The gold bull case used to be a list. War. Debt. A weak dollar. Fear. A chart that only went up. On October 7, 2026, a note from The Market Ear cut the list down to one line. Central banks keep buying. Maybe they do. The dollar is rising. Gold is breaking down. The old tailwinds are fading. The question that is left is not whether gold is a hedge. The question is who buys the next ounce.
That is the whole idea, and it is enough. When a bull case has one buyer, the price is no longer a story about the world. It is a story about whether that buyer stays in the room. Goldman Sachs still has a path that rises about 23 percent by the end of 2027, and the bank says nearly all of that rise comes from official buying. The Market Ear's point is the cruel one. A flow that powerful is also a flow the forecast cannot spare. If the buying slows, there is no second plot.
This is not a call to sell gold, and it is not a call to buy a dip. It is a test of what you think you own.
The case that got simple
Simple is not the same as strong. A simple case is a case with fewer ways to be right. For two years, an owner of gold could be right for many reasons at once. Central banks were buying. Private investors were buying. The dollar was not always in the way. Stress in some other market could be used, after the fact, as the reason the metal went up. That stack let people disagree and still hold the same trade.
The stack is thinner now. The Market Ear put the price under its long trend line and well under the 200-day average. It showed the dollar and gold, which had moved together in opposite directions since last autumn, with a wide gap opened by the latest dollar strength. The dollar, on that chart, says gold should take another leg lower. A separate chart marked a bout of stress in France. Yields jumped. Gold did not. The old line, buy gold when things get messy, failed the week it was supposed to be obvious.
None of those pictures is a forecast by itself. A trend line is a line someone drew. A moving average is a memory of the last 200 closes. A gap versus the dollar can close because the dollar falls, not because gold does. France can be a local bond story with nothing to say about a global reserve asset. Take them together, though, and they do one job. They take the easy excuses off the table. If the dollar is a headwind, and chaos is not a bid, and the chart is already under the long average, the bull case has to name its buyer. The note names one. Central banks.
One flow, almost the whole path
Goldman has said gold can still be higher at the end of 2027. In September the bank's end-2027 target was $5,400 an ounce. From the price then, that was about 23 percent. The bank was explicit about the source. Official buying, running far above the old pace, was credited with nearly all of that gain. Private funds and traders were the thin slices. The picture The Market Ear ran on October 7 shows the same shape. A rising blue block of central-bank purchases. A sliver for gold funds. A smaller sliver for speculative bets. The path is not a choir. It is a solo.
The size of the solo is the part that should bother you. Before 2022, Goldman put average official buying near 17 tonnes a month. The bank's nowcast, on a three-month seasonal basis, has been near 91 tonnes a month. That is more than five times the old pace. The Market Ear says Goldman now assumes the pace averages 60 tonnes a month through 2026 and 2027. In a September note, as Kitco reported it, the working figures were 50 tonnes a month in 2026 and 40 in 2027. The exact assumed tonne is a moving target. The structure is not. The base case is no longer the old world of 17 tonnes. The base case is a pace that, a few years ago, would have been the surprise.
That is the trap in the phrase "the new normal." A catalyst is something you did not count on. An assumption is something the model needs in order to print the target. Central banks have been the catalyst. Goldman's path turns them into the assumption. The Market Ear said it cleanly. At some point, extraordinary demand becomes the assumption rather than the catalyst. You can believe the banks will keep buying and still see the risk. The forecast does not have a second engine. Nearly all of the upside is one flow. If the flow is merely large, instead of five times the old normal, the 23 percent is not a view about gold. It is a view about a customer.
You cannot see the customer
The customer is also hard to count. Goldman's nowcast is not the same thing as the line a central bank prints in its reserve report. The bank has said China, in particular, appears to buy more metal than the official tally shows. In July the nowcast put total official buying near 44 tonnes, against that 17-tonne pre-2022 average, with China a large part of it. Goldman thought the real Chinese purchase that month was far above the public number. The three-month pace near 91 tonnes is a model of metal moving, not a press release.
For an investor, that is not a comfort. It is a blindfold on the only bid you have left. You are asked to underwrite a buyer who does not trade on a screen you can watch, who can pause without a headline, and whose true size is a nowcast rather than a fact published on a schedule. The people who dislike central banks are not wrong to dislike them. They are wrong if they think dislike is an analysis. The bull case now asks you to depend on the institution you trust least. You do not have to like the bank. You do have to decide whether you are willing to own a price that needs the bank to keep doing something rare, quietly, for another year and a half.
There is a second discomfort. A buyer this large is not a fan of gold. A central bank buys for its reserve, its politics, and its fear of someone else's currency. It does not buy because your chart looks oversold. It does not owe you a higher price. It can slow down because it has enough, or because the price is already high, or because the dollar is the asset it wants that month. The structural story calls this diversification. The price story calls it a bid. A bid can stop. Diversification is just the polite name for the bid while it lasts.
