The headline is a search. It is not a plan. Gold is down about 27 percent from its record of $5,602.23 an ounce on January 29, 2026. On October 7 it traded near $4,100. A drop that large meets the old rule of thumb for a bear market in the metal. It does not meet the rule for a list of gold stocks to buy. The shares of the big miners are a different tape. The VanEck Gold Miners fund was still up on the year in early October, even after a one-month drop of about 11 percent. You can be in a bear market for the ounce and only a correction for the equities. People who shop the headline as if both were the same crash will buy the wrong thing for the wrong reason.
Here is the only idea that earns the five names. A bear market does not create five gold mining stocks to invest in. It creates one test, run five times. Can this company still fund its mines, its sustaining capital, and its promises if the gold price stays near $4,100, or goes to $4,000, and if its own costs do not stand still? Four of the names below passed the last quarter with wide margins. One of them has already cut the plan. The list is the test. It is not the purchase.
Nothing here is a recommendation to buy or sell any stock, any fund, or the metal. Costs cited are what the companies reported for a past quarter, or what they guided. A past margin is not a future margin. All-in sustaining cost is not the cash in the bank.
Are gold mining stocks a good investment in a bear market
Are gold mining stocks a good investment in a bear market? Only if you can say which bear you mean, and only if the mine is still a business at the new price. Gold investing in the metal and gold mining investment in the shares are not the same bet. The metal has no costs. The shares are the gap between the gold price and the cost of pulling it out, times a balance sheet, a mine plan, and a multiple that shrinks when people are scared.
In the boom, that gap was enormous. One desk lined up four majors against a second-quarter average gold price near $4,512. Agnico Eagle's reported all-in sustaining cost was $1,459, which left an implied spread near $3,053. Newmont's was $1,621, a spread near $2,891. Barrick's was $1,866, a spread near $2,646. AngloGold Ashanti's was $2,039, a spread near $2,473. Implied spread is just the gold price minus the stated cost. It is not profit. It ignores taxes, the mix of mines, hedges, copper credits, and the fact that a company does not sell every ounce at the quarterly average. It is still the right first cut. Even the most expensive of those four was earning, on that crude math, more than $2,400 an ounce over its sustaining cost. A bear market that takes gold from $5,600 to $4,100 does not make those mines poor. It makes the spread smaller. Smaller is not the same as gone.
That is why "good investment" is the wrong grade. At $4,100, and with those costs unchanged, the spreads would still be roughly $2,600 at Agnico, $2,500 at Newmont, $2,200 at Barrick, and $2,100 at AngloGold. Kinross, after it raised its own cost guide, would still sit near $2,200 over the middle of its new range. None of that is a forecast. Costs are rising at three of the four majors. Kinross has already lifted its cost guide and cut its ounces. The bear market in the gold price is real. The bear market in the business, so far, is a margin trim and, in one case, a missed plan. Gold stocks that you buy because the chart of the metal looks "cheap" after a 27 percent fall are a speculation on the chart. Gold stocks that you study because the cost line is still far under $4,100 are a study of a business. Only the second one survives a further drop.
What the phrase "best" is doing
Best gold mining stocks is a phrase that sorts tickers by hope. In a drawdown it should sort them by what would kill the story. A low cost that is stable kills fewer stories than a low cost that just jumped. A balance sheet with net cash kills fewer stories than a balance sheet that needs the gold price to stay high so the dividend can stay high. A company that cuts guidance in public is uglier this week and more honest than a company that has not reported the quarter yet. The five names are here because each one has a fact you can grade in the next set of numbers. They are not here because a model ranked them one through five.
The gold price outlook for the next month is still a rates story. Yields have been high. The dollar has been firm. December hike odds were elevated on the morning gold made a two-month low. None of that changes the ore in the ground. It changes the multiple people will pay for the ore. Gold mining companies can have record margins and falling shares at the same time. September showed a version of it. New York gold futures fell 6.4 percent. A set of large gold miners fell 12.7 percent. Not one of those fifteen names rose. The business did not get 12 percent worse in a month. The price of the business did. If you need the shares to rise because you bought them, that gap will hurt you. If you are asking whether the business survives $4,000 gold, the gap is the opportunity to look, not an instruction to buy.
