The headline number is a lower average for a whole year, and the bank’s own path steps down after it. A stock does not collect the average. It collects the price minus the cost, if the ounces show up.
This is not a recommendation to buy gold, buy a miner, or sell one. A bank forecast is an opinion. CIBC has already cut this opinion once this year, by a wide margin. A mine can miss ounces. A cost can rise. A share count can grow. Nothing here is personal advice, and nothing here is a price target for a stock.
The sentence people will repeat is the large one. Gold could reach $5,500 by 2028. The sentence in the October report is smaller, and it is a cut. Commodity analysts at CIBC Capital Markets, led by Anita Soni and Cosmos Chiu, now expect gold to average $5,500 an ounce in 2028. In their earlier forecast that average was $6,000. The 2027 average was cut to $5,200 from $6,500. The fourth-quarter 2026 forecast was cut to $4,500 from $6,500. A number that survived a cut is not the same thing as a number that was raised.
An average is not a peak
Reach is a verb for a high point. Average is a verb for a year. If gold spends 2028 at $5,500 on the bank’s math, some months will be above that and some will be below it. The year can “reach” $5,500 in the spring and still average less. It can touch $6,000 for a week and average $5,200. CIBC did not publish a high for the year. It published a mean. The headline smooths that difference until it disappears. Put it back.
There is a second difference, and it is the one that should slow a buyer down. Gold already traded near $5,600 earlier in 2026. CIBC itself noted that the metal had fallen nearly 24 percent from that record. A news note on October 19 marked spot gold at $4,147, more than 25 percent under the early high. By the next day one cost-curve snapshot had the price near $4,122, and Kitco described the market as clinging to support around $4,100. From about $4,150, a $5,500 average is roughly a third higher. It is also still below a price the market has already printed. The forecast is a recovery toward a neighborhood the tape has visited, not a voyage into a new country.
Do the path, not just the one year. Kitco’s account of the same outlook has the average at $4,800 in 2029, $4,500 in 2030, and $4,200 in 2031. On that account, 2028 is the high year of the averages, and then the line steps down. A person who buys a miner because “gold is going to $5,500” has skipped the years after the year. If the bank is right, the fat year is a visit. The long number after it is closer to $4,200 than to $5,500.
They are still above a consensus figure, if the October 19 note read the report correctly. That note puts the crowd at $4,620 for 2027 and $4,451 for 2028. Kitco’s writeup does not repeat those two crowd numbers. On the note’s math, CIBC is about $580 above the 2027 crowd and about $1,050 above the 2028 crowd. Being above the crowd is not the same as being right. In February this same bank was looking for a $6,000 average in 2026 and a $6,500 average in 2027. The October cut is the market marking their homework. A forecast that moved by $1,000 or more in eight months will move again.
Why the pencil came down
The cut has a cause, and the cause is not a sudden dislike of gold. Persistent inflation, made worse by the U.S.-Iran conflict and by high energy prices, pushed the Federal Reserve into a tighter stance. The Fed raised rates by a quarter point in September and left the door open to more. Higher real rates and a stronger dollar raise the cost of holding a metal that pays nothing. The 10-year Treasury yield moved above 5 percent. Bonds pay you to wait. Gold does not.
That is the headwind they name. They have not thrown out the old story. Their line, as carried by the October 19 note, is this: “Despite this monetary policy headwind, we believe elevated geopolitical risk, fiscal concerns and continued diversification away from US Treasuries will underpin gold demand and support a recovery through 2027 and 2028.” Geopolitics, debt, and a slow walk away from Treasuries are the three legs they kept. The rate path is the leg they had to shorten.
They also leave a door for a friendlier Fed later. If growth weakens, or if inflation cools enough, the Fed could ease more than the market now prices. They say real rates below 2 percent have been a supportive setting for gold in the past. A drop in inflation-adjusted yields is the catalyst they want for 2027 and 2028. Want is not a calendar. The September hike went the other way. The data since have not forced a reversal. A forecast that needs an easing the Fed has not promised is a forecast with a hinge.