The dollar is not helping, and chaos is not either
Since last autumn, The Market Ear says, gold and the dollar have moved in a tight inverse pair. That is the old macro link, and it has not been repealed. The latest rise in the dollar has opened a gap. Gold has not fallen as far as the dollar's move, on that chart, would imply. Gaps like that close. They can close with a weaker dollar, which would spare the metal. The note's reading is the other close. The dollar says gold should take another leg down.
You do not have to obey a gap. You do have to notice what the gap replaces. If gold were being lifted by a private panic, a strong dollar would be a detail. If gold is being held up by a slow official bid, a strong dollar is the wind on the only sail. This week's tape fits the second picture. Yields are high. The market has been pricing another hike. Gold has been near $4,100, far under the January high, and leaning on long-term lines instead of making new ones. The official bid can be real and still lose a week to the dollar. A bid that loses weeks can still win years. It cannot be asked to win every week and also to justify a target that has no other source.
The France chart makes the same point from the other side. Stress showed up. Gold did not catch a bid from it. The simplistic hedge failed in public. That does not mean gold never hedges anything. It means you cannot use the word hedge as a substitute for a buyer. In a flow market, gold hedges what the marginal buyer is afraid of. Right now the marginal buyer, in Goldman's own split, is a reserve manager. Reserve managers were not buying French political noise this week. They were, if the nowcast is right, buying a slow stack of tonnes. Those are different trades. An owner who wanted a panic hedge did not get one. An owner who wanted a central-bank hedge is long a program, not a headline.
The options market has already started to doubt the path
The bull story, The Market Ear says, is still intact. The confidence around the path is not. Two option facts carry that distinction.
First, call positioning. Goldman has said gold call open interest is still around three times the old average. The bank's own point is that this can feed on itself. Dealers who sell calls buy the metal to hedge. The hedge lifts the price. The higher price brings more calls. Central banks buy, calls stay large, dealers hedge, and the price rises. That is reflexivity. It is a fine description of a rising market. It is also a description of a market that can fall by the same machine. If the official bid cools, the calls do not have to go to zero for the hedge to flip. Dealers who are long because of the calls they sold become sellers when the calls lose value. The flow that amplified the rise amplifies the decline. Reflexivity is not a friend. It is a gear. Gears work in both directions.
Second, the skew. The Market Ear shows gold's option skew moving sharply higher. Investors are paying up for downside protection. They have not abandoned the story. They are spending money to be wrong less painfully if the path breaks. That is what a one-buyer market looks like when the buyers of the story get nervous. They do not need a new theory. They need a put. A market that is long the dream and long the insurance is a market that knows the dream has a single point of failure.
You can respect both facts without turning them into a trade. Elevated calls are not a sell signal. Rising skew is not a sell signal. They are a description of a crowd that still owns the upside and has started to pay for the floor. The floor they are arguing about is no longer abstract.
The line both sides can say out loud
The Market Ear's last point is the one that should change how you read the rest. The bulls and the bears may agree on the next level that matters. In the note's telling, Goldman's hawkish scenario takes gold toward $4,070. UBS says pullbacks toward $4,000 should be used to add. Two houses, two verbs, one number. One house calls $4,070 a bad case. The other calls $4,000 a place to buy. The Market Ear calls $4,000 the line in the sand.
Look at where the metal already is. Through this decline, spot has been trading near $4,100, with prints on either side of that as the day moved. A hawkish case of $4,070 is not a story about next year. It is a story about a percent or two. The add-zone at $4,000 is not a distant sale. It is the next round number under a market that is already under its 200-day average. You are not waiting to see Goldman's stress case. You are standing next to it. The argument has compressed into a band that a single bad week can cross.
That is why "buy the dip at $4,000" is not an answer to the note. It is one side of the agreement, using the same line as a comfort. A line both sides can see is a line that will be tested, not a line that has been resolved. If gold closes under the long trend and $4,000 comes into play, as the chart note says, the test is not technical. The test is the buyer. Does the official bid show up in size near that number, or was the five-times pace a feature of a higher price? You will not get a press release the same afternoon. You will get the nowcast later, and the price now.
What you actually have to underwrite
An investor does not need a new target. An investor needs a sentence he can say without borrowing it from a bank. Here is the sentence the October 7 note forces.
I am long gold only if I am willing to own a market whose upside, on Goldman's own split, is official buying that has to stay far above 17 tonnes a month. I am not long gold because France is a mess, or because the chart once looked like a runaway, or because a dip toward $4,000 was called a gift. If I cannot live with a pace that slows toward the old world, I do not own the 23 percent. I own a metal that is already below its long average, with the dollar pointing the wrong way, and with options priced as if someone expects the path to fail.
That sentence does not tell you to sell. It tells you what would make the hold honest. A hold is honest if you have decided that 40, 50, or 60 tonnes a month is a reasonable thing for reserve managers to do, even when the dollar is strong and the price is no longer making highs. A hold is a story you inherited if the reason is that central banks "always" buy. They do not always buy at five times the old pace. They have been buying at that pace. The model now needs them to keep it up. Those are different facts.
