, as he read it, sat just under 5.3 percent. He asked if 6 percent next year is imaginable, and whether the bond is saying inflation has moved out of the headline and into the core.
Day would not pretend he had a theory of that divergence ready. Oil, he said, is trading the war with Iran, the posts from the president that swing from one day to the next, how much crude is getting through, and how much is left in storage. Those are near-term facts. They do not have to match the 10-year. Treasury Secretary Scott Bessent had said core inflation gauges were still relatively muted, and that oil had not really seeped into the rest of the basket. Day thinks the bond market is looking past that. He thinks buyers see inflation as stubborn, and he thinks they are right.
A yield of 6 percent, in his view, is mostly a supply question. Short bills can always be sold to people who want a parking spot. The 10-year and beyond cannot. If the Treasury shifts issuance back toward the long end, where it used to live, he expects the long yields to rise to clear it. The department can choose not to. It can stay in bills. The choice is the forecast. The courage to lock in a 30-year loan is the other half, and he has already said he does not have it at 5.5 percent.
The index was adjusted down
Day's inflation view starts with a method, not a slogan. The statistics bureau changed the way it builds the consumer price index again. He said he is not accusing anyone of a plot. The plain result, as he described it, is that this year's index is lower than it would have been under the old method. Once you adjust a car's price because the new car does more than the old one, you have left the receipt. The receipt is what a household pays. He would rather the bureau get it right and still be read with that limit in mind.
Kevin Warsh, at the Fed, has been clear in Day's hearing that inflation is stubborn and that bringing it to 2 percent is the priority. Day's chain is short. If that is the goal, the serious tool is a higher policy rate. A bond buyer who believes the chain buys the higher yield now, before the Fed gets there. That is how a stubborn index and a determined chair show up as a rising long rate, even on a week when oil is falling.
Why the metal did not do what the textbook says
Here is the idea. Real yields have moved up. Oil has been high. The dollar has been strong. The chair has talked about more hikes. Textbooks say gold falls in that weather. Day said it has been remarkably resilient over the last two or three months. He does not expect it to break out toward $5,000 again while yields are backing up like this. He expects a soft market. A range. Volatility. Not a collapse, and not a new high.
The buying he points to is official and steady. China, he said, posted in September what he thought was its largest increase in reported gold buying in about three years. Reported purchases have risen month after month even as the price has come down. He does not know if that continues. He reads it as institutions using the lower price, not as tourists chasing a spike.
The cash in his own accounts changed the pace, not the direction. Two widely held stocks were acquired for cash in June. The proceeds landed in the accounts he runs. He has been able to keep buying and still hold more cash than he normally would. If that cash had not arrived, he said, he would be even more cautious. The gold fund cannot do this. A narrow fund, under the rules, stays invested. The separately managed accounts can wait. Waiting, for him, is not a forecast that gold is finished. It is respect for a yield that has moved ahead of the official inflation rate.
A war is not a bid
Lin asked the question gold owners get wrong. If a geopolitical shock is supposed to lift gold and keep it up, what does this year say, given that the Iran war did not?
Day's answer was that wars which do not come out of a clear sky tend to hurt gold after they start. He used two cases. Russia massed tanks and troops on the Ukraine border. Gold rose hard for weeks. He himself thought they would not invade. They did. Within about a week of the crossing, gold peaked, and then it fell for six or nine months. Iran was the same shape. The bombing was not a secret that nobody had imagined. Gold had already been rising. It peaked, in his telling, the day after the bombs started, and then it gave the gain back.
He offered reasons, and they are mechanical. The dollar is still a refuge. A crisis sends people into it, and a stronger dollar, other things equal, is a weaker gold price. He said that rush was obvious this year. The second reason is liquidity. In the Gulf, he said, gold in Dubai traded at a 25 to 30 percent discount to the world price. If your metal is in a vault in a city that might be hit, you may sell it and move the money to London or New York. Turkey's central bank, worried the war would spread and the lira was falling, sold gold to support the currency. Day noted they had been buying for years, were sitting on a profit, and were probably above their own target. Selling, he said, is not a betrayal of the case for gold. One reason you hold it is so that you can sell it when you need cash. An emergency reserve that cannot be spent is a souvenir.
So a future war, if it looks like these, is not a plan for a higher gold price. The rise comes while the threat is building. The peak is near the first shot. The months after belong to the dollar and to whoever needs to raise cash. People who buy the headline on the day of the headline are late to a trade that already happened.
What would actually restart it
Lin asked for the regime that makes a multi-year bull run, if a war will not hold the price up. Day wanted the chart from 1971 on a log scale, because the scale changes the boast. From 1971 to the 1981 peak, he put the gain at 2,358 percent. From the trough around 2001 to 2011, he put it at 680 percent. The bull market of the last few years, on that scale, is not an extraordinary one. He is confident, and he said he may be wrong, that what we are in now is a correction in the middle of a cycle, not the end of one.
The trigger he named is the end of a war, not the start. An end would, at least for a while, take the war premium out of oil and out of the dollar. Oil down means less pressure on inflation. A dollar that resumes the decline he saw in 2025, not a crash, just the loss of a refuge bid, would ease the gold price from the other side. If those two let the Fed step back from hikes, he thinks gold can turn up again. He is describing a sequence. He is not promising the week.
