A stock that holds up is not the one that falls the least today. It is the ounce whose cost is still far under the new price, after you take out the hedge, the partner, and the capital the all-in number left out.
This is not a recommendation to buy or sell any gold mining stock, fund, bar, or futures contract. It is not personal advice. It is not a gold price forecast. It is not a ranking of the best Canadian gold stocks. Costs, ounces, and guides change. Read the filing. A study is not a mine.
The one idea is this. “Holds up” is a subtraction. Gold price minus the company’s own cost, on the ounces you actually own. A green day is not that subtraction. A Canadian address is not that subtraction. A low all-in sustaining cost that leaves a growth bill outside the line is not the whole subtraction. Five names below fail the test in five different ways. None of them is a winner because the metal fell.
Start with the print, not the adjective
On October 22, 2026, Dow Jones Market Data said front-month Comex gold settled at $4,042.80 a troy ounce. The drop was $87.70, or 2.12 percent. Dow Jones called it the largest one-day dollar and percentage decline since September 28. It was the lowest settlement since August 3. It was the third down day in four.
The same note put that close 23.98 percent under the January 29 settlement high of $5,318.40. Year to date, the front month was down $282.80, or 6.54 percent. Month to date, it was down 2.71 percent. The close was still 1.44 percent above the 2026 settlement low of $3,985.60, from July 16, and 2.41 percent above the 52-week low of $3,947.70, from November 4, 2025.
Hold those distances apart. A 2 percent day is gold price volatility. A 24 percent slide from the January high is a gold price decline on that series. A 6.5 percent drop from the start of the year is a third fact. Live screens on the same day showed other numbers, in the $4,040s and the $4,050s, and an early quote near $4,115. Use one series. The series in this piece is the settlement. Do not average it with a morning quote and call the blend a fact.
$4,100 was a headline because it is round. The settlement was $57.20 under that line. A trader with a stop can care. A mill sells ounces across a quarter. It does not invoice a round number. Gold stock performance on the day of the break is a mood. The margin is the arithmetic you can still do after the mood leaves.
What people mean when they ask which stock holds up
The question sounds like a shopping list. Which Canadian gold stocks fall the least if gold keeps falling? Which ticker is safest? Which one is the buy?
Those are three different questions, and the tape answers none of them. A stock can “hold up” on a screen because it is hard to sell, because it already fell last month, or because a buyer showed up for an hour. None of that is a cost. None of that is an ounce.
A gold investment outlook that starts with “Canada” has the same hole. Agnico Eagle, Alamos Gold, and IAMGOLD have mines outside Canada. Lundin Gold’s mine is in Ecuador. Kinross, which is not one of the five tests here, is Canadian-listed and pours gold on other continents. Artemis Gold’s Blackwater mine is in British Columbia, and its growth bill is still mostly ahead of it. A head office in Toronto or Vancouver does not put the tonne in your account. It does not set the power price, the permit, or the royalty.
There is also a bad tonne count in circulation. A line keeps saying Goldman Sachs tracked 79 tonnes of central bank buying in August, and that gold stocks are therefore set for another rally. The World Gold Council’s August figure, published October 6, was 39 tonnes of reported net buying, and 170 tonnes year to date. China added 20. Poland added 8. Uzbekistan added 8. Kazakhstan added 7. Thirty-nine is not seventy-nine. A central bank’s bar, bought in August, is not a bid for your share on October 22. It does not pay your all-in cost. It does not decide which gold mining stock holds up.
So the gold stocks to watch, in this piece, are not the five that a model likes. They are five ways the phrase “holds up” can lie.
The only subtraction that counts
Revenue moves with the gold price, on the ounces you sell. Cost does not have to. Profit is the gap. If the gap is wide, a small gold price decline is a small cut in profit. If the gap is narrow, the same decline is a large cut. That second case is operating leverage. At a $4,043 settlement, most of the guides below are still in the first case. The drama people want from a break of $4,100 is not in the arithmetic yet.
