UBS did not publish a treasure map. On October 9, 2026, an Investing.com account of the bank's work named six preferred gold mining stocks as the sector heads into a year the bank thinks will cost more to mine. The names are Newmont, Barrick, Endeavour Mining, SSR Mining, Skeena Resources and Franco-Nevada. Newmont is the preferred senior. Barrick is the other large name, even though UBS says Newmont got the better of their Nevada deal. Endeavour is the mid-tier with a free-cash-flow yield the bank puts above 10 percent at spot, and with country risk the bank does not pretend away. SSR is a cash pile looking for a life. Skeena is a mine being built. Franco-Nevada is a royalty company, which means it does not wear the diesel bill the way a digger does.
That is the idea, and it is the only one. A list made for 2027 is a cash filter for a year of higher unit costs. It is not a ranking of who has the most gold in the ground. Gold mining stocks that cannot still pay, or still pour, after energy and other costs rise are outside the logic of this note, even if the rocks are exciting. This piece explains the list as reported. It is not a recommendation to buy or sell any share, any metal, or any fund.
The year the list is built for
The price path in the same account is a warning before it is a gift. Gold hit a record near $5,400 an ounce in January 2026. It corrected to around $4,000 in July. It rebounded to roughly $4,700 in August. It closed September near $4,170. Anyone who owned the metal, or the miners, lived all four of those numbers. The GDX fund holds a basket of gold mining companies. It swung from a peak-to-trough loss of about 40 percent to a gain of about 50 percent at different points in the year. That is leverage in both directions. It is also why a bank that still likes the sector tells clients to be selective instead of full.
UBS's view of the setup, as reported, is two sentences that have to be held together. Valuations are generally reasonable. Margins stay attractive if the gold price holds above $4,000 an ounce. Then the second sentence. Sustained pressure from energy and from broader costs is expected to push unit costs higher in 2027. A margin that looks wide at $4,170 can narrow without the gold price falling, if the cost to pull an ounce rises. Gold mining profitability is a spread. The list is about the spread, not about the headline price.
A separate account of UBS projections, published by interactive investor on October 8, put the bank's gold price deck at $4,800 an ounce in 2027. It then put $4,500 in 2028 and 2029, and $4,000 in 2030. Silver was put at $77.50 in 2027, against a spot price then near $60. Those are the bank's numbers, not a promise, and they are not the same thing as a peak-year fantasy. A deck that walks down toward $4,000 over four years is a deck in which the miner who spends the cash windfall on a trophy, or on a