The headline says nearly $40. The tape says more, and the extra dollars do not change the point. On Wednesday, October 14, 2026, spot gold closed at $4,104.80 an ounce. On Thursday, October 15, live spot quotes sat near $4,156 to $4,159. That is a rebound of about $54 from the close, not $40. December futures were quoted near $4,183.90 at 9:15 a.m. New York time, a larger move off the same bad day. Use the print you can name. Do not round it down to make a headline tidy, and do not round it up into a new bull market.
Here is the only idea. Thursday's gold price rebound was a yield move. A yield move is not a mine. Falling Treasury yields gave bullion back a day it lost to hot U.S. inflation data. They did not change gold production costs at a single Canadian pit. The five gold mining companies below are worth watching because their cost sheets and their ounce guides are already in print. They are not worth watching because gold bounced.
This is not a recommendation to buy, sell, or hold Agnico Eagle, Kinross, Alamos Gold, Lundin Gold, Artemis Gold, gold bullion, or any other security. It is not investment advice, not a solicitation, and not a gold price forecast. Company guidance is a target, not an outcome. Mines miss. Costs rise. Yields reverse. You can lose money.
What the yield actually did
The U.S. 10-year Treasury yield closed Wednesday at 5.186 percent. In the morning of Thursday, a widely followed quote put it near 5.103 percent, down a little more than 8 basis points, with a day range that had already touched about 5.077 percent. Wednesday's own range had been violent. The 10-year opened near 5.222 percent, spiked as high as 5.334 percent, and still finished at 5.186 percent. A later market note on Thursday described the yield as having slipped from around 5.3 percent toward 5.08 percent.
That is a retreat. It is not a collapse. The same 10-year yield had printed a 52-week high near 5.368 percent on October 7. A week later it was still above 5 percent. Year to date, as of Thursday morning, the yield was up roughly 93 basis points. Over one year it was up roughly 113 basis points. Falling Treasury yields on one Thursday do not erase a year of rising Treasury yields. They erase a slice of Wednesday.
Wednesday's slice had a cause. U.S. inflation data came in hot. Consumer prices had already pushed traders to put a Federal Reserve rate hike, later this month, back on the table. Gold fell with that news. It closed under the prior day's $4,160.79 and, during the session, traded below $4,100 for the first time since early August. Silver had a worse day. One futures account put an intraday low near $55.56, about 6 percent down, a three-month low. Silver's Wednesday close was reported near $55.64.
Thursday did not cancel the inflation. A producer-price report for September was also hot: up 1.3 percent on the month against a forecast near 0.6 percent, and up 6.3 percent on the year against a forecast near 5.4 percent. Yields fell anyway. The bid in bonds came from somewhere else. A selloff in technology shares sent people toward haven assets. Bonds caught that bid. Gold rose with the bonds. The dollar was reported down nearly half a percent. Ole Hansen at Saxo Bank called the yield reversal dramatic and the gold recovery real, and he added that the technical outlook for gold was still fragile.
Hold those two facts apart. Federal Reserve interest rate policy did not change on Thursday morning. No vote was taken. Hike odds had risen on Wednesday because of inflation. Yields then fell because stocks broke. A policy path and a one-day yield are not the same object. Anyone who trades the gold price rebound as if the Fed had changed its mind is trading a stock-market scare and calling it a policy shift.
A nominal yield is not the real yield
Gold does not compete with a headline yield. It competes with a real yield, the nominal yield minus the inflation investors expect to keep. This article does not have Thursday's Treasury inflation-protected print in hand. It will not invent one. What it has is a nominal 10-year still near 5.1 percent, a hot consumer-price day on Wednesday, and a hot producer-price day on Thursday.
That mix can cut either way, and you should not pretend it cuts only one. If expected inflation jumps by more than the nominal yield, the real yield falls, and gold often likes that. If the Fed answers the inflation with a higher policy rate, and the market believes it, the real yield can rise later, and gold often does not like that. Thursday's tape only settled the first few hours. Bonds were bought because equities were sold. Inflation was not revised down. The real-yield question is still open. Real interest rates and gold are the pair that matters. Treasury yields and gold prices are the pair the screen shows. They rhymed on Thursday. Rhyme is not a model.