Run one example before the names. Take a cost of $1,500 an ounce. At $4,042.80 the gap is about $2,543. The day before, gold was $87.70 higher, at $4,130.50. The gap then was about $2,631. The price fell 2.12 percent. The gap fell about 3.3 percent. The share can fall faster than the metal. It does not fall like a mine that used to earn $200 an ounce. The starting gap is too wide.
Now drop the price to $3,000, because “holds up” usually means a larger fall than one day. The gap at a $1,500 cost is $1,500. From about $2,543, that is a cut of about 41 percent, on a price cut of about 26 percent. That is leverage. It is not what October 22 did. It is the test people skip when they rank stocks by the day’s percent change.
Three leaks sit beside that math.
The first leak is the word “all-in.” All-in sustaining cost is a company-built number. It often includes the cost of keeping today’s mine running. It often leaves out the pit, the mill, or the expansion you are paying for so the mine is bigger later. Two companies can both say AISC and not mean the same thing. By-product credits can make a gold cost look low. If those credits were priced at a metal price that has since fallen, the gold cost rises on paper without the rock changing.
The second leak is whose ounce it is. A table can be 100 percent of a mine, 70 percent, or 85 percent. Gold-equivalent ounces fold in copper or silver at a price the company assumed. A hedge can sell an ounce at a price that is not the settlement. Your share gets what is left after the partner, the state, the royalty, the hedge, and any new shares.
The third leak is time. A settlement pays ounces sold near that price. It does not pay ounces in a 2028 study, and it does not pay growth tonnes that are still a construction site. Do not paste $4,043 into a footnote that used $2,600, or $4,500, and call it the same bet. Do not paste January’s $5,318 in either. The high is a date. It is a planning price only if the engineer wrote it down.
Five tests, not five winners
The order below is not a gold stock performance ranking. It is not a list of the best Canadian gold stocks to buy. It is the order of a question. How wide is the printed gap? What did the company leave out of the cost? Whose ounce is it? What already broke before this week’s gold price decline?
Lundin Gold: the wide gap, in one country
Lundin Gold trades as LUG in Toronto. The mine is Fruta del Norte, in Ecuador. One mine. That is the whole production story in the pages used here.
In its second-quarter 2026 presentation, the company put all-in sustaining cost at $1,176 per ounce of gold sold. Cash operating cost was $1,016. Production was 118,994 ounces. Ounces sold were 110,385. That $1,176 is the quarter. It is not the year.
The company’s outlook page puts 2026 all-in sustaining cost at $1,110 to $1,170 an ounce, and cash operating cost at $900 to $960. The footnote assumes gold at $4,000 and silver at $44. The second-quarter all-in cost of $1,176 already sits a few dollars above the top of the year’s band. A guide and a result are not the same object. Do not average them into a number the company did not print.
At the October 22 settlement, a $1,176 cost leaves a gap of about $2,867. The top of the year guide, $1,170, leaves about $2,873. A 2 percent drop in the gold price barely moves either gap. On this one subtraction, Lundin is the sort of name people mean when they ask which stock holds up. The cost is low. The gap is wide. The gold price decline of one day does not close it.
The test it fails is concentration. A low cost in a single country is a wide margin and a single point of failure. Power, permits, politics, a mill, a road. Any one of those can close a gap that the gold price left open. The gold price forecast, whatever a desk publishes next, does not insure Ecuador. A Canadian listing does not either.
There is a second, smaller leak. The year guide is built at $4,000 gold. The settlement is $4,042.80. Those two prices are close. They are not the same price, and the silver credit in the footnote is $44, not whatever silver prints tomorrow. If the credit shrinks, the gold cost inches up. It does not, by itself, turn $1,170 into $2,000. Say the size of the thing.