The same energy shock that supports a fear bid for gold also feeds the inflation that makes the Fed hawkish. War can lift the metal and lift oil in the same season. Oil then lifts the consumer price index. The index lifts rate fears. Rate fears sell the metal. CIBC is describing that knot, not untying it. A reader who takes only the $5,500 and leaves the knot has not read the note.
The demand they need, in tonnes
A price average is the output. The inputs are tonnes. One October 19 account of the bank’s note says the bullish years assume gold exchange-traded fund demand of 1,000 tonnes in 2027 and another 1,000 tonnes in 2028, against just 35 tonnes for 2026. The same account puts central-bank buying at 1,000 tonnes a year in each of those two future years. Kitco’s writeup of the report does not repeat those forward tonnes. Treat them as one newsroom’s reading, not as a figure two outlets have both printed.
Hold those future tonnes next to the tonnes Kitco says the report already counted. Global gold-backed funds took in a record $30.7 billion in the third quarter, equal to 211 tonnes. The quarter before that, the funds had let 45 tonnes go. Bars and coins totaled 784 tonnes in the first half, one of the stronger first halves on record, with China and India still buying metal as a store of value. Central banks bought 345 tonnes in the first half. The second quarter alone was 289 tonnes, about five times the first quarter.
The two readings do not snap together without a seam. A third quarter of 211 tonnes, the figure Kitco reports, cannot share a year with a full-year assumption of 35 tonnes unless the other three quarters are a large net sale, or unless the 35-tonne figure is a different line. A reader should not add them. If the forward line is real, one thousand tonnes of fund demand in a single year is a very large year, and one thousand tonnes of official buying is a very large year too. The first half of 2026, at 345 tonnes on Kitco’s account, annualizes nearer 700 if the pace holds, not 1,000. On that reading the bank needs the official sector to speed up from an already busy half, and it needs Western funds to stay buyers after a year in which they have already sold and bought.
Futures traders have been the weak hand. Kitco says CIBC noted that speculative length has faded as yields and the dollar rose. The split the bank likes is physical buying that holds while the futures crowd steps back. That split can support a recovery. It can also flip. If the bars-and-coins buyer in Mumbai or Shanghai pauses, and the futures seller does not, the floor the bank is counting on is a story about two different customers. One of them can leave.
Mine supply is the other input, and here Kitco’s account of the bank is modest. It expects world mine output to rise only about 1 percent a year. High prices push companies to stretch old pits and build new ones. Grades fall. Permits take years. Disruptions happen. A 1 percent supply gain does not flood a market. It also does not guarantee the price. If the October 19 reading is right, and the funds buy 200 tonnes instead of 1,000, the gentle supply line will not save the average. If that 1,000-tonne line was misread, the hole in the story is the reading, not the mine.
What a miner actually receives
How a $5,500 gold price could affect mining stocks is a subtraction problem. The miner does not receive $5,500. The miner receives $5,500 minus the cost of getting the ounce out and keeping the mine open, then minus tax, then divided by the shares. If the cost is $1,500, the margin on the ounce is $4,000 before tax. If the cost is $2,200, the margin is $3,300. If the ounce is never mined, the margin is zero. The stock is a claim on that remainder, not a claim on the bank’s average.
Costs do not sit still while the gold price is invited higher. The reason CIBC cut the gold forecast is high energy, a firm dollar, and a hawkish Fed. Diesel, power, reagents, and wages are the miner’s bill. The same forces that delayed the gold price can fatten the cost line. A $1,000 rise in gold that arrives with a $300 rise in all-in sustaining cost is a $700 gift, not a $1,000 gift. All-in sustaining cost, the industry’s AISC, is the full bill of keeping the mine in business. It is not the only bill. It leaves out some growth spending. It is still the cleanest public yardstick, and it moves.
Use today’s neighborhood as the base, and label it as a sketch. At $4,150 gold, a mine with an AISC of $1,500 has about $2,650 of spread. At a $5,500 average, if the cost did not move, the spread would be about $4,000. That is a large widening. It is also a fantasy if you freeze the cost. The companies below have already raised cost guidance, lost ounces to a wall or a seismic event, or both, in a year when gold was still historically high. The spread is real. The freeze is not.