A gold price forecast that starts from this bounce is a forecast about the next yield, and then about the next inflation print, and then about whether the Federal Reserve follows through. None of those is a fact about a Canadian mill. Precious metals investing that skips that chain is a story about a day. The gold market outlook for the next year is a different document. It has to say what happens to real yields if the Fed hikes into a 5 percent 10-year, and what happens if it does not. This piece does not make that bet.
Where the metal sits, without a story pasted on it
Gold bullion prices near $4,159 on Thursday afternoon were about 23 percent under the all-time closing high of $5,405 set on January 29, 2026. They were about 6 percent under where 2026 began, on the year-to-date reading attached to Wednesday's close. The 52-week range in that same record ran from $3,948.50 to $5,405. Thursday's price was roughly $210 above the low of that range. A market can be well off its high, still well above its low, and up $54 on the day. All three are true. None of them is a signal to own a miner.
Silver bounced too, and by less in dollars and more in percent. Thursday morning quotes put spot silver near $56.70 to $57.20 after a Wednesday close near $55.64. A metal that can drop 6 percent and retrace a slice the next morning is not a stable input for a mine model. Mines sell gold and sometimes silver over quarters. They do not sell Thursday.
The honest gold price recovery, if you insist on the phrase, is this: bullion took back most of one bad session. It did not take back January's high. It did not take back the year's loss. It did not take back the cost increases already written into 2026 guidance at three of the five companies below. Watch the session if you trade the session. Do not file the session as a change in the ore.
Why these five, and why not a junior list
The five Canadian gold stocks are Agnico Eagle, Kinross, Alamos Gold, Lundin Gold, and Artemis Gold. They are producers. They have mills, costs, and guidance. That is the point of a watchlist on a yield day. You can put the bounce next to a number the company already published and see whether the bounce is large or small.
This is not a list of junior gold mining stocks. A junior, in the strict sense, does not have a mine in steady production. It has a drill program, a study, or a permit. A $54 move in bullion does not grade its core and does not fund its next season. If you came for juniors because the headline said "stocks to watch," you came to the wrong shelf. The juniors will feel a gold mood. They will not show you, this week, whether Thursday paid for a tonne of ore. These five will at least show you a cost.
They are also not a ranking. The order below is the order of the lesson, not the order of a buy. Agnico shows a low cost and a wall that moved. Kinross shows a guide that was cut by more than Thursday gave back. Alamos shows a seismic event that repriced a year. Lundin Gold shows a fat grade that the reserve does not promise to keep. Artemis shows a very low sustaining cost and a very large growth bill that sustaining cost does not include. Canadian gold mining stocks are not one trade. Canadian gold stocks with the same postal code can have opposite problems.
Agnico Eagle: the margin is wide, and the wall still moved
Agnico Eagle Mines, TSX and NYSE: AEM, is the scale name on this list. In the second quarter of 2026 it produced about 856,000 ounces. All-in sustaining costs were $1,459 an ounce. Total cash costs were $1,054 an ounce. Both were inside the company's guidance for the period. The realized gold price that quarter was $4,483. The margin between what it sold and what it spent to sustain the ounces was roughly $3,000 an ounce.
Put Thursday's spot next to that same cost. At $4,159 and an AISC of $1,459, the indicative margin is about $2,700 an ounce. The gold price rebound of about $54 adds about $54 an ounce to that margin. It does not add $54 to every ounce still in the ground, and it does not restore the realized price of the second quarter. The second quarter was a record price quarter. October is not. A cost advantage can survive a lower price. It cannot be described as if the record quarter were still the invoice.
The ounce problem at Agnico is not the yield. On July 1, rock moved in the Barnat pit wall. About a million tons. The company said roughly 370,000 ounces became inaccessible. Management spread that loss over three years: about 60,000 ounces in 2026, and about 115,000 ounces in each of 2027 and 2028. Full-year 2026 production was still guided inside the original range of 3.3 million to 3.5 million ounces, but toward the lower end. A wall does not care that the 10-year yield fell 8 basis points.
