Gold mining stocks and the gold price are not the same bet. In the three years through late September 2026, the big-miner fund GDX compounded at about 49 percent a year. The bullion fund GLD compounded at about 30 percent. Over one year the gap was about 29 percent against about 15 percent. Over three months it was about 21 percent against about 5 percent. Then the last month reversed the courtesy. GLD fell about 6.6 percent. GDX fell about 9.3 percent. Since 2006, the full life of that comparison, the bar won. GLD compounded near 9 percent a year. GDX compounded near 5.5 percent. The experts who see more upside in the shares are talking about the first set of numbers. The record that includes the bad years is the second. Both are true. A headline that keeps only one of them is a sales pitch.
On October 1, 2026, the spot gold price was still above $4,100. Morning quotes were near $4,159. Later quotes sat around $4,170 to $4,180. September had taken more than 6 percent off the metal. A January 29 closing high of $5,405 still sat about 23 percent above the market. The 10-year Treasury yield was around 5.2 to 5.3 percent. Shares of miners, through GDX, ended September near $88, after trading above $101 early in the month, and the fund’s year-to-date gain was only about 2 percent. The gear that had paid for three years had just taken a bite. That bite is the subject. It is not a footnote.
This piece has one theme. The extra upside in mining stocks is the margin, and the margin only widens if the gold price rises faster than costs. A mine with heavy fixed costs turns a small change in the gold price into a larger change in profit. That is why experts say gold mining stocks offer more upside in a gold rally. It is the same reason they offer more downside in a gold price correction. September was the downside. The three-year record was the upside. The twenty-year record shows the upside is not free, because costs, new shares, and bad mines ate it. More upside is a statement about the next move in the gold price. It is not a trait the shares own by themselves.
Nothing here is advice to buy or sell gold, a miner, or a fund. Leverage cuts both ways. A forecast can be wrong. A low-cost mine can still issue shares until your slice of the profit is small.
Why gold miners could outperform gold
The mechanism is old, and it does not require a guru. Picture a mine whose all-in sustaining costs are $1,800 an ounce. That figure is an illustration, not a promise about any one company. At a gold price of $4,000, the margin is $2,200. If the gold price rises by $400, to $4,400, the price is up 10 percent. The margin is up about 18 percent, from $2,200 to $2,600. Profit, if costs stay put, rises faster than the metal. If the gold price falls by $400, the margin falls by about 18 percent while the metal falls 10 percent. Gold miners vs gold is this ratio. Gold vs gold stocks is the same ratio read from the other side. The shares are a claim on the spread. The bar is a claim on the price.
Costs do not stay put. That is the leak in the illustration. Fuel, labor, steel, royalties, and sustaining capital move. If the same $400 rise in gold arrives with a $200 rise in all-in sustaining costs, the margin rises by only $200, not $400. The gear shrinks. If costs rise as fast as the gold price, the gear is gone, and you took equity risk for the return of a bar, minus the headaches. Gold mining profitability is not the gold price. It is the gold price minus a stack of bills that have their own cycle. AISC gold miners publish that stack for a reason. An investor who will not read it is guessing the only number that decides whether the expert was right.
There is a second lever, and it is not operations. It is the multiple. Gold stock valuation is what the market will pay for a dollar of that margin. In a rising gold bull market, investors pay more for the same margin because they think the margin will keep growing. The share can rise by the profit gain plus the multiple gain. That double lift is why a three-year compound rate near 49 percent was possible while gold itself, through GLD, compounded near 30 percent. In a scare, the multiple shrinks while the margin shrinks. September’s deeper loss in GDX is that double cut. Gold stocks performance is not a smoother version of the gold price. It is a louder one.
A third lever sits in the ground, and it is the one Rick Rule stressed in late September. Years of thin exploration left the industry short of new mines. A new mine in a place with courts and communities can take more than a decade. If the gold price is substantially higher over five years, a producer who needs ounces may buy them rather than wait. Single-asset mines become targets. That bid is not operating leverage. It is a scarcity premium on gold resources that can actually be mined. It can make a share outperform even if its own costs are ordinary. It can also fail to arrive. Logical takeovers are often late, or they arrive in shares that dilute the buyer’s owners. Gold supply does not care that a slide called a project “strategic.”