There is a second leak. Governments take more when the price is high. Royalties, income tax, and special mining taxes are not frozen at the rate that applied at $1,800 gold. A province or a host country can change the split. The ounce can be hedged, or it can be unhedged. Do not assume the big producers are a clean bet on the spot price. Alamos used $92 million in the second quarter to buy back and close the last 35,000 ounces of legacy hedges. A hedge book, even a small one being shut, means part of the move never reached the income line. Where the book is gone, the shares take the full move, up and down. That is good in a $5,500 year. It is the reason the shares fell when gold fell from about $5,600 toward $4,100. Operating leverage works in both directions. The bank’s average does not turn it off.
CIBC did not, in the notes reviewed here, hand out a list of five stocks to buy on the back of $5,500. The five names below are a curve, chosen because their latest reported costs and their latest operational misses show the spread between a cheap ounce and a dear one. They are Canadian in the sense that matters to a Toronto reader. They are listed here. They are not a portfolio. They are not ranked. A lower cost is not a buy signal. A higher cost is not a sell signal. Both are facts you can check against the next quarterly report.
Agnico Eagle, and a wall that gold cannot fix
Agnico Eagle is the large Canadian producer whose second quarter of 2026 still looked, on the numbers, like the sort of quarter a $5,500 believer wants to own. Production was 856,000 ounces. Total cash costs were $1,054 an ounce. All-in sustaining costs were $1,459. Free cash flow was more than $1.3 billion. Shareholder returns were $625 million, including $400 million of buybacks. One account of the quarter puts cash at $3.5 billion. Another stresses net cash of about $3.3 billion. Those are not the same pile. Fitch had already moved the rating to A- from BBB+, on April 29, before this quarter was reported. The upgrade is real. It is not a July event.
Then the pit answered. On July 1 a rock movement in the Barnat pit wall put roughly 370,000 ounces out of reach. That total is the figure two accounts of the quarter share. A call summary splits it into about 60,000 ounces in 2026 and about 115,000 in each of 2027 and 2028, and it describes about a million tons of rock. The press-note version checked here does not repeat the split. Use 370,000 as the hard number. Treat the year-by-year cut as what management was reported to have said on the call, and check it against the filing before you rely on it. Management said 2026 production would still land inside the original range of 3.3 million to 3.5 million ounces, but toward the low end. The same call summary says Canadian Malartic’s second-half cash-cost guidance went up. That line, too, needs the filing.
Read that against the bank’s fat year. Ounces Agnico cannot reach do not collect $5,500, whatever the average turns out to be. A higher price does not put a wall back. It can pay for a different mining sequence. It can make a lower-grade tonne worth blasting. It cannot repeal the July 1 movement. A call summary also records a death at Upper Beaver on May 1, and says management called three deaths in a year unacceptable. That is a report of the call, not a second document. A cost per ounce that ignores the people who produce it is not a full account of the business. Cash, whether the gross pile or the net one, is a cushion. It is not a promise that the next wall, or the next shift, goes as planned.
At the second-quarter cost of $1,459, a move from about $4,150 gold to a $5,500 average would widen the spread by roughly $1,350 an ounce if nothing else changed. The low end of 2026 guidance is 3.3 million ounces. That is a 2026 figure, not a 2028 mine plan. Multiply the two and you get a very large pile of extra revenue before tax, on the back of an envelope, and you have mixed two years. The envelope is the problem. The 370,000 ounces behind the wall are already the reason guidance sits at the low end. Use the company for scale. Do not use the envelope as a forecast of its earnings.
Lundin Gold, where the cost is low and the country is not Canada
Lundin Gold sits at the cheap end of the second-quarter cost curve. The company’s own release put all-in sustaining costs at $1,176 an ounce for that quarter. A compilation of reported costs placed that figure among the lowest on a list that ran from under $1,000 to well above $2,500. The same release reaffirmed 2026 cost guidance of $1,110 to $1,170 an ounce, so the quarter was above the band the company still expects for the year. A low cost is the cleanest way to show why the same gold price is not the same business. At $4,150, a $1,176 cost leaves about $2,970 of spread. At $5,500, if the cost held, the spread would be about $4,320. The cheap mine keeps more of every extra dollar. That is the arithmetic. It is not a view on the shares.