Agnico Eagle: the gap stayed, the ounces moved to the low end
Agnico Eagle trades as AEM in New York and Toronto. On July 29, 2026, the company left full-year total cash costs at $1,020 to $1,120 an ounce and all-in sustaining costs at $1,400 to $1,550. The midpoint of that all-in band is $1,475. It said payable gold production for 2026 was now expected near the lower end of 3.3 million to 3.5 million ounces. The reason it named was a preliminary redesign of the Barnat pit at Canadian Malartic, after a rock-mass movement reported on July 2.
At $4,042.80, a $1,475 cost leaves a gap of about $2,568. The top of the band, $1,550, leaves about $2,493. The bottom, $1,400, leaves about $2,643. The second quarter was inside the band. Payable production was 855,816 ounces. Total cash costs were $1,054. All-in sustaining costs were $1,459. One good quarter is not the second half. July already said the year sits toward the low end of the ounce range because of Barnat.
This is the cleanest large-company version of “holds up” in the set, and it is still not a buy call. The cost band did not break when the gold price did. The company has more than one mine, so one pit is not the whole share. Canadian Malartic is in Quebec. Other ounces are not. Do not hear “Canadian gold stock” and picture one map.
What did change, before this week’s tape, was the capital bill. July said 2026 capital spending, excluding capitalized exploration, is now $2.6 billion to $2.8 billion. The prior band was $2.2 billion to $2.4 billion. The company tied the increase to construction at Hope Bay, approved on May 19. Capitalized exploration stayed at $290 million to $330 million. Do not add those two bands into a total the company did not print. Do not keep the May total, which included exploration under the old capital plan. That page was replaced.
Hope Bay capital is not inside the $1,475 midpoint. A wide margin that is then spent on the next mine is not cash you receive. “Holds up” on an all-in cost can be true while the cash flow does not hold up, because the company chose to build. That choice can be rational. It is still a choice. The gold price decline does not make it for you, and it does not unmake it.
Artemis Gold: the lowest printed cost, and the bill beside it
Artemis Gold trades as ARTG on the TSX Venture Exchange. The mine is Blackwater, in central British Columbia. On January 14, 2026, the company guided 2026 production at 265,000 to 290,000 ounces and all-in sustaining cost at US$925 to US$1,025 per ounce sold. It described that band as one of the lowest in the industry. The same note said the all-in figure includes about US$65 an ounce of regional and corporate costs, and that most of those are non-cash. Read that before you treat $925 as cash out the door.
On August 6, 2026, the company said it was keeping that production guide and that cost guide. It also said all-in cost for the year was expected to trend toward the higher end of the band, on inflation and higher oil, partly offset by work inside the mill. Second-quarter production was 74,063 ounces. Sales were 78,126 ounces. Cash cost was US$813 per ounce sold. All-in sustaining cost was US$955. Year-to-date all-in cost was US$1,013. The quarter’s all-in margin, on the company’s own line, was US$2,519 an ounce.
At the October 22 settlement, a $1,025 cost, the top of the guide, leaves a gap of about $3,018. A $955 second-quarter cost leaves about $3,088. On the printed subtraction alone, this is the widest gap in the five. People looking for the stock that “holds up best” will stop here. Do not stop here.
Of the 78,126 ounces sold in the second quarter, 49,052 were sold into the spot market, 6,246 were delivered into a gold stream, and 22,828 were delivered into hedge programs. The hedge piece was 15,828 ounces under a mandatory program and 7,000 under a discretionary one. Spot was about 63 percent of ounces sold. Stream plus hedge was about 37 percent. The hedge alone was about 29 percent. The release said gold sold into the spot market realized US$4,392 an ounce. It did not state the price received on the second-quarter hedge deliveries.