The cash problem, if it is a problem, is the opposite of distress. The second quarter produced record free cash flow of more than $1.3 billion. Shareholder returns were $625 million, including $400 million of buybacks. Cash reached about $3.5 billion. Net cash was about $3.3 billion. Fitch moved the rating to A- from BBB+. A company in that position does not need a $54 bounce to fund the next quarter. What it needs is the wall to stay where the new plan says it stays, and the rest of the mines to replace the ounces Barnat will not give.
There is a harder line in the same quarter, and it is not financial. A worker, Daniel Giroux, was killed at Upper Beaver on May 1. Management said there had been three fatalities in a year, and called that unacceptable. A gold price recovery does not belong in that sentence. Safety is not a margin. If you watch Agnico, watch the ounces lost to the wall, the cost staying near $1,459 or not, and whether the fatality record changes. Do not watch it as a levered bet on Thursday's yield.
Kinross: the guide cut was larger than the bounce
Kinross Gold, TSX: K and NYSE: KGC, gave the cleanest arithmetic on this list, and it did it in September, not on Thursday. On September 24 the company cut its 2026 and 2027 production outlook. The new range is 1.84 million to 1.86 million attributable gold-equivalent ounces a year. The old range was 1.9 million to 2.1 million. From the old midpoint to the new one, the cut is about 8 percent. All-in sustaining costs for 2026 were raised to $1,850 to $1,900 an ounce, from $1,730. Third-quarter production was flagged at about 425,000 gold-equivalent ounces.
The causes were named. Weather and metallurgical trouble at La Coipa in Chile. Weaker mining performance at Round Mountain in Nevada. Those are operating facts. A Treasury yield in New York did not make the pad at La Coipa wet, and a yield retreat will not dry it.
Now the dollars. The old AISC figure was $1,730. The new midpoint is $1,875. The increase is $145 an ounce. Thursday's bounce in spot gold was about $54. The cost guide took back more, per ounce, than the gold price rebound gave back. Even if you use the low end of the new range, $1,850, the increase from $1,730 is $120. Still more than $54. A trader who buys Kinross because bullion bounced is buying a stock whose own cost sheet moved by a larger amount in the other direction, a month ago.
The margin is still wide. At $4,159 gold and $1,875 AISC, the indicative spread is about $2,284 an ounce. That is not a broken mine. It is a mine whose owners were told to expect fewer ounces and a higher cost than the prior slide. Wide is not the same as unchanged. Precious metals investing that stops at "they still make money at $4,000" will miss a guide cut that is sitting in plain sight.
Kinross raised the 2026 shareholder-return target to 50 percent of free cash flow, from 40 percent. That is a choice about cash that arrives, not a promise that the ounces will. If production lands at 1.84 million instead of the old midpoint near 2.0 million, there are fewer ounces to throw off cash, even at a higher price. A higher payout ratio on a smaller pot is not automatically more cash. Read the next production report before you spend the ratio.
Alamos: a seismic event rewrote the year
Alamos Gold, TSX and NYSE: AGI, cut its 2026 plan on July 29. Production guidance fell to 510,000 to 560,000 ounces, from 570,000 to 650,000. All-in sustaining costs rose to $1,775 to $1,875 an ounce, from $1,500 to $1,600. Total cash costs rose to $1,175 to $1,275, from $1,020 to $1,120.
The company named the reasons. A seismic event at Young-Davidson cut expected mining rates and grades. Recovery of ounces at La Yaqui Grande slipped in time. Labour and contractor costs in Canada ran hot. Young-Davidson's own 2026 guide fell to 100,000 to 115,000 ounces, from 155,000 to 175,000. Its AISC guide for the year went to $2,500 to $2,600 an ounce. That is not a low-cost Canadian mine in 2026. It is a mine spending to hold ground.
The rest of the company is not Young-Davidson. The Island Gold District was still guided at 290,000 to 310,000 ounces, with AISC of $1,550 to $1,650. Mulatos was guided at 120,000 to 135,000 ounces, with AISC of $1,125 to $1,225. The blend is what produces the company-wide $1,775 to $1,875. A blend can hide a sick part inside a healthy average. If you watch Alamos, split the average.