Why gold mining stocks offer more upside
Experts who say the shares have more upside are usually stacking those three levers and then assuming the gold price goes up. Operating leverage. A richer multiple. A buyer for scarce ounces. If all three work, gold mining stocks vs gold is not a close race. The shares win by a lot. Rule has pointed at the 1970s as history, not as a promise. Gold rose on the order of 25-fold. The old Philadelphia gold and silver index, a cousin of today’s miner funds, rose on the order of 49-fold. He is not forecasting a repeat. He is saying the direction of the leverage, in a long nominal bull market, has been toward the shares. Gold mining stocks vs physical gold, in that history, was a bet that paid more and hurt more. Anyone selling the 49-fold figure as a gold price prediction for this decade is leaving the interview.
The condition is the whole sentence. More upside if the gold price rises faster than costs, and if the multiple does not collapse, and if the company does not issue a river of new shares. HSBC, in early October, cut its average gold price forecast for 2026 to $4,490 and its 2027 average to $4,825. Read the word average. A year that opened above $5,000 can average $4,490 even if autumn is ugly. The bank also said near-term pressure can continue and that the price may be nearing a bottom, with central banks more willing to buy if the price nears $4,000. That is a gold price outlook for the metal. It is not a gold mining stocks outlook. A higher average gold price helps miners only if they are still producing, still solvent, and still owned by the same number of shares when the average is realized.
Gold demand is the other half of the metal’s case, and it does not move the shares one for one. Jewelry buyers step back when the price is high. Investors in coins and bars can panic or stampede. Central banks buy on a slower clock. A producer does not sell to a central bank’s press release. A producer sells into a spot market. If gold demand from officials puts a floor under the price, the floor helps every miner’s margin at once. It does not pick the best mine. Gold investment options that start with the bar and stop there are buying the floor. Gold investment options that go on to the shares are buying the floor plus the gear, the multiple, and the mistakes.
What the twenty-year record is trying to say
If the gear were free, GDX would have beaten GLD since 2006. It did not. About 9 percent a year for the bar, against about 5.5 percent for the miners, over that long window, is the bill for the times the gear ran backward. Costs rose. Projects ran late. Companies issued shares at the bottom to stay alive. Some of the largest names of one cycle were not the largest names of the next. Gold stocks vs gold, over a full cycle that includes a bear market, can be a worse compound because the downside years take back more than the upside years give. A three-year window that starts after a low will hide that. Late September 2026 is inside a strong three-year window and immediately after a bad month. Both windows are on the screen. Choosing the flattering one is not analysis.
Risk is the other line on the same page. Over three years, GLD’s volatility was about 21 percent. GDX’s was about 40 percent. The worst drawdown in that window was about 26 percent for the bar and about 39 percent for the miners. GDX’s beta to GLD was about 1.6. Correlation was high, near 0.83, which means they usually move together, and the miner moves more. Precious metals stocks are not a diversifier against gold. They are gold with a volume knob. Precious metals mining stocks turn the knob past one. If you wanted a quieter hold, you already own the wrong tool. If you wanted a louder hold, you have to admit the volume knob was turned up in September, and it was not pleasant.
Gold stock investment that pretends the knob only turns up is how people buy miners at the end of a gold rally and then feel cheated. The cheat was in the brochure. The fund’s own long-term numbers were public. Gold stocks 2026 sit on top of that history, not instead of it. A good gold stock outlook says the next leg up would likely be larger in the shares than in the metal, and the next leg down would likely be larger too. Anything else is a gold mining stocks outlook 2026 written by someone who needs a buyer.
Costs are the plot
All-in sustaining costs are the industry’s attempt to stop miners from quoting a flattering cash cost and ignoring the capital that keeps the mine alive. Cash cost is the cost of pulling the next ounce. All-in sustaining costs add the capital that replaces what you just mined, plus royalties and the corporate bill that does not sit at the pit. A company can look cheap on a cash cost and expensive on an all-in cost. The all-in number is the one that has to sit under the gold price for the gear to be real.
West Red Lake Gold’s Madsen mine is a useful, public example because the costs are high enough to see. In the second quarter of 2026 the company reported cash costs near $2,000 an ounce and all-in sustaining costs near $3,284, inside a guidance range of about $2,800 to $3,600. Gold near $4,160 still leaves a margin. It is not the margin of a great mine. Guidance for the year was 35,000 to 45,000 ounces. If those costs fall toward the bottom of the range, the gear improves. If they stay at the top, a lot of the gold price is spoken for before a shareholder is paid. AISC gold miners at the high end of the industry are a geared bet that only works if the gold price stays well above that band. They are not “more upside” in the abstract. They are more upside only after a wide spread, and less upside, fast, if the spread closes.