It did state prices for what was left. For the rest of 2026, 27,984 ounces remained under the mandatory hedge at a weighted average of $2,830 an ounce. The last 7,000 discretionary ounces were delivered in July at a weighted average of $3,345. The release’s own rule is that amounts are in Canadian dollars unless it says otherwise, and those two prices are not labeled as U.S. dollars in the lines used here. Do not convert them. A low all-in cost on a streamed or hedged ounce is a cost against a price that is not the October 22 settlement. Until the currency on those hedge prices is explicit, you also do not know how many U.S. dollars Blackwater actually received for that slice.
The larger leak is the cash the all-in line does not include, and the unit it is counted in. Sustaining capital guidance was $5 million. The release counts amounts in Canadian dollars unless it says otherwise, and that $5 million is not labeled as U.S. dollars. Growth capital, in the August update, was expected at $685 million to $755 million in that same default unit, to be funded from operating cash flow. That included $110 million to finish a $120 million Phase 1A expansion, $385 million to $435 million toward a $1.44 billion Expanded Phase 2, and $190 million to $210 million of other expansion capital, mostly tailings, water, and fleet. The all-in cost of US$925 to US$1,025 is a different currency. Do not set a Canadian-dollar build beside a U.S.-dollar cost and call them one number. Do not divide $685 million to $755 million by the ounce guide and place the result next to US$1,025. That division mixes units and overstates the cash burden in U.S. dollars.
Year-to-date sustaining capital expenditures and lease payments were about $19 million in the release’s default unit, against the $5 million sustaining-capital guide. The year-to-date line includes leases. It is not, by itself, proof that the $5 million guide is already broken. It is proof that the small sustaining line and the cash that actually left are not the same sentence.
Do not add the growth bill to all-in sustaining cost and pretend you have invented a new cost metric. The expansions are meant to last past this year. Do notice that the cash still leaves, in Canadian dollars, on a scale of hundreds of millions, while the U.S.-dollar all-in cost stays near $1,000. A producing mine’s sustaining cost, such as Lundin’s $1,176, is not the same object as a U.S.-dollar all-in cost with a Canadian-dollar build sitting beside it.
On October 13 the company said Blackwater produced 69,546 ounces in the third quarter, up 14 percent from the same quarter of 2025. Year to date, production was 205,532 ounces, and the company said it was on track for 265,000 to 290,000. The mill treated 1.48 million tonnes. Grade averaged 1.55 grams per tonne. Recovery was 90 percent. Phase 1A, a 33 percent lift in design throughput to 8 million tonnes a year, was 77 percent complete, with commissioning planned for the fourth quarter. Together with Expanded Phase 2, the company said the two steps would more than triple current processing capacity and lift annual production above 500,000 ounces in 2029.
To reach the low end of this year’s guide, the fourth quarter needs about 60,000 ounces. The third quarter was 69,546. The company says the year is on track. A 2029 figure of more than 500,000 ounces is not this year’s invoice. Gold price volatility between now and 2029 is not in that sentence. Neither is the cash required to get there.
So Artemis can have the widest printed gap and still be the name where “holds up” is easiest to misuse. The cost guide was written when the company itself, in January, was illustrating a margin at about US$4,500 gold. The settlement is now $4,042.80. The gap is smaller than the one in that January illustration, and it is still wide. The growth bill is in Canadian dollars. The hedge and the stream sold ounces away from the spot print. A falling gold price describes neither.
Alamos Gold: the cost rose before the metal fell
Alamos Gold trades as AGI in Toronto and New York. On July 29, 2026, it cut the year’s production guide and raised the year’s cost guide. This happened months before the October 22 settlement. If you are looking for what “holds up,” look at the July page first. The gold price decline this week did not write it.
Full-year production was lowered to 510,000 to 560,000 ounces, from 570,000 to 650,000. Total cash costs were raised to $1,175 to $1,275 an ounce, from $1,020 to $1,120. All-in sustaining costs were raised to $1,775 to $1,875, from $1,500 to $1,600. The company named three drivers. Lower mining rates and grades at Young-Davidson after a seismic event. The timing of ounce recovery at La Yaqui Grande. Higher labor and contractor costs in Canada, plus rehabilitation and extra ground support at Young-Davidson in the second half.