The arithmetic against Thursday is the same shape as Kinross, and larger. The old AISC midpoint was $1,550. The new one is $1,825. The increase is $275 an ounce. The gold price rebound was about $54. The seismic event, in cost terms, was worth about five Thursdays, and that is before counting the ounces that will not be mined this year. At $4,159 and an AISC of $1,825, the indicative margin is still about $2,334 an ounce. Again: still profitable is not the same as unhurt.
The company has also published a path that gets cheaper later. Its 2027 guide, as of the September presentation, shows production of 650,000 to 730,000 ounces and AISC of $1,325 to $1,425. For 2028 it shows 755,000 to 835,000 ounces and AISC of $1,200 to $1,300. Those are company targets. They assume Island Gold ramps, and they assume Young-Davidson returns toward its old rate. A target for 2028 is not an ounce in 2026. Third-quarter production was expected, as of the July release, at 115,000 to 140,000 ounces, with AISC higher in that quarter because of sustaining capital and the lower output. The next print will tell you whether the cut was the cut, or the start of a longer cut.
Lynn Lake is in the growth budget and not in 2026 production. Growth capital for 2026 was guided at $657 million to $720 million. Sustaining capital was guided at $228 million to $255 million. AISC captures sustaining capital. It does not capture the growth bill. A gold mining stocks outlook that quotes only AISC will miss the cash that is leaving to build the next mine. That distinction matters more at Alamos and at Artemis than it does at a company that is mostly harvesting.
Lundin Gold: the quarter was fatter than the reserve
Lundin Gold, TSX: LUG, runs Fruta del Norte in Ecuador. Its published 2026 guidance is 475,000 to 525,000 ounces. Cash operating costs are guided at $900 to $960 an ounce. All-in sustaining costs are guided at $1,110 to $1,170. The mill is guided at 5,500 tonnes a day. The head-grade assumption in that guide is 8.3 grams per tonne. Recovery is assumed at 91 percent. Sustaining capital is $75 million to $90 million. Near-mine exploration is $56 million. Regional exploration is $21 million. The cost guide itself assumes $4,000 gold and $44 silver. Thursday's spot, near $4,159, is only a little above the price the company used to draw the cost line. This is not a case of a mine priced for $2,000 gold suddenly living at $4,000. The guide already lived near here.
The third quarter was a record, and the record is the trap. Fruta del Norte produced 153,736 ounces. It sold 148,326 ounces at an average realized price of $4,299. The plant treated 535,711 tonnes, about 5,823 tonnes a day, above the 5,500-tonne guide. The grade was 10.1 grams per tonne. Recovery was 88.3 percent, under the 91 percent assumption. Nine-month production was 392,472 ounces. To land inside the full-year range, the fourth quarter needs about 83,000 ounces at the low end and about 133,000 at the high end. The low end is a slow quarter by this mine's recent standard. The high end is another strong one.
Ten-point-one grams is not the plan. The plan on the guidance page is 8.3 grams. The reserve grade reported with the company's reserve work is near 7 grams, not 10. A quarter can be rich because the stope in front of the crew is rich. The next stope does not owe you the last stope. Jamie Beck has said, in public, that the mine has to go back toward reserve grade. A gold price rebound does not change that sentence. If the grade fades from 10.1 toward 8.3, and later toward 7, the ounces per tonne fall. The cost per ounce rises, because the same work produces less gold. That is a larger lever than $54 on the screen.
At the midpoint of the AISC guide, $1,140, and at $4,159 gold, the indicative margin is about $3,019 an ounce. Lundin is the low-cost producer on this list after Artemis, on a sustaining-cost basis, and it is a single-asset company. One mill. One country. One grade profile that is, this year, running hot. Single-asset margins look wonderful until the asset hits a bad quarter, a power problem, or a tax dispute. Ecuador is not a footnote. The company has been in a live argument with the state over a contract adjustment tied to high prices. This article will not retry that whole case. It will say only this: a high realized price is not free if the contract shares the surplus. Read the filings on that point before you treat $4,299 as the price the shareholder kept.