Fuel is the clean leak, and two large producers have published it. Barrick’s 2026 planning case has used about $70 West Texas oil, and its own sensitivity has put roughly $12 an ounce on cash costs for a $10 move in oil, once that move reaches the diesel. Agnico Eagle has said most of the diesel for its Nunavut mines is already bought through the 2027 sealift, about 30 percent above its 2026 budget after hedges. A falling oil price on a screen does not refund a barrel already purchased. Gold mining companies with northern logistics can see the gold price rise and the margin rise by less than the textbook. Canadian gold mining stocks are full of this kind of fact. It is boring. It is the difference between a slogan and a margin.
Producers, developers, explorers
The gear is not the same in every corner of the gold mining sector. Gold producers have a margin today. Gold producer stocks move with that margin and with the multiple. They are the heart of GDX. At the end of September the largest weights included Newmont, Agnico Eagle, and Barrick, with Wheaton and Franco-Nevada, the royalty model, close behind. Those are gold stocks to watch if you want the sector’s beta without picking a single pit. They are not a list of buys. Royalty companies are a different machine. They take a slice of revenue and skip most of the operating cost. They have less gear and fewer operating accidents. In a wild gold rally they often lag the dirtiest producer. In a cost spike they often hurt less. Gold mining stocks to watch, if the phrase means anything, start with this split. Do you want the full gear, or a thinner gear with fewer ways to fail.
Gold developers have the deposit and the bill, and not the margin. Their value is a set of future ounces, discounted at a rate that just went up with bond yields. A 5 percent long bond is a tax on patience. A project that works at $4,100 gold can still be a bad stock if it must raise money every year until first pour. Gold mining stocks outlook language that lumps developers with Newmont is hiding the raise. The raise is the story. If the gold price rises and the share count rises faster, the holder of the old shares can lose while the expert was “right” about gold.
Gold exploration stocks are a step further out. Many have no mine and no study that a bank will lend against. Junior gold stocks are priced as options on a discovery, a gold price, and a financing window. Junior gold mining companies fail often. That is not a slur. It is the base rate. A portfolio of attempts can make sense inside a speculative slice. One name bought because miners “have more upside” is a category error. TSXV gold stocks are where that error is most expensive, because the exchange is built for early stories and the stories do not owe you a mine. Mining stocks to watch at that end of the list are a research project. They are not a substitute for the bar, and they are not a substitute for Agnico.
Canadian mining stocks dominate the respectable middle of this market. Agnico, Barrick, Kinross, Wheaton, and Franco-Nevada are how a global investor owns Canadian gold stocks without meaning to. They bring Canadian assets, Canadian diesel, and Canadian royalties into a U.S. ticker. Gold mining stocks 2026, for many portfolios, are already a Canada bet. That can be a good bet. It is still a bet on costs and on the gold price, not a patriotic fact.
The upside experts skip when they say upside
Rule’s five-year case and the three-year fund record are the bullish file. The file has a cover sheet he also wrote. Rising U.S. rates are a near-term headwind. He has treated bullion as savings and the shares as a speculation that requires work. He has said he is re-valuing what he owns and holding cash, because volatility is when he buys, and you cannot buy a sale with a full hand. A man who is bullish on miners over five years and who wants cash for a drop is not contradicting himself. He is describing the path. Gold mining investment that cannot sit through a 9 percent month in the miners is not the investment those experts described. It is a trade that will fire them at the low.
There is a version of more upside that is smaller and more honest than a new record. Gold does not have to return to $5,405 for the shares to rise. A move from about $4,160 back through $4,200 and on toward $4,300, if costs hold, widens margins and can lift the multiple. With a beta near 1.6, a 5 percent move in the metal can be something like an 8 percent move in the fund, if the relationship holds and nothing else goes wrong. It often does not hold for a single week. It is a tendency, not a contract. Gold stocks vs gold on a given Friday can be anything. Over a year in which the metal trends, the tendency has been real. Over a year in which the metal chops and costs rise, the tendency has been a trap.
Gold production is the constraint that makes the tendency plausible. The world does not spin up a new 300,000-ounce mine because the price had a good month. Gold supply from existing pits can rise a little, through higher grades and extra shifts. It cannot erase a deficit on a headline. If gold demand, including official buying, stays firm while new mines stay late, the price has a reason to hold a higher floor. Miners with real margins are the equity way to own that floor with gear. Miners with costs near the price are the equity way to own a rumor of the floor. The expert comment worth keeping is the one that separates those two.