The mine lines matter more than the company line. Island Gold District production was set at 290,000 to 310,000 ounces, against a prior 290,000 to 330,000, with all-in cost of $1,550 to $1,650. Young-Davidson was cut to 100,000 to 115,000 ounces, from 155,000 to 175,000, with all-in cost of $2,500 to $2,600, against a prior $1,730 to $1,830. Mulatos District was set at 120,000 to 135,000 ounces, with all-in cost of $1,125 to $1,225. Lynn Lake sits in the capital table, at $140 million to $160 million of growth spending, and not in the 2026 production line.
At $4,042.80, the top of the new company all-in band, $1,875, leaves a gap of about $2,168. That is still wide. It is also about $275 an ounce worse, at the midpoint, than the old band of $1,500 to $1,600. The midpoint moved from $1,550 to $1,825. This week’s gold price drop was $87.70. The cost increase the company already booked is larger than the day’s price drop. A stock can “hold up” against a 2 percent tape and still be a worse business than it was in the spring, because the cost moved first.
Young-Davidson is the mine that does not match the company average. At a $2,600 all-in cost, the gap against $4,042.80 is about $1,443. At $2,500 it is about $1,543. Still positive. Much thinner than Island Gold’s gap at $1,650, which is about $2,393, or Mulatos at $1,225, which is about $2,818. If you own the share, you own all three, plus the Lynn Lake bill that produces nothing in the 2026 ounce line. The company all-in cost hides the mine that feels a lower gold price first.
The company said production should be stronger in the second half as Island Gold ramps, and that third-quarter production was expected between 115,000 and 140,000 ounces. It also said third-quarter all-in costs were expected to rise, on lower production and the timing of sustaining capital. A rising cost in the quarter after a guide was already raised is not a stock that has finished adjusting. It is a stock whose next print can still get worse without gold falling another dollar.
Island Gold had a strong second quarter on the company’s account. Underground mining and milling rates drove what it called record quarterly production of 67,500 ounces at that operation, and Magino’s mill averaged nearly 8,900 tonnes a day. A good mill does not cancel a seismic event at another mine. Gold stock performance will trade them as one ticker. The costs do not.
IAMGOLD: the cost column is already above the guide
IAMGOLD trades as IMG in Toronto and IAG in New York. Two pages are in play, and they are not the same date. The guidance page, in the view used here, carries no date. On October 21, 2026, the company updated Côté’s ounces and did not, in that note, restate the cost table.
The undated guidance page puts attributable production for 2026 at 720,000 to 820,000 ounces. Côté, at a 70 percent share, is 270,000 to 310,000 ounces, with 119,600 attributable ounces in the year-to-date column. Westwood, at 100 percent, is 110,000 to 130,000, with 68,600. Essakane, at 85 percent, is 340,000 to 380,000, with 183,500. The attributable total in that column is 371,700 ounces. That Côté figure of 119,600 is the pre-third-quarter number. On October 21 the company said Côté produced 64,400 attributable ounces in the third quarter, 92,000 ounces at 100 percent, and 184,000 attributable ounces year to date, 262,900 at 100 percent. Add 64,400 to 119,600 and you get 184,000. The same note said 2026 Côté production, on the 70 percent share, was expected at the low end of 270,000 to 310,000 ounces. The ounce test moved the day before the gold settlement. The cost test on the guidance page did not get a new sentence that day.
Consolidated all-in sustaining cost on that guidance page is $2,000 to $2,150 per ounce sold. The year-to-date column puts it at $2,195. That column lines up with the pre-third-quarter ounce count. It was not restated in the October 21 note. Do not call $2,195 a cost through October 22. Do call it a result that was already above the top of the year’s guide before the third quarter was added. Côté’s figure in that column is $2,094, against a guide of $1,775 to $1,925. Essakane’s is $2,159, against $2,000 to $2,150. Westwood’s is $1,921, against $1,950 to $2,100, so that one mine is inside its band and slightly under the low end. The company total in the column is not.