Agnico Eagle: the cost that went the right way
Agnico Eagle is the cleanest of the five tests, and clean is not a synonym for safe. In the second quarter of 2026 it reported all-in sustaining cost of $1,459 an ounce. A quarter earlier the figure was $1,483. Over a year in which other majors paid what one commentator called the boom tax, Agnico's stated cost edged down. Against a quarterly average gold price near $4,512, the implied spread was about $3,053, the widest of the four majors in that comparison. At a gold price near $4,100, and with the cost held flat, the spread would still be about $2,640. That is not a distressed mine. It is a rich mine in a cheaper metal.
The test is whether $1,459 survives the quarter it is about to report. A public calendar had Agnico's next numbers on October 28. Oil has been expensive during this war. Labor and royalties do not freeze because the gold price fell. A cost that ticked down for one quarter can tick up for the next without the company doing anything foolish. If the October print holds near $1,459, the bear market has not touched the machine. It has only touched the share price. If the print jumps the way Newmont's did, from a low number to a much higher one, then the "margin king" label was a quarter, not a trait.
There is a second test that has nothing to do with the cost line. Agnico is a large, liquid gold stock. In a risk-off week it will fall with the other gold mining stocks, often by more than the ounce. High quality did not spare the group in September. A low sustaining cost does not hedge a multiple. It hedges the question of whether the company can keep the lights on. Those are different questions. Investors who buy Agnico because it is "the best" and then sell it on an 11 percent month have used the quality to justify a trade they could not hold. The cost figure is only useful if the holding period is long enough for the cost to matter more than the month.
Newmont: the cost that jumped
Newmont is the test of a single strange number. Its reported all-in sustaining cost was $1,029 in the first quarter of 2026 and $1,621 in the second. That is a jump of almost $600 an ounce in one quarter. Against the second-quarter gold price, the implied spread was still about $2,891. At $4,100, with costs stuck at $1,621, the spread would be about $2,480. The business is not broken by that arithmetic. The story is damaged if $1,621 is the new normal and the next print is higher again, because the market had just watched the cost line leave the basement.
A calendar had Newmont's next report around October 22. That print answers the only question that matters for this name in a bear market. Was $1,621 a mix, a one-time item, a copper or by-product swing, or the cost of running the portfolio at a high gold price, with contractors and energy priced for the boom? Companies do not always separate those causes in the headline number. The filing does, if you read the bridge from cash cost to all-in cost. A buyer who stops at $1,621 has a label. A reader who finds the bridge has a fact.
Newmont is also a size test. It is the largest name in most gold-miner indexes. When the index falls 11 percent in a month, Newmont is a large part of the reason, or a large part of the cushion. You do not get to own "gold mining stocks" as an idea and skip the biggest one. You also do not get to treat size as safety. A big company can run a cost jump just as well as a small one. The bear market does not care about the market cap. It cares whether the next all-in cost is closer to $1,029 or closer to $1,800. Until that number is out, Newmont is a question, not a holding you can defend with last quarter's margin.
Barrick: the cost that has been climbing
Barrick's reported all-in sustaining cost went from $1,708 in the first quarter of 2026 to $1,866 in the second. The direction is the fact. The level is still low against $4,100. The implied spread in the second quarter, using a $4,512 gold price, was about $2,646. At $4,100 and a cost of $1,866, it would be about $2,230. That is a wide business. It is a narrower business than Agnico's, and it is a business whose cost rose about $160 in one quarter. Two quarters of that pace would put the cost near $2,000 without any drama in the geology. The war has been a drama in the oil price. Energy is inside the cost.
Barrick is also a company the market argues about for reasons that are not the cost line. Asset sales, projects, and the copper mix have all taken turns as the story. This article does not need those arguments. In a bear market for the gold price, the cost trend is the argument that pays. A rising all-in cost plus a falling gold price is two squeezes on the same margin. Neither squeeze, at today's gap, threatens the ability to run the mines. Together they threaten the idea that last year's margin was the new base. People who underwrite Barrick at a $3,000 spread will be disappointed by a $2,200 spread even if $2,200 is still excellent. Disappointment is how gold stocks fall when the metal falls less.
The grade for Barrick, before any new number, is simple. If the next all-in cost is flat to down, the climb was a quarter. If it is up again, the climb is a path, and the path is the thing a bear market punishes. You do not need a target price to know which print is which.
AngloGold: the expensive end of a rich group
AngloGold Ashanti had the highest all-in sustaining cost of the four, at $2,039 in the second quarter. It also had the clearest upward slope. The sequence in that same comparison ran $1,720, then $1,805, then $1,955, then $2,039. Four prints, each higher. The implied spread against a $4,512 gold price was still about $2,473. At $4,100 it would be about $2,060 if the cost stuck. At $3,940, near the summer troughs some technical notes have marked, it would be about $1,900. The mine does not shut. The multiple can still compress, because the market pays up for margins that are expanding and pays down for margins that are only large.