The expansion question is the other watch item. The guidance page still shows 5,500 tonnes a day as the 2026 mill. The third quarter already ran faster than that, at 5,823. Running faster on high grade is a gift. Building a larger mill so the gift can be replaced with more tonnes of ordinary grade is a different decision, with a capital bill and a recovery risk. Recovery in the record quarter was 88.3 percent, not 91. More tonnes at a lower recovery is not an automatic win. Watch the grade, the recovery, and the capital, in that order. The yield is fourth.
Artemis Gold: the sustaining cost is low, and the build is not
Artemis Gold, TSX: ARTG, owns the Blackwater mine in British Columbia. It is the youngest producer on this list, and the one a reader might mistake for a junior. It is not a drill story anymore. It is a mill that is still being expanded. That is a specific kind of risk, and it is a better kind to watch on a yield day than a company with no mill at all.
The second quarter of 2026 was a record: 74,063 ounces. All-in sustaining costs were US$955 an ounce, inside a full-year guide of US$925 to US$1,025. The company kept 2026 production guidance at 265,000 to 290,000 ounces. Cash cost in the quarter was US$813 an ounce. The AISC margin against the realized price that quarter was about US$2,519 an ounce, on the company's own unit table.
The third quarter, reported in mid-October, was 69,546 ounces, up 14 percent from a year earlier. Nine-month production was 205,532 ounces. The ore treated was 1.48 million tonnes. The grade was 1.55 grams per tonne. Recovery was 90 percent. Compare that with the second quarter: 1.34 million tonnes, 1.86 grams, 92.2 percent recovery. The mill ran harder. The grade fell. The recovery fell. Output fell from the record, and was still a solid quarter.
The rest of the year is arithmetic, not a slogan. To hit the low end of guidance, 265,000 ounces, the fourth quarter needs about 59,500 ounces. That is below the third quarter. To hit the midpoint, about 277,500, it needs about 72,000 ounces, a little above the third quarter and a little below the second-quarter record. To hit 290,000, it needs about 84,500 ounces, roughly 14 percent above the record quarter. The company can make the range without a hero quarter. It cannot make the top of the range without one. A $54 gold move does not add those ounces.
Now the bill that AISC does not show. Growth capital for 2026 was guided at $685 million to $755 million. Inside that, the phase-two expansion was $385 million to $435 million, phase 1A was about $110 million, and other expansion capital was $190 million to $210 million. Sustaining capital, excluding lease payments, was about $5 million. Resource expansion and exploration was $15 million to $20 million. A mine with a $955 sustaining cost and a $700 million growth year is not a $955 mine in the only sense a shareholder cares about, which is cash left over. Gold production costs of the sustaining kind are the cost of keeping today's ounces. Growth capital is the cost of tomorrow's ounces. Quote both, or quote neither.
Artemis also bought protection. Gold put options cover more than 80 percent of available spot sales each month into the second quarter of 2027. The premiums were $36.3 million. A put is a floor. It is not, by itself, a ceiling. The company did not, in the passage used here, sell away the upside. A bounce in gold bullion prices can still reach the shareholder on the unhedged slice, and on the slice above the floor. What the put does change is the downside story. A violent drop toward the 52-week low hurts a hedged seller less than an unhedged one, and the premium is the price of that sleep. Do not describe Artemis as locked out of Thursday's rally. Do not describe it as unexposed to a gold drop either. Read the strike prices in the filing. They are not in this article, and a hedge without a strike is a slogan.
At $4,159 gold and a $955 AISC, the indicative sustaining margin is about $3,200 an ounce. That is the widest sustaining margin on this list. It is also the margin of a mine in British Columbia that is spending heavily to get bigger, at a grade that just fell from 1.86 grams to 1.55 in one quarter. Grade at a bulk mine moves with the bench in front of the shovel. One quarter is not a reserve. The watch item is whether 1.55 was a dip or a new run rate, and whether the expansion stays inside $685 million to $755 million.