Cash cost tells the same story in a sharper way at Côté, on the same undated page. The year-to-date column shows cash cost including royalties of $1,622 an ounce, against a full-year guide of $1,200 to $1,350. Excluding royalties, that column shows $1,301, against $900 to $1,050. Those are not October costs. They are the page’s running total beside the old ounce count. The second half would have to be a very different mine for the year to land inside those bands. October 21 said processing rates had improved, including a September average above nameplate, and still put the year at the low end of the ounce range. Improvement is a claim about tonnes. The cost column is a result the company has not replaced. Do not let the old guide erase it, and do not pretend the October note rewrote it.
At $4,042.80, the guidance page’s $2,195 all-in figure leaves a gap of about $1,848. The top of the guide, $2,150, leaves about $1,893. The business is not, on these numbers, losing money at a $4,000 gold price. It was missing its own cost plan on the page that still holds the cost, while the gold price was still far above that plan. “Holds up” against a falling price can be true in the same year that the cost column does not hold up. Those are different sentences. A buyer who wants the first sentence and ignores the second is shopping for comfort. And the ounce sentence changed on October 21: Côté’s year was guided to the low end, on a newer count than 119,600.
Ownership is the other sentence. Costs on the page are attributable: 70 percent at Côté, 100 percent at Westwood, 85 percent at Essakane. The capital table shows Côté at the 70 percent share, Westwood at 100 percent, and Essakane at 100 percent, with a full-year total of $380 million sustaining and $120 million expansion, $500 million in all. Do not put an 85 percent ounce next to a 100 percent capital line and call them the same tonne. Your share is the attributable ounce, after the partner. The page is careful about that on the cost lines. A reader in a hurry will not be.
Côté is in Ontario. Westwood is in Quebec. Essakane is in Burkina Faso. The Canadian gold stock is three places. A cost problem at Essakane is not a cost problem at Côté, except inside the one consolidated line the market trades. If you want the name that “holds up,” you have to say which mine. The ticker will not say it for you.
Put the five gaps on one page
Use the October 22 settlement of $4,042.80. Use the cost the company printed. Round to the nearest dollar. These are not targets. They are subtractions.
Lundin Gold, second-quarter all-in cost of $1,176. Gap about $2,867. Year guide top $1,170. Gap about $2,873. One mine, in Ecuador. Guide built at $4,000 gold.
Agnico Eagle, full-year all-in band $1,400 to $1,550, left in place July 29. Gap about $2,493 to $2,643. Midpoint gap about $2,568. Ounces guided near the low end of 3.3 million to 3.5 million after Barnat. Capital excluding exploration raised to $2.6 billion to $2.8 billion.
Artemis Gold, guide top US$1,025, with the year expected toward that high end. Gap about $3,018. Second-quarter all-in cost US$955. Gap about $3,088. Of 78,126 ounces sold, 49,052 were spot, 6,246 were stream, and 22,828 were hedges. Growth capital of $685 million to $755 million is in the release’s Canadian-dollar default. It is not a U.S. dollar add-on to the all-in cost.
Alamos Gold, new all-in band $1,775 to $1,875. Gap about $2,168 to $2,268. Young-Davidson’s band is $2,500 to $2,600. Gap about $1,443 to $1,543. The company cost was raised in July, by more, at the midpoint, than gold fell on October 22.
IAMGOLD, guidance-page all-in cost of $2,195 in the year-to-date column, already above the guide of $2,000 to $2,150. That cost was not updated on October 21. Gap about $1,848. Côté’s cost in that column is above its own guide. Côté’s ounces were updated: 184,000 attributable year to date, and the year expected at the low end of 270,000 to 310,000.