The shares are the same bet with a gear. Gold miners do not have a separate buyer. They have costs, and they have a price of gold that is set by the same flow. A hundred dollars on the ounce, from $4,100 toward $4,000, does not close a large mine. It does change the story the shares were bought for. The shares fell harder than the metal on the way down because they are a claim on the path, not a claim on the tonnes in a vault. If the path is one buyer, the shares are a leveraged claim on that buyer. You do not get to own the miners as a way around the question. You get the question in a louder voice.
Funds of the metal are the quieter voice of the same question. They are not a central bank. They are the sliver on Goldman's chart, not the blue block. If private buying is the small part of the forecast, a fund inflow will not replace a slowdown in official tonnes. It can change a week. It cannot replace the assumption.
What would prove the note wrong
A one-buyer case can fail in the way that makes the bulls right. The official pace can stay near the nowcast, not merely near the assumption. The dollar gap can close because the dollar falls. Private buyers can return in size, so the blue block stops being the whole picture. Gold can hold the long trend line, and $4,000 can stay a number people talk about instead of a number they trade through. Any of those would mean the simple case was early, not false.
The case can also fail in the way the note fears. The nowcast can slip from 91 toward 60, and from 60 toward something that looks like the old 17, and the target can stay printed on a page while the price goes to the line both sides already named. The calls can unwind. The skew can be the smart money, not a hedge bought out of habit. A close under the trend line would not be magic. It would be the chart agreeing with the dollar that the single buyer did not show up that week.
You do not have to guess which failure you get. You have to know which one you are positioned for. If you need the 23 percent, you need the tonnes. If you need gold to hedge a messy headline, this month already told you that hedge did not arrive. If you need $4,000 to be a sale that other people give you, remember that the other people include a bank that thinks $4,070 is what the hawkish path looks like. You would be buying their stress case and calling it a discount.
The close
Gold bulls have one big problem, and it is not a mystery. The problem is that the bull case has become one buyer. Central banks are that buyer. Goldman says they are almost the entire reason to expect a higher price by the end of 2027. The buying has been more than five times the old monthly pace. The forecast now treats a still-huge pace as normal. The dollar says lower. A messy tape in France did not bring in a new bid. The calls can magnify either direction. The options market is paying more for downside. And the level the bulls would buy, and the level the hawkish case would visit, are almost the same number, sitting just under a price the metal is already testing.
Who is the next buyer? If you do not have an answer that is not "the same central banks, at a pace the model cannot lose," you do not have a bull case. You have a customer. Customers leave. The line in the sand is $4,000 only because that is where you find out if this one stayed.
A note on sources and limits
The framing, the charts, and the $4,000 line are from The Market Ear, "Gold Bulls Have One Big Problem," October 7, 2026. The note says gold is well under its 200-day average and flirting with a long trend line, that a close under that line brings $4,000 into play, that gold and the dollar have moved inversely since last autumn with a gap now open, and that a bout of French stress did not lift gold. It says Goldman expects gold to rise 23 percent through the end of 2027, with nearly all of that rise from central-bank buying. It puts the nowcast at 91 tonnes a month on a three-month seasonal basis, against a pre-2022 average of 17 tonnes, and says Goldman assumes 60 tonnes a month through 2026 and 2027. It says call positioning is about three times the historical average, that skew is rising as investors pay for downside, that Goldman's hawkish scenario is toward $4,070, and that UBS says pullbacks toward $4,000 should be used to add. Those are the note's readings of Goldman, UBS, and LSEG charts. They are not this article's forecasts.
Goldman's September work, as reported by Kitco on September 23, 2026, assumed average official purchases of 50 tonnes a month in 2026 and 40 in 2027, with a year-end 2026 fair value of $4,900 that other reports said was later trimmed toward $4,650, and an end-2027 target of $5,400. The same coverage put the three-month nowcast near 91 tonnes a month and the pre-2022 average near 17. July's nowcast was reported near 44 tonnes, with China buying more than the official disclosure on Goldman's estimate. The 60-tonne assumption in The Market Ear and the 50-and-40 split in the September reports are both attributed, not blended into a fake single number. The "about 23 percent" figure was measured from the price when that target was discussed. From a lower spot price, the same target is a larger percentage. It is still, on Goldman's split, a central-bank percentage.
Spot near $4,100 on October 7 is from the day's trading range, which moved. It is used only to show that $4,070 is close to the market, not as a precise last sale. No moving-average level is invented here beyond The Market Ear's statement that price is well under the 200-day. This is not investment advice and not a solicitation to buy or sell any security or commodity. Gold can fall through $4,000 or leave the line behind. Central-bank tonnes can stay high or drop. Readers should read the primary notes and speak with a licensed adviser before any decision.