This is the name that teaches the difference between "survives" and "leads." AngloGold survives $4,000 gold on these figures with room to spare. It does not lead a group whose costs are $1,459. In a bull market the gap is a detail. In a bear market the gap is the spread between the stock that people defend and the stock they sell first when they want less gold exposure. Being sold first is not a verdict on the ore. It is a verdict on the cost slope. A company that adds about $300 an ounce to its sustaining cost across four quarters is telling you the boom got into the cost base. The gold price has now left the boom. The cost base may not have.
The falsifier is a quarter that breaks the slope. A print under $2,000 would say the climb paused. A print over $2,200 would say the bear market in the metal and the inflation in the mine are arriving together. Neither print is a reason, by itself, to buy or sell. Each print tells you whether the cushion you think you own is the cushion you own.
Kinross: the plan that already broke
Kinross is on this list because it failed a test in public, not because it is the fifth "best" name. On September 23 it cut attributable production guidance for 2026 and 2027 to about 1.84 million to 1.86 million gold-equivalent ounces a year. The old range was 1.9 million to 2.1 million. The new range sits about 8 percent under the old midpoint, and 2 to 3 percent under the old floor. It raised 2026 all-in sustaining cost guidance to $1,850 to $1,900 an ounce, from $1,730. Third-quarter output was flagged near 425,000 ounces. The shares fell about 11 percent the next day, to about C$34. Desjardins cut its target and kept its own rating. That rating is theirs. It is not a finding that the cut was small.
The causes were operational, which is the point. At La Coipa in Chile, winter weather through the third quarter slowed mining and milling. Higher copper grades and weaker recoveries in parts of the sulphide ore forced a change to both years. The company is stockpiling high-copper material. A pre-feasibility study is looking at a flotation circuit. At Round Mountain in Nevada, mining performance was weaker. Paracatu and Tasiast, the two big low-cost mines, were still expected to produce a combined 1.1 million ounces, a fifth year in a row, in line with the prior guide. Total operating and capital costs were said to be on track even with higher oil prices. So the portfolio did not collapse. Two assets missed. The guidance moved anyway. That is how a bear market in operations starts, before the gold price has finished its own bear market.
Kinross raised the 2026 return-of-capital target to 50 percent of free cash flow, from 40 percent, in the same update. A higher payout ratio on a smaller production plan is not a gift. It is a choice. It can be a sign that the remaining cash flow is real and the board wants it in shareholders' hands. It can also be a sign that the company would rather defend the stock with a percentage than with ounces. Free cash flow is what is left after the mines are kept whole. If the cost guide of $1,850 to $1,900 is met and gold stays near $4,100, the spread is still wide, near $2,200 at the middle of the range. The ounces are the hole. Fifty percent of a smaller free-cash-flow number can be less cash than 40 percent of the old one. Read the dollars, not the percent.
This is the most useful name on the page for anyone asking whether gold mining stocks are a good investment in a bear market. The bear market people talk about is the chart of the metal. The bear market an owner lives through is a weather season, a recovery rate, and a guidance cut that arrives on a Wednesday. Kinross showed you that cut while gold was still well above $4,000. The next reports from the other four have not had to show you theirs. A list that includes only the companies that have not yet disappointed is a list of companies you have not finished reading.
The arithmetic at $4,100, labeled as arithmetic
Put the five cost figures next to one gold price so the romance drops out. These are subtractions. They are not earnings, not targets, and not bids.
Agnico, $1,459 against $4,100, leaves about $2,640. Newmont, $1,621 against $4,100, leaves about $2,480. Barrick, $1,866 against $4,100, leaves about $2,230. AngloGold, $2,039 against $4,100, leaves about $2,060. Kinross, at the middle of a new guide near $1,875, leaves about $2,225. The order of the cushion is Agnico, then Newmont, then Barrick and Kinross close together, then AngloGold. The order of the worry is different. Agnico's worry is whether the low cost holds. Newmont's worry is the $600 jump. Barrick's worry is the climb. AngloGold's worry is the climb that has lasted four prints. Kinross's worry is that the worry already happened, in ounces as well as in cost.