A valuation that can be checked
Gold stock valuation can be done on a page. Take the all-in sustaining cost. Subtract it from a gold price you are willing to underwrite, not from the high you wish for. What remains is the margin. Ask how many ounces the mine can actually sell, not how many a slide once hoped for. That is revenue minus cost, before tax and before the interest on a build. Then look at the share count this year against the share count three years ago. If the count is up a great deal, your piece of the margin shrank. Then look at net debt. A margin that exists only to pay a lender is not yours yet.
Do this at two gold prices. One near today’s, about $4,160. One lower, near $4,000, which several desks still treat as a possible test if jobs or yields surprise. A company that is fine at $4,160 and strained at $4,000 is not “more upside.” It is a conditional. A company that is fine at both, with a stable share count, is the sort of gold producer the upside thesis needs. You will not find many of the second kind on the most exciting list. You will find them among large gold producers and some royalties, and you will still have to read the fuel note. Canadian gold stocks in the senior group are the practical place to run the page. TSXV gold stocks rarely survive the second price.
The multiple comes last, not first. If the market is already paying a rich price for the margin, the expert’s upside may have been spent. If the market is paying a low price because September scared it, the upside may still be ahead, and so may another scare. Gold mining stocks outlook work that starts with a target and never opens a cost line is a story. Gold price forecast work that never asks what the mine keeps is the same story in a different suit.
When the experts are wrong
They are wrong, for a holder of shares, in four ordinary ways. The gold price falls and the gear magnifies it. Costs rise and eat the gear. The company issues shares and hands the gear to new owners. Or the gold price rises and the shares still lag because the multiple was already high, or because a single asset failed. September was a small version of the first. A diesel bill already signed is a small version of the second. A developer’s annual raise is the third. A pit wall, a permit, or a government that rewrites a contract is the fourth. Panama’s long fight over Cobre Panamá is a reminder that a copper pit can be legal on Monday and a political problem for years. Gold mines are not exempt. Geology does not vote. Governments do.
They can also be right about the sector and wrong for you, if your horizon is the month. The gold mining stocks outlook 2026 that says “more upside” is a multi-year claim sitting on top of a week in which the metal failed to hold a rally on good inflation news. Soft inflation cut October hike odds and the gold price still gave the gain back, because the long bond and the dollar did not cooperate. A five-year bullish expert and a bad October can both be accurate. The investor who needs them to match will sell the five-year claim to escape the bad October, then buy it back after the easy part of the gold rally, which is how the twenty-year record got built.
Precious metals investment has a simpler alternative, and it should be said plainly. Own the metal if you want the price and you do not want the mine. Own a broad set of gold producer stocks if you want the gear and you can stand a 40 percent volatility and a deeper drawdown. Own a speculative slice of juniors only if you have already assumed it can go to zero. Mixing these into one idea called “gold” is how people discover, in a down month, that they owned the knob and thought they owned the bar. Gold miners vs gold is a choice. It is not a synonym.
What would confirm the upside
The upside case does not need a new all-time high to start proving itself. It needs a gold price that stops falling, costs that stop rising, and a share count that stops growing. A close back above $4,200, the shelf September broke, would say the metal’s bounce is more than a pause above $4,100. A stable or falling all-in cost in the next quarter would say the margin is allowed to widen. A major that buys a single-asset mine for a premium would say the scarcity Rule described is a market, not a speech. None of these is a buy signal by itself. Together they are the tape version of the expert case.
The case fails, in practice, if $4,000 gives way and costs stay high. Then the gear runs backward, high-cost mines cut guidance, and juniors cannot finance. It also fails quietly if the gold price rises and every gain is paid out in new shares and higher contractor bills. Gold mining profitability is a residual. Residuals can disappoint in a bull market. The gold bull market of the last few years already did some of that disappointing, which is why the long-term fund math still favors the bar even after a brilliant three years for the miners. Gold stocks performance has to be read in more than one window or it will lie.
Gold investment, done with a clear head, picks the window on purpose. If your window is five years and you think the price stays high while new mines stay late, the shares of real margins have a case to beat the metal. If your window is the next jobs report, you are trading a beta of 1.6 into a number you do not know. If your window is twenty years and you will not rebalance, the historical edge has been the bar, because the mines found ways to waste the edge. The expert is not wrong to see more upside in the shares. The expert is incomplete until the window, the cost, and the share count are in the same paragraph as the upside.