The same $54, five different ways
Line the bounce up against the cost sheets. The figures below use Thursday's spot near $4,159 and the cost numbers already cited. They are indicative. They are not a forecast of earnings. They ignore silver, copper, royalties, hedges, and the difference between produced ounces and sold ounces.
Agnico, at the second-quarter AISC of $1,459, shows a margin near $2,700. The bounce added about $54. The Barnat wall removed ounces, on the company's count, for three years. Fifty-four dollars does not put those ounces back.
Kinross, at the new AISC midpoint of $1,875, shows a margin near $2,284. The September cost increase, from $1,730 to that midpoint, was $145. The bounce was about $54. The guide cut was the larger event.
Alamos, at the new AISC midpoint of $1,825, shows a margin near $2,334. The July cost increase, from a $1,550 midpoint to $1,825, was $275. The bounce was about $54. The seismic event was the larger event. The 2027 and 2028 cost targets, if they are met, would matter more than Thursday. They are not met yet.
Lundin Gold, at an AISC midpoint of $1,140, shows a margin near $3,019. The third-quarter grade was 10.1 grams against an 8.3-gram guide assumption and a reserve nearer 7. A move from 10.1 grams toward 8.3 is a cut of about 18 percent in grade, before recovery. That is not a $54 problem. It is a tonne problem. Recovery already missed the 91 percent assumption, at 88.3 percent, in the record quarter.
Artemis, at a $955 second-quarter AISC, shows a sustaining margin near $3,200. Growth capital of $685 million to $755 million is outside that margin. The third-quarter grade was 1.55 grams after a 1.86-gram quarter. The fourth quarter needs about 59,500 ounces to protect the bottom of guidance and about 84,500 to reach the top. The bounce does not mill the difference.
A list of gold stocks to watch is a list of these gaps. The common gap is the same. The gold price recovery of one session is smaller than the operating news already on the page at Kinross, Alamos, Agnico, and, in a different way, at Lundin and Artemis. If the next session takes the $54 back, the cost sheets remain. If the next session adds another $54, the cost sheets remain. The sheets are the object.
What a stock is not required to do
Bullion was up a little more than 1 percent from Wednesday's close to Thursday's spot quotes near $4,159. Mining shares are not required to rise 1 percent when bullion rises 1 percent. They are not required to rise 3 percent either, which is the old rule of thumb people repeat when they want leverage without reading a balance sheet. Share counts, hedges, debt, growth capital, and local taxes sit between the ounce and the shareholder. Agnico's cash pile does not move one-for-one with the gold price. Artemis's puts do not move one-for-one with it. Kinross's lower ounce guide does not move one-for-one with it.
This article does not quote Thursday's share prices for the five. A price at noon is stale by the close, and a close is stale by the next CPI print. If you want the stocks, look up the prints yourself, then put them next to the cost and the guide. A stock that fell on a day gold rose is not automatically cheap. A stock that rose more than gold is not automatically right. The gold mining stocks outlook that matters is whether the next quarter's ounces and costs look like the guide. The outlook that does not matter is whether Thursday's tape "confirmed" the story.
What to write on the card
Write the metal with the hour. Wednesday, October 14, 2026: spot gold closed at $4,104.80, down from $4,160.79, and traded below $4,100 during the day. Thursday, October 15: spot quotes near $4,156 to $4,159. December futures near $4,183.90 at 9:15 a.m. New York time. The rebound from the close was about $54, not $40. The January 29 closing high was $5,405. Thursday was about 23 percent under that high. The 52-week low in the same record was $3,948.50.
Write the yield with the hour. The 10-year closed Wednesday at 5.186 percent, after a high of 5.334 percent. Thursday morning it was quoted near 5.103 percent, about 8 basis points under the prior close, with a low near 5.077 percent. The 52-week high yield was about 5.368 percent on October 7. The retreat is real. The level is still high. Year to date the yield was up about 93 basis points.
Write the cause as two days, not one slogan. Wednesday: hot U.S. inflation data, higher hike talk, a stronger dollar, gold down. Thursday: hot producer prices as well, up 1.3 percent on the month and 6.3 percent on the year, and yields down anyway because a technology selloff bid for bonds. The dollar eased. Hansen called the gold recovery fragile. Federal Reserve interest rate policy did not change in the morning. A speaker was due at the IMF meetings. A speech is not a vote.