The mine is Fruta del Norte, in Ecuador, not in Ontario. The company is Canadian. The rock, the permit, the power line, and the political weather are Ecuadorian. A Toronto listing does not move the pit. CIBC’s demand story is global. So is the risk that sits under a low AISC. A cost number from one quarter does not describe the next royalty negotiation, the next community dispute, or the next rainy season. Lundin’s place on this page is the left side of the curve. The left side still has a country attached.
A low cost also changes how you should hear “$5,500.” A mine that already clears nearly $3,000 an ounce at today’s price does not need the bank’s average in order to be a profitable mine. It needed a gold price far below $4,000 for that. What $5,500 would change is the size of the surplus, and the fight over who receives it. Host governments notice surpluses. So do unions. So do shareholders who want the cash now. A cheap ounce can be a political ounce. Watch the cost. Watch the contract around the cost.
Alamos Gold, and a plan that is not yet a mine
Alamos Gold’s second-quarter all-in sustaining cost was reported at $1,728 an ounce. That is several hundred dollars dearer than Lundin and a few hundred dearer than Agnico’s quarter. The spread over $4,150 is still wide, about $2,400. The spread over $5,500 would be wider. The year in between has already been revised.
The cut is in the company’s own July materials, not only in a conference slide. Full-year 2026 production guidance was lowered to 510,000 to 560,000 ounces, from 570,000 to 650,000. At the midpoints, that is about a 12 percent cut. All-in sustaining cost guidance was raised 18 percent, on the company’s own words, to $1,775 to $1,875. The cause was operational. A seismic event at Young-Davidson in June limited expected mining rates to about 5,000 tonnes a day in the second half, and the timing of ounces at La Yaqui Grande slipped. A claim that the mine then climbs back above 7,000 tonnes a day, and a claim that the chief executive said he should have guided Magino more conservatively, come from a conference recap. They are not in the guidance release checked here. The release is the better guide to 2026 than a bank’s average for 2028.
The same July deck is more careful about the later years than a slogan is. It shows 2028 production of 755,000 to 835,000 ounces, and all-in sustaining costs of $1,200 to $1,300. A later report of company comment points at about a million ounces a year by 2030, helped by the Island Gold shaft and the Lynn Lake build. That million is not the 2028 line on the July page. The 2028 cost, if it arrived, would sit under CIBC’s 2028 gold average with a wide gap. It is a plan. Shafts sink late. Builds cost more than the slide. A seismic event has already moved one mine’s daily tonnes. A reader who treats the 2028 cost target as a fact is repeating the mistake the headline makes with $5,500. Both numbers are forward. Both have already been preceded by a cut to the nearer numbers.
Alamos is useful on this list because it shows the two clocks. The near clock is fewer ounces and a higher cost. The far clock is a bigger company with a lower cost, landing in the same year CIBC has marked as its high average. You can watch both. You should not pay today for both, as if the second clock cannot slip. Growth that arrives in 2029 collects whatever gold price exists in 2029. On CIBC’s own path, that later average is $4,800, then $4,500, then $4,200. A growth plan aimed at the fat year can miss it and land in the thinner years. Timing is part of the cost.
Kinross, where the guidance moved the wrong way
Kinross is the example of a cut that happened in the miner, not in the bank. In September the company reduced its 2026 and 2027 outlook to about 1.84 million to 1.86 million attributable gold-equivalent ounces a year, from a prior range of 1.9 million to 2.1 million. That is about an 8 percent cut from the old midpoint. All-in sustaining costs for this year were reset to $1,850 to $1,900 an ounce, up from $1,730. The second-quarter reported cost, before that reset, was $1,821. Third-quarter production was flagged near 425,000 ounces, under what several desks had modeled.
The causes were local and ordinary, which is the point. Winter storms at La Coipa in Chile left mining rates and mill throughput below plan into September. The company also hit higher-than-expected copper grades and poorer recoveries in some sulphide ore, and it is stockpiling some of that copper-rich rock. A conference recap put the copper grade in a band of 5 to 7 percent. The September news account does not. Use the company’s words, not the band. Round Mountain in Nevada mined worse than planned. Phase X, the higher-grade underground, was still scheduled to start contributing in 2028. That is not the same sentence as “production returns to 2.0 million ounces in 2028.” The second sentence is not in the September report. What is in it is a smaller company for 2026 and 2027, the years in which CIBC hopes the price recovers.