Now move the gold price and leave the costs alone, which is the optimistic case. At $4,000, every spread shrinks by $100. Nobody on this list is in trouble. At $3,940, near a band some technical desks have called a multi-month floor, the spreads shrink by about $160 from the $4,100 case. Still nobody is in trouble on cost. At $3,315, a stress case a Bank of America technical note floated in July if 2026 were a major top, Agnico's spread would still be near $1,850 and AngloGold's near $1,280, using these same costs. Even the stress case does not zero the margin. It does cut the margin in half from the boom, and halves are how dividends, buybacks, and project approvals get reviewed. The danger in this bear market is not that these five mines stop. The danger is that investors paid a boom multiple for a boom margin, and the margin is now a step down while the multiple is still being argued.
Move the costs and leave the gold price at $4,100, which is the pessimistic case the recent trend supports. Another $200 on AngloGold's cost, in line with its recent pace, takes the spread from about $2,060 to about $1,860. Another $200 on Barrick does the same kind of trim. A Newmont print that repeats the jump, rather than reversing it, is a larger event than $200. Kinross has already spent its $150 to $170 of cost increase in the guide. The gold price outlook can be flat and the stocks can still fall if the cost prints keep rising. That is the bear market inside the bear market. The ounce is the weather. The cost is the roof.
What "to invest in" is allowed to mean
Gold stocks to buy is a command. This piece will not give it. Gold investment in a miner, if it is an investment and not a bounce trade, is a decision you can write down in one sentence. You know the last all-in cost. You know whether that cost is rising. You know the next date the company has to speak. You know you can hold the shares through a month like September, when the group fell about twice what the futures fell. You are not using a 27 percent drop in the metal as a coupon.
If you cannot write that sentence for Agnico, do not let a low cost write it for you. If you cannot write it for Newmont, do not let the index weight write it for you. If you cannot write it for Barrick or AngloGold, do not let a still-wide spread hide a rising cost. If you cannot write it for Kinross, do not let a higher payout ratio hide a smaller mine plan. The five names are five chances to see whether your sentence is true. They are not a basket.
There is a sixth choice that the headline leaves out, and it is the honest one for many readers. A fund of gold mining stocks, such as the large-miner ETF, owns these problems in a blend. You will not catch Kinross's cut in time to duck one stock. You will also not need to be right about which of the four cost lines breaks next. You will get the group's gearing, which in September was about two dollars of share-price pain for each dollar of gold-price pain. That gearing is the product. Calling it a bear-market bargain because the metal is down 27 percent and the fund is not is a comparison of two windows. Check the window before you call it cheap.
Why the five costs are not one ruler
A reader who ranks these companies by the subtractions above has already gone too far. All-in sustaining cost is not a single ruler. Agnico, Newmont, Barrick, AngloGold, and Kinross do not produce the same mix. Some ounces are gold. Some are gold-equivalent ounces that fold in copper or silver. Kinross states its guide in gold-equivalent ounces, which is why a copper-recovery problem at La Coipa can move a "gold" number. A company with a large copper credit can print a lower gold cost in a quarter when copper is strong, and a higher gold cost when that credit shrinks. Newmont's jump from $1,029 to $1,621 may have a bridge like that inside it. The bridge is the document. The single number is the headline.
Sustaining cost also leaves out the spending that grows the company. A new shaft, a new plant, or a study of a flotation circuit is not in the sustaining line the way a tire and a stope are. Kinross can say operating and capital costs are on track and still be spending on a pre-feasibility study that does not sit in the $1,850 figure the way an owner imagines. A low all-in sustaining cost plus a large growth budget can consume the cushion the subtraction just showed you. The bear market test is not only "cost versus $4,100." It is "cost, plus the capital the company refuses to cut, versus the cash the mine will actually produce."
Jurisdiction is the third gap in the ruler. La Coipa's winter is a Chilean fact. Round Mountain is a Nevada fact. Paracatu is a Brazilian fact. Tasiast is a Mauritanian fact. A portfolio that averages them into one cost will hide the asset that missed. The 8 percent cut at Kinross came from a piece of the portfolio, not from every mine. The same will be true if one of the other four disappoints. Ask which mine moved the number. A company-wide all-in cost that rises because one pit missed is a different problem from a company-wide cost that rises because every pit got more expensive. The first can reverse with the weather. The second is the boom tax, and it does not reverse because the weather improved.