Write the five costs, and nothing softer. Agnico: second-quarter AISC $1,459, about 856,000 ounces, 2026 guide 3.3 to 3.5 million ounces toward the low end after the Barnat wall. Kinross: 2026 and 2027 guide 1.84 to 1.86 million gold-equivalent ounces, AISC $1,850 to $1,900, up from $1,730. Alamos: 2026 guide 510,000 to 560,000 ounces, AISC $1,775 to $1,875, up from $1,500 to $1,600, after the Young-Davidson seismic event. Lundin Gold: 2026 guide 475,000 to 525,000 ounces, AISC $1,110 to $1,170, third-quarter grade 10.1 grams against an 8.3-gram assumption. Artemis: 2026 guide 265,000 to 290,000 ounces, AISC US$925 to US$1,025, second-quarter AISC US$955, growth capital $685 million to $755 million.
What an investor can do with a bounce
You can refuse to let a headline round the tape. Nearly $40 was the frame. About $54 was the spot move off the close. Almost $80 was the futures quote at 9:15 a.m. against Wednesday's cash close. Pick one print, name the hour, and stop. A gold price forecast built on a rounded headline is a forecast built on a slogan.
You can put the bounce next to the cost change before you put it next to a target price. At Kinross the cost guide rose by more than the bounce. At Alamos it rose by about five times the bounce. At Agnico the cost is low and the wall removed ounces. At Lundin the cost guide is low and the grade is above the guide. At Artemis the sustaining cost is low and the growth bill is large. That is the whole watchlist. Gold stocks to watch, on a day like this, are stocks whose last operating sentence was bigger than the day's price sentence.
You can keep real interest rates and gold in the right order. The screen showed Treasury yields and gold prices moving together on Thursday, yields down and gold up. The inflation data did not cooperate with a simple "Fed is done" story. Producer prices were hot on the day bonds rallied. If you cannot say what happened to the real yield, say that you cannot say it. Do not fill the blank with the nominal move.
You can ignore the leverage fairy tale. These gold mining companies do not owe you two or three dollars of share price for every dollar of gold. Cash, hedges, grade, and capital stand in between. Artemis's puts are a floor on a large slice of sales into mid-2027, not a ban on upside. Lundin's realized price of $4,299 in the third quarter is not the price in a contract dispute, and it is not the price at an 8.3-gram head grade. Agnico's $3.5 billion of cash is a fact about the past quarter's gold price as much as it is a fact about the next one.
You should not do the flattering thing. Do not call a 5 percent 10-year "low" because it fell from 5.33 percent. Do not call $4,159 "cheap" because $5,405 exists, or "expensive" because $3,948 exists. Do not turn five producers into a junior basket. Do not turn a junior's drill hole into a substitute for these cost sheets because the junior moves more on a loud day. Do not treat 2027 and 2028 AISC targets at Alamos as 2026 cash. Do not treat Lundin's 10.1-gram quarter as the reserve. Do not treat Artemis's $955 AISC as the full cash cost of a year that includes up to $755 million of growth capital.
What would make this reading wrong
The reading is wrong if Wednesday's spot close was not $4,104.80, or if Thursday's spot was not near $4,156 to $4,159 in the hours cited. Live prices move. A later print at $4,120 or at $4,220 changes the size of the bounce. It does not change the method. Compare whatever the close was with the cost guides. The method fails only if you use an unnamed "nearly $40" and never check it.
The reading is wrong if the 10-year did not close Wednesday at 5.186 percent, or was not near 5.10 percent on Thursday morning. A quote at 8:34 a.m. New York time is not the close. If Thursday's yield ended higher than Wednesday's, the "yields retreated" clause is a morning clause, not a day clause. Say which hour you mean. The link from yields to gold is also not a law. Gold can rise with yields, and it can fall with them. One Thursday is one Thursday.