Kinross raised the share of free cash flow it aims to return in 2026, to 50 percent from 40 percent. That is a choice about the cash it does earn. It does not replace the ounces. Desjardins kept a buy rating and cut its target. The rating is someone else’s opinion. The cut in ounces and the rise in cost are the company’s. At a $1,875 midpoint cost, gold at $4,150 still leaves a spread of about $2,275. Gold at a $5,500 average would leave about $3,625 if the cost stayed. The September revision is the warning that the cost does not stay, and that the ounce count does not stay either.
Gold-equivalent ounces are their own small trap. An ounce-equivalent mixes gold with other metals using a price. If copper or silver moves, the “ounce” moves even when the gold in the mill does not. Kinross’s trouble at La Coipa was in part a copper problem inside a gold mine. The equivalent-ounce guidance is still the right public number to quote. It is not a pure gold count. A $5,500 gold price does not fix a circuit that is fighting copper.
Barrick, the large ounce that still has to be mined
Barrick Mining sits near the middle of the senior cost pack. Its second-quarter all-in sustaining cost was reported at $1,866 an ounce, in a compilation that cites the company release. The company’s own page puts 2026 all-in sustaining cost guidance at $1,760 to $1,950 an ounce. The midpoint is $1,855. Kinross’s new midpoint is $1,875. Barrick’s midpoint is a little lower, not higher. The ranges overlap, and Barrick’s top, $1,950, does sit above Kinross’s top. Do not say Barrick is the dearer of the two. Say the bands cross. At $4,150 gold the spread over the second-quarter cost of $1,866 is about $2,280. At $5,500 it would be about $3,630 if costs froze. Scale is the other fact. Barrick’s own 2026 gold guidance is 2.90 million to 3.25 million attributable ounces. A few hundred dollars of margin on about three million ounces is a different sum from the same margin on half a million ounces. Scale is also a different set of countries, pits, and labor forces that can each miss.
Mark Hill was named president and chief executive on October 19, after serving as interim. A conference recap had him saying the third quarter would be a fourth straight quarter on guidance, and that investors wanted something closer to eighteen months of that. That sentence is one room’s notes. It is not on the guidance page. The page is enough. In a year when gold made a record and then gave a large piece of it back, the market’s question to the large producers was not “what if gold is $5,500.” It was “can you do what you said.” Barrick’s place on a watchlist is that question. A bank average two years out does not answer it. The next four quarters do. Gold guidance of 2.90 million to 3.25 million ounces, and an all-in cost band of $1,760 to $1,950, are the numbers those quarters have to meet.
Barrick is not a pure call on the CIBC gold number for another reason. Copper sits in the business. The company guides to 190,000 to 220,000 tonnes of copper this year. The October 19 note, and not the Kitco writeup, has CIBC lifting copper to about $6.50 a pound for the fourth quarter and for 2027 and 2028. Take that metal price as one newsroom’s reading. Take the tonnes from the company. A company that sells both metals is two forecasts, not one. If gold recovers on the bank’s path and copper holds, the mix helps. If gold stalls and copper is the metal that was right, the stock is not a failed gold bet. It is a different bet wearing a gold headline. Know which ounces you think you own.
The curve, said in one place
Line the second-quarter costs up and the sermon gets short. Lundin Gold at $1,176. Agnico Eagle at $1,459. Alamos at $1,728. Kinross at $1,821, with guidance now $1,850 to $1,900. Barrick at $1,866. The gap from the cheapest of these five to the dearest is about $700 an ounce. At any gold price you pick, that $700 is the difference in spread before tax. A $5,500 average does not close it. It sits on top of it. The cheap mine keeps the extra $700. The dear mine does not catch up by wishing.