The reports that grade the test
The next three weeks do more work than this article. Newmont's report, if it lands near October 22, tells you whether $1,621 was a spike. Agnico's, if it lands near October 28, tells you whether $1,459 was a habit. Kinross has already spoken and will be graded on whether the third quarter really is about 425,000 ounces and whether La Coipa's winter is in the past or still in the mill. Barrick and AngloGold will be graded on the slope. A single good quarter does not end a climb. A single bad quarter does not end a low-cost mine. Two prints in the same direction start to be a path.
While you wait, the gold price will twitch on the Federal Reserve, the dollar, and the war. Those twitches will move all five stocks. They will not answer the test. A rally of $150 in the ounce makes every spread look smarter and does not fix a recovery rate in Chile. A drop of $150 makes every spread look dumber and does not mean Agnico's orebody changed. Separate the tape from the filing. The bear market in the tape is a multiple. The bear market in the filing is a cost and an ounce. Only one of those can break a mine.
The close
Five gold mining stocks do not become investments because the gold price fell from $5,602 to about $4,100. They become a short list of questions. Agnico Eagle's question is whether a $1,459 sustaining cost is a habit or a quarter. Newmont's is whether a jump from $1,029 to $1,621 snaps back. Barrick's is whether a climb from $1,708 to $1,866 stops. AngloGold's is whether four rising prints are the cost of the boom, still in the system after the boom in the metal has cracked. Kinross's question is already on the page. It cut the ounces, raised the cost, raised the payout ratio, and watched the stock drop about 11 percent. That is what a bear market looks like when it is a business, not a chart.
Are gold mining stocks a good investment in a bear market? They are a possible investment when the cushion over cost is still wide at the gold price you are willing to underwrite, and when you have read the one fact that would shrink it. They are a bad investment when the reason to buy is that a headline offered five tickers and the word bear. At $4,100, none of these five is a broken mine on the arithmetic above. One of them is a broken plan. The other four have not reported the quarter that this gold price will be judged by. Wait for the number. The list can wait with you.
A note on sources and limits
The gold price path is from October 7, 2026 market reporting. Spot set a high of $5,602.23 on January 29 and traded near $4,100 on October 7, a decline of about 27 percent. A 20 percent drop from that high would have stopped near $4,480. September's 6.4 percent decline in New York gold futures, and the 12.7 percent decline in a set of large gold miners, are from mining.com's October ranking. The VanEck Gold Miners fund's early-October snapshot, up on the year and down about 11 percent over one month, is from public ETF data around October 6. Those figures move.
Second-quarter 2026 all-in sustaining costs used here are the figures compiled from company reports by SilverTrade's miner-margin desk: Agnico Eagle $1,459, Newmont $1,621, Barrick $1,866, AngloGold Ashanti $2,039. The same desk showed Agnico at $1,483 in the first quarter, Newmont at $1,029, Barrick at $1,708, and AngloGold's sequence of $1,720, $1,805, $1,955, and $2,039. It used a second-quarter average gold price near $4,512 to describe implied margins of about $3,053, $2,891, $2,646, and $2,473. Implied margin is price minus stated cost. It is not earnings. Report-date mentions of about October 22 for Newmont and October 28 for Agnico are from that same public calendar and can slip. They are not company promises reproduced here in full.
Kinross figures are from its September 23, 2026 update, as reported by The Northern Miner and other outlets. Attributable guidance moved to about 1.84 million to 1.86 million gold-equivalent ounces for 2026 and 2027, from 1.9 million to 2.1 million. All-in sustaining cost guidance moved to $1,850 to $1,900, from $1,730. Third-quarter output was indicated near 425,000 ounces. The return-of-capital target moved to 50 percent of free cash flow, from 40 percent. La Coipa weather and metallurgy, Round Mountain, and the Paracatu and Tasiast combined 1.1 million ounces are from those reports. The share drop of about 11 percent to about C$34 on September 24, and the Desjardins target change, are from contemporaneous coverage of that update. Desjardins' rating is theirs.
Subtractions at $4,100, $4,000, $3,940, and $3,315 are illustrations. They assume the cited cost does not change. Costs will change. The $3,940 area is a technical reference from early-October bank commentary on summer troughs, not a forecast. The $3,315 figure is a July Bank of America technical stress case, not a base case. This article does not adopt either number.
This is not investment advice, not an offer, and not a solicitation to buy or sell any security. Gold and gold mining stocks can fall further. Costs can rise. Guidance can be cut again. Readers should read the company filings, check the live gold price, and speak with a licensed adviser before any decision.