Who keeps the extra upside
Even when the math works, the owner of the share is not always the person who keeps it. A company can earn a wider margin and spend it on a project that never pours. It can earn it and pay it to a government that raises the royalty after the mine is built. It can earn it and use it to buy a worse mine at the top of a gold rally, which is a habit the sector has never quite lost. Gold mining profitability at the pit and gold stocks performance in your account are separated by those choices. The expert who stops at the pit is describing the ore. The shareholder lives after the board meeting.
This is why the royalty and streamer model sits in the same funds and does not behave like a dirty pit. A streamer pays a low fixed amount for a slice of production and skips the diesel, the pit wall, and most of the capital blowout. The gear is milder. The disasters are fewer. In a strong gold bull market the pure producer often wins the race, because its margin is fatter. In a year when costs jump, the streamer often loses less. Gold investment options are really a menu of how much of the mine’s trouble you are willing to eat in exchange for how much of the gear. There is no setting that is only the gear.
There is also a tax and a currency between the ounce and you. A Canadian miner reporting in U.S. dollars, with costs in Canadian dollars, can see a move in the exchange rate widen or shrink the margin before the gold price does anything. A U.S. holder of that share then takes a second currency bet on the way back. Gold mining stocks vs physical gold hides this. The bar in a vault does not have a Toronto payroll. The share does. None of this kills the case for the shares. It explains why a clean 1.6 beta is a long-run average, not a week you can budget.
So the practical split is dull, and dull is the point. If you want the gold price and you will not read a cost note, own the metal. If you want more upside and you will read the all-in cost, the debt, and the share count, a basket of real producers is the honest version of what the experts are describing. If you want the story of a discovery, admit you own an option and size it like one. Gold stocks to watch, under that split, are not five tickers from a headline. They are the names that still have a margin at a gold price you would hate to be wrong about. Everything else is a louder way to be early.
The close
Gold miners can outperform gold. They have done it, by a wide margin, over one year, three months, and three years into late September 2026. They can also underperform, and they did it in the last month, and they did it over the long life of the miner fund versus the bullion fund. The reason is one math problem. A fixed-cost mine turns a move in the gold price into a larger move in profit, and the market then adds a multiple on top. When the gold price rises faster than all-in sustaining costs, that is more upside. When it does not, the same structure is more downside. September showed the second. The prior three years showed the first.
Experts such as Rick Rule add a second idea. The industry did not explore enough, so a higher gold price over five years may be paid not only as margin but as takeover bids for ounces that already exist. That can be real. It is not a schedule, and it is not a reason to own every junior. Gold producers with wide margins and stable share counts are the clean expression. Gold developers are a race against the raise. Gold exploration stocks are options. Canadian gold stocks in the senior group are where much of the world’s miner beta already sits. High-cost restarts are a lesson in how fast a $3,000 all-in cost uses up a $4,160 gold price.
The theme is the filter. Do not buy mining stocks because an expert said more upside, as if upside were a property of the ticker. Buy the logic only if you accept the knob turns both ways, and only if the margin survives a gold price you would still respect if you are wrong. Gold mining stocks vs physical gold is a choice about that knob. The bar does not have a diesel bill. The mine does. The expert who forgets the bill is selling the three-year chart. The record since 2006 is still on the page.
Important information
This article is for information and education only. It is not investment advice and not a recommendation to buy, sell, or hold gold, mining shares, or any fund. Miners can fall more than gold. Costs can rise. Share counts can rise. Past outperformance does not predict the next year.
Fund comparisons for GDX and GLD, including approximate compound returns, volatility, drawdown, and beta, are from a published comparison dated around September 25, 2026, and from the GDX fact sheet near September 30, 2026. They are rounded and will not match every data vendor. Gold prices, the September decline, yield levels, and HSBC’s average forecasts of $4,490 for 2026 and $4,825 for 2027 are from early October 2026 market reports. Rick Rule’s comments are drawn from the published summary of his September 23, 2026 Palisades interview and from other 2026 interviews, not from a claim that he recommended any security named here. Company cost figures are from those companies’ disclosures. The $1,800 cost example is an illustration, not a forecast or a company estimate. This article does not consider any reader’s finances.