Now lay the operational cuts on the same line. Agnico’s 2026 ounces are aimed at the low end of 3.3 million to 3.5 million after Barnat, and about 370,000 ounces are out of reach. Alamos cut this year’s ounces to 510,000 to 560,000 and raised this year’s all-in cost, while its 2028 cost of $1,200 to $1,300 is a July plan, not a mine. Kinross cut two years of production and raised this year’s cost. Barrick’s 2026 band is 2.90 million to 3.25 million ounces at an all-in cost of $1,760 to $1,950, and the next quarters have to land in it. Lundin’s low cost is the company’s own second-quarter figure, and the mine is in Ecuador. Its 2026 cost guidance, $1,110 to $1,170, is actually below the quarter. Five different ways to fail a gold forecast that might itself be right.
This is the answer to the question of how $5,500 would affect the stocks. It would widen margins for the ounces produced, more for the low-cost ounce than for the high-cost ounce, and only after costs, tax, and the share count. It would not restore a wall. It would not undo a seismic event. It would not make a guidance cut into a beat. It would not arrive on January 1, 2028, and sit there. It is an average for one year on a path that the same bank has stepping down in 2029, 2030, and 2031. A share price that capitalizes one fat year as if it were a permanent price will be wrong even if CIBC is right.
What has to go right, and what can still break
For the average to land, a few large things have to happen together. Official buyers have to keep taking hundreds of tonnes a year. If the October 19 reading is right, the forward assumption is closer to a thousand. Funds have to buy on a scale that makes a 211-tonne quarter look normal rather than special. Mine supply has to stay near a 1 percent crawl, which is Kitco’s account of the bank. Real rates have to come down from the restrictive setting that caused the October cut, or the other buyers have to ignore the rates. The dollar has to stop being the bid that sold gold this autumn. Energy prices have to stop forcing the Fed’s hand. None of those is a small assumption. All of them can be true. They have not been true together in 2026, which is why the pencil came down.
The forecast can also be too low. CIBC was too low before it was too high. In an earlier note it had to chase the market up toward $6,000 and $6,500. Then it had to chase the market down. A bank that has been on both sides of the metal in the same year is not a compass. It is a record of how hard the year has been. If official buying accelerates, if the Fed is forced to ease by a growth break, and if the dollar gives way, an average of $5,500 can look cautious next to a spike. The long-term numbers Kitco reports after 2028, down toward $4,200, say the bank does not expect a spike to become the new floor. You can disagree with that fade. You should know the fade is in the same writeup as the number in the headline.
The stock forecast can be wrong even when the gold forecast is right. That is the ordinary way mining shares disappoint in a bull market. The price arrives. The ounces do not. The cost arrives with the price. The country changes the tax. The company issues shares to build the growth that was going to collect the price. Dilution is a silent cut to the per-share ounce. None of the five names is accused here of a financing. The point is structural. Equity is a residual. Residuals disappoint more often than the metal does.
There is a cleaner way to own the average, and it is not a stock. A bar, or a fund that holds bars, moves close to the price and does not have a pit wall. It also does not have a mine’s leverage. If gold goes from $4,150 to $5,500, the bar is up about a third. A miner with a $1,500 cost can be up by more than that on the ounce, and can still be down on the share if the mine stumbles. People who want the bank’s gold number can buy the metal. People who want the residual can buy the miner. Mixing the two, and calling both “a gold investment,” is how the headline’s five stocks get treated as five tickets to $5,500. They are not tickets. They are businesses.
The crowd is lower, and the crowd can be wrong too
CIBC’s 2028 average sits about $1,050 above the crowd figure in the October 19 note, $4,451. The 2027 gap on that same note is about $580, with the crowd at $4,620 and the bank at $5,200. Kitco did not reprint the crowd figures. A gap that large, if the note read the report right, is not a rounding error. It is a different view of how long the headwind lasts. If the crowd is right, gold in 2028 is only a little above where it trades this week. If CIBC is right, the year is a third higher than a $4,150 spot. The stocks will not split the difference for you. They will follow the path that actually prints, and they will follow it with a lag, a cost, and a mood.