He anchored it in real rates, which is the part Lin had already named. In the 1970s, Arthur Burns raised rates for years and gold still rose, because the rates stayed behind inflation. Only when the yield gets ahead of inflation, a positive real rate, does the metal feel it. Day thinks that is where we are. Yields have moved meaningfully above the official consumer price index. He also thinks that index is too low. If yields keep rising and the official numbers do not, gold feels it. No speech fixes that. The spread does.
Where he is willing to add
Higher rates have not made him swap seniors for juniors, or the reverse, as a rule. He is looking at one company at a time. The pattern he wants is a bad quarter, a weak reserve update, or some other headline in the last couple of months that cut the shares by more than the facts deserved. Those are the adds.
The example he gave was Franco-Nevada. A week or two before the interview, Panama's ministerial commission published its report on Cobre Panamá, the copper mine owned by First Quantum, where Franco-Nevada holds a stream. The tape of the interview called the owner First Majestic. The company is First Quantum. The report's first recommendation, issued September 30, is a negotiation toward a restart that would fund an orderly closure, at no cost to the state, with no extension of the mine life. Day called that pairing perverse. A document about reopening that leads with closure invited every headline to print the closure. The shares of the miner and of Franco-Nevada fell hard. He added to Franco-Nevada into that fall.
Read the add as a method, not as a tip. He was not arguing the politics of Panama. He was arguing that a strange sentence in a government report can mark a stock down past the change in the asset. The stream is still a stream. The mine is still shut until a deal exists. The gap between those two sentences is where he said he hunts. Big miners, big royalties, small explorers. The size does not sort the list. The exaggeration does.
Sentiment does change the default. Gold shares are weaker than they were earlier in the year, when the metal was near $5,000. If a large producer falls 20 percent, he is more inclined to buy it than if a small explorer falls 20 percent. By that default, a softer tape shifts him toward the larger names. Lin suggested the opposite, that weak sentiment is when the small names are cheapest. Day agreed that is true when retail has given up. He does not think retail has given up. The week before, he had been at two gold conferences in Colorado, both with record crowds and a record number of generalist investors, and at a mining forum in Vancouver with what he called a strong, enthusiastic room. Capitulation is an empty room. These rooms were full.
One seasonal fact sits on top of that. Widely held stocks that are down on the year are likely to be sold for tax losses, because so few shares show a loss at all. A name that has declined all year can be sold for the calendar, not for the ore. He flagged it. He did not say every December sale is a gift. He said the calendar will add selling that has nothing to do with the mine.
The idea, once
Day will not lend to Washington for 30 years at 5.5 percent. He thinks another quarter-point is coming, that it will not break the stock market by itself, and that it will tighten the screw on the Treasury's interest bill, on floating credit, and on households who live on the card. He thinks the AI borrowing that competed with Treasuries is slowing, and that a 6 percent 10-year, if it comes, comes from more long-dated supply, not from a mystery. He thinks inflation is stubborn, the official index is softer than the old method would have shown, and bond buyers are trying to get ahead of a Fed that has named 2 percent as the job.
None of that is a gold target. The fact is that this mix should have broken the metal, and it has not. China has been adding as the price fell. The dollar and the need for cash, not the war itself, explain why gold peaked when the shooting started and then sagged. The next leg up, in his sequence, wants a quieter war, a softer dollar, cheaper oil, and a Fed that can stop. Until yields stop backing up, he expects a range. In that range he buys slower, holds more cash than usual, and adds where a headline did more damage than the asset. A full conference hall is not a clearance sale. A stock cut by a clumsy sentence might be.
Gold should have broken. It did not. That is the information. It is not an instruction to pay up.
A note on sources and limits
This account follows a mid-October 2026 interview on David Lin's program with Adrian Day, president of Adrian Day Asset Management and portfolio manager of the EuroPacific Gold Fund. Yields near 5.3 percent on the 10-year, the well-bid auction, the quarter-point hike he expects, the $4 trillion issuance figure, the sub-2 percent 10-year of a decade ago, the card rates in the 20s, and the $1,000 emergency claim are his or Lin's statements from that conversation. They are not a new official release.
The September FOMC minutes, released October 7, 2026, are the source for the line Day and Lin discussed, that heavy private debt for AI infrastructure had added to term premiums. Off-balance-sheet AI borrowing, the Intel force majeure, and corporate coupons of 6 to 8 percent are Day's characterization. The statistics-bureau method change is his description of the result. It is not a charge of misconduct.
China's September buying, the Dubai discount of 25 to 30 percent, and the Turkish central bank's gold sales are Day's account of those markets. The 1971–1981 gain of 2,358 percent and the 2001–2011 gain of 680 percent are his reading of a log chart. Black Monday, October 19, 1987, and the 22 percent one-day drop in the Dow are the historical event Lin used as the anniversary.
Cobre Panamá is owned by First Quantum Minerals, not First Majestic. The interview misspoke the name. Franco-Nevada holds a stream on the mine. On September 30, 2026, a Panamanian ministerial commission recommended negotiations toward a restart that would finance an orderly closure, with no extension of mine life. Franco-Nevada and First Quantum both disclosed the recommendation that day. Day said he added to Franco-Nevada after the shares fell. This piece does not know his size, his price, or whether a restart will happen.
Nothing here is investment advice or a solicitation to buy or sell any security or metal. A resilient gold price can still fall if real yields keep rising. A headline discount can be deserved. Readers should read the filings and speak with a licensed adviser before any decision.