If you rank them by the width of the printed gap, Artemis is first and IAMGOLD is last. That ranking is the trap. Artemis has the hedge book and the growth bill. IAMGOLD still has a four-figure gap and a year that is missing its cost plan. Lundin’s gap is wide and the mine is alone. Agnico’s gap is wide and the capital bill went up. Alamos’s company gap is wide and one mine’s is not. “Best” depends on which leak you refuse to count. A gold price forecast does not settle that. A falling price does not either.
What a larger drop would do, without calling it a forecast
This is not a gold price forecast. It is the same subtraction at two other prices, so you can see the gear.
At $3,500, Lundin’s $1,176 cost leaves a gap of about $2,324. Agnico’s $1,475 leaves about $2,025. Artemis’s $1,025 leaves about $2,475. Alamos’s $1,875 leaves about $1,625. IAMGOLD’s guidance-page figure of $2,195 leaves about $1,305. That figure was not updated on October 21. Every name in the five is still in the black on that page. The order of the gaps changes only a little. The share prices would not have to follow the order. Shares trade on mood, tax, capital, and the next headline. The ounce does not.
At $3,000, Lundin’s gap is about $1,824. Agnico’s is about $1,525. Artemis’s is about $1,975 on the U.S. dollar all-in cost, before the stream, the hedge, and the Canadian-dollar growth cash. Alamos’s company gap, at $1,875, is about $1,125. Young-Davidson at $2,600 is about $400. IAMGOLD at the guidance page’s $2,195 is about $805. The thin gaps get thin first. That is what “holds up” means if you insist on a stress, not a shopping list. It still does not make Young-Davidson the company, and it still does not turn Artemis’s US$1,025 into a cash cost after a Canadian-dollar build of $685 million to $755 million.
At $2,500, IAMGOLD’s guidance-page figure of $2,195 leaves about $305. That figure was not updated on October 21. Alamos at the top of its band leaves about $625. Young-Davidson at $2,600 does not leave a gap. Lundin, Agnico, and the Artemis guide still do. A price of $2,500 is not a prediction. It is a way to see that a 2 percent day under $4,100 and a price near the 2026 cost of the higher-cost mines are different events. People use the same word, “correction,” for both. The word is too small.
Gold price volatility is not the same as a broken mine
A gold price decline of 2 percent feels large on a screen because the dollar move is $88. Against a $1,176 cost, $88 is about 7.5 percent of the cost and about 3 percent of the gap. Against a $2,195 cost, $88 is about 4 percent of the cost and about 5 percent of the gap. The higher-cost mine feels the same dollar more. It does not, at this settlement, feel a loss.
The January high is the other exaggeration. From $5,318.40 to $4,042.80, the metal gave up about $1,276. On Agnico’s $1,475 midpoint, the gap went from about $3,843 to about $2,568. That is a real cut, about a third. It is not a shut mine. Anyone who marks a gold stock from the January high and calls the whole path a failure is marking a peak, not a business. Anyone who marks it from $4,100 and calls the whole path a bargain is marking a round number.
Oil sits in both stories. It can lift the inflation fear that weighs on gold, and it can lift diesel at the pit. Artemis said higher oil was one reason its all-in cost was drifting toward the top of the band. Alamos raised costs for labor, contractors, and rehab, not for a one-day gold print. Agnico raised capital for Hope Bay. IAMGOLD’s cost column was already hot on a page that does not mention October 22 at all, and Côté’s ounce guide was cut to the low end the day before. The mines moved before the headline. The headline did not create them.
A Canadian address is not a margin
If the question is “best Canadian gold stocks,” the honest split is this.
Artemis is the mine in British Columbia, with the lowest printed all-in band in this set, a stream, a hedge book, and a growth budget in Canadian dollars that dwarfs the sustaining line. Holding up on a U.S. dollar cost is not the same as holding up on cash.