Consensus is not a floor. It is an average of other opinions, and those opinions were also too low on the way up and too slow on the way down. In February CIBC lifted its 2026 full-year average to $6,000, from an October figure of $4,500, and it put 2027 at $6,500. Eight months later the new $4,500 is not that full-year average restored. It is a fourth-quarter forecast. Same digits. Different claim. The pencil still traveled to $6,500 and came back. Anyone who bought the February number as a destination paid for a forecast that the same desk has now retired. That is the useful history. It is not a reason to ignore the October number. It is a reason to treat it as a scenario with a date on it.
The fourth-quarter call is the near test. $4,500 against a spot near $4,100 to $4,150 is a rise of roughly 8 to 10 percent before year-end. That is a much smaller claim than $5,500. It can fail in a month. If the year ends near $4,100, the bank is already behind the first checkpoint, and the 2027 and 2028 averages will be asked to do more of the work. If the year ends near $4,500, the checkpoint is met and it still says nothing decisive about 2028. Watch the near number. It is the one the calendar can grade.
Five names, and the thing they are not
A watchlist is a list of things to check. It is not a basket, a weighting, or a view that these five beat the other fifty gold names on the Toronto market. Agnico, Lundin Gold, Alamos, Kinross, and Barrick are here because the latest public costs and the latest public misses are concrete. A royalty company would be a different subtraction. It does not run the pit, and it does not wear the full cost. A junior with no mill is a third subtraction. It sells hope and a treasury, and a $5,500 average does not pour gold into a hole that has not been drilled. Leaving those out is a choice about the theme. The theme is the ounce you can cost, not the story you can tell.
None of these five was picked because CIBC told you to buy it in the October gold note. If the bank has a separate rating on any of them, that rating is a different document, with a different date, and it can conflict with the commodity desk. Do not staple a stock rating to a metal average. Desjardins can like Kinross and still cut the target. A forum crowd can applaud a growth slide and still be early. The only common document is the gold price, and even that document disagrees with itself across the years.
Position size is the part a headline never does. A 10 percent drop in gold from $4,150 is about $415 an ounce. On a $1,500 cost, that is a large slice of the spread. On a share that has already run with the metal, the equity move can be larger than the metal move. The way down from $5,600 to $4,100 was the lesson, delivered this year, not in a textbook. People who size a miner as if $5,500 were a floor will meet the residual on a bad week. The residual does not send a letter first.
A cut, said once
CIBC says gold could average $5,500 in 2028. It used to say $6,000. It says $5,200 for 2027, not $6,500. It says $4,500 for the last quarter of this year, not $6,500. After the high year, its own averages step down. The metal is trading near $4,100 to $4,150, a long way under the $5,600 area it already reached. The gap from here to the 2028 average is real. The gap from the old forecast to the new one is also real. Both gaps belong in the same paragraph.
The demand case they kept is geopolitics, fiscal worry, and a diversification away from Treasuries, plus official buying and a tight mine supply. The case they marked down is the one where high oil and a hawkish Fed do not get out of the way. A gold price prediction that ignores the mark-down is a slogan. A gold price target that pretends an average is a peak is a different slogan. Canadian gold stocks do not need either slogan. They need a cost, a mine plan, and a gold price they can survive if the average never comes.
Watch Agnico for whether the low end of 3.3 million to 3.5 million ounces holds after the 370,000 ounces at Barnat. Watch Lundin Gold for whether a $1,176 quarterly cost stays near a guidance band of $1,110 to $1,170, at a mine that is not in Canada. Watch Alamos for whether 510,000 to 560,000 ounces is the last cut, and whether the July plan of $1,200 to $1,300 all-in costs in 2028 is still a plan next year. Watch Kinross for whether 1.84 million to 1.86 million ounces and a $1,850 to $1,900 cost are the floor of the revision. Watch Barrick for whether gold stays inside 2.90 million to 3.25 million ounces and an all-in cost of $1,760 to $1,950. Those are dates and tonnes. They are a better watchlist than a single year on a bank’s page.
If $5,500 arrives as an average, the low-cost ounce will have thrown off more cash than the high-cost ounce, and both will have thrown off more than they throw off near $4,150, provided the costs do not eat the move. If it does not arrive, the companies with the cleaner mines and the stronger cash will still be companies. The forecast will be a paragraph in an old note, revised again. Hold the distinction. The bank changed its number. The mines still have to make theirs.