The reading is wrong if the producer-price figures were revised, or if the 1.3 percent monthly gain and the 6.3 percent yearly gain were not the releases traders had in front of them. It is also wrong if the technology selloff was not the bid under the bond rally. Haven stories get told after the fact. The inflation story and the stock-selloff story can both be true. If a later account shows yields fell for a different dominant reason, use that reason. Do not keep a cause because it sounds clean.
The reading is wrong if Agnico's second-quarter AISC was not about $1,459, or if the Barnat ounces were not about 370,000, split as the company said. Those figures come from the company's quarter and from contemporaneous accounts of management's own split. A later reserve update that puts the ounces back, or takes more, changes the Agnico section. The fatality count is a disclosed operating fact. It is not a valuation input in this piece.
The reading is wrong if Kinross did not raise AISC to $1,850 to $1,900 from $1,730, or did not cut the ounce guide to 1.84 million to 1.86 million. The $145 gap between $1,730 and the new midpoint is the whole Kinross point. If the prior figure of $1,730 was a target that had already been abandoned, the comparison shrinks. Use the September 24 disclosure, not a memory of an older slide.
The reading is wrong if Alamos did not cut 2026 production to 510,000 to 560,000 ounces and did not raise AISC to $1,775 to $1,875. Those ranges are from the company's July 29 release and its September presentation. The 2027 and 2028 ranges are company targets in that presentation. They are not consensus, and they are not a promise. If a later guide walks them down, the "it gets cheaper later" sentence has to be marked as withdrawn.
The reading is wrong if Lundin's third-quarter grade was not 10.1 grams, or if production was not 153,736 ounces, or if the guidance page does not show 8.3 grams and AISC of $1,110 to $1,170. Nine-month output of 392,472 ounces implies a fourth-quarter requirement of about 83,000 to 133,000 ounces to hit the range. If some of those nine-month ounces were later reclassified, redo the subtraction. The reserve grade near 7 grams is the company's reported reserve, not a guess about the next stope. A new reserve that lifts the grade toward 10 grams would retire the mean-reversion warning. None has been cited here.
The reading is wrong if Artemis's second-quarter AISC was not US$955, or if 2026 growth capital was not guided at $685 million to $755 million, or if third-quarter output was not 69,546 ounces and nine-month output was not 205,532. The fourth-quarter ounces required are subtraction from the company's own range. The put options are a floor unless the filing also shows sold calls that this article did not use. If the structure is a collar, the upside is capped, and the sentence that says Thursday's rally can still reach the shareholder has to be limited to the uncapped portion. Read the strikes.
The reading is wrong if you treat every AISC as the cash cost of the business. AISC is a sustaining measure. It leaves out growth capital, and it can leave out items a shareholder still pays. Comparing a $54 gold move with an AISC change is fair. Comparing a $54 gold move with the full cash outflow of a building year is a different sum, and at Artemis and Alamos it is the more important one.
The idea, once
Gold rebounded on Thursday, October 15, 2026, because Treasury yields retreated. Spot was near $4,159, about $54 above Wednesday's close of $4,104.80. The 10-year yield, which had closed at 5.186 percent after spiking to 5.334 percent, was near 5.10 percent the next morning. Hot U.S. inflation data had knocked the metal down. A stock selloff knocked the yield down. The Federal Reserve had not voted. The real yield was not in the print this article will pretend to know. That is the whole price story. It is a day.
The five Canadian gold mining companies are not a day. Agnico's second-quarter cost was $1,459 an ounce, and a pit wall took ounces out of three years. Kinross raised its cost guide by $145 an ounce at the midpoint and cut its ounces, a larger mark than $54. Alamos raised its cost guide by $275 an ounce at the midpoint after a seismic event at Young-Davidson. Lundin Gold's sustaining-cost guide sits near $1,140, and its record quarter ran at 10.1 grams against an 8.3-gram plan and a reserve nearer 7. Artemis Gold's sustaining cost was $955, and its growth bill for the year was guided at $685 million to $755 million. Falling Treasury yields can lift gold bullion prices for an afternoon. They do not repair a wall, a grade, or a guide. Watch the cost sheet. The bounce already happened. It was the small number.