Alamos has Island Gold and Young-Davidson in Ontario, Mulatos in Mexico, and Lynn Lake as a spend. The Canadian mines are not one cost. Young-Davidson’s new all-in band starts at $2,500. Island Gold’s starts at $1,550. The ticker averages them.
Agnico’s Barnat issue is in Quebec, and the company is larger than that pit. The cost band held. The ounce guide moved to the low end. The capital guide moved up.
IAMGOLD’s Côté and Westwood are in Canada. Essakane is not. The cost column on the guidance page is already above the guide, and it was not rewritten when the October 21 ounce note arrived. You cannot see that if you only ask whether the company is Canadian.
Lundin is Canadian-listed and not a Canadian mine. It has the wide gap and the single-asset risk. Leaving it out of a “Canadian gold” list because the pit is in Ecuador would be a mistake about risk. Including it because the head office feels local would be a mistake about geography. Say both.
How to watch without turning the list into a purchase
Write five lines, one per name. Not a price target. A cost, a date, and a leak.
Lundin Gold. Second-quarter all-in cost $1,176 per ounce sold. Year guide $1,110 to $1,170, at $4,000 gold. Leak: one mine, in Ecuador.
Agnico Eagle. July 29 all-in band $1,400 to $1,550. Ounces near the low end of 3.3 million to 3.5 million. Capital excluding exploration $2.6 billion to $2.8 billion. Leak: Barnat, and a build at Hope Bay that the all-in line does not contain.
Artemis Gold. Guide US$925 to US$1,025, trending toward the high end. Second-quarter all-in cost US$955. Growth capital $685 million to $755 million in Canadian dollars, not in the U.S. dollar cost. Leak: 6,246 streamed ounces and 22,828 hedged ounces out of 78,126 sold, and a build the all-in line does not contain.
Alamos Gold. July 29 all-in band $1,775 to $1,875, up from $1,500 to $1,600. Young-Davidson $2,500 to $2,600. Leak: a cost that rose before gold fell, and a company average that hides that mine.
IAMGOLD. Guidance-page all-in cost $2,195 in the year-to-date column, against a guide of $2,000 to $2,150. Not restated on October 21. Côté at 70 percent, Essakane at 85 percent. October 21 put Côté at 184,000 attributable ounces year to date and at the low end of 270,000 to 310,000 for the year. Leak: a cost column the year was already missing, an ounce count that moved, and a capital line at Essakane that is not on the same ownership basis as the ounces.
Then write the gold price from one series. The settlement used here is $4,042.80 on October 22. If the next report replaces a cost, your old gap is over. If a desk publishes a gold price forecast, it is their forecast. It is not a cost. It is not an ounce. It does not pay the partner, the hedge, or the expansion.
The answer the headline does not want
The headline asks which five Canadian gold stocks could hold up best while gold prices fall. The useful answer is that “best” is the wrong word, and “hold up” is a gap.
On the printed all-in cost alone, against $4,042.80, every name in this set still has a four-figure gap. Artemis’s guide is the widest. IAMGOLD’s guidance-page cost is the narrowest, it was already above the company’s own band, and it was not restated when Côté’s ounces were. Lundin’s gap is wide because the cost is low, and the risk is one mine. Agnico’s gap is wide, the ounces slipped to the low end of the range, and the capital bill rose. Alamos’s company gap is wide, and Young-Davidson’s is the first to thin if the price falls further.
None of that is a gold price forecast. None of it is a list of the best Canadian gold stocks. A stock that falls less than its peers tomorrow can still be the one whose cost guide broke in July. A stock with the lowest all-in cost can still be spending $685 million to $755 million of Canadian dollars that the U.S. dollar cost line does not show. A central bank’s 39 tonnes in August will not choose between them. A story about 79 tonnes will not either.
If you need one line, keep this. Gold prices are falling. The ounce holds up when its cost is still far below the new price, the ounce is yours, and the cash the cost left out still fits. Watch that. Do not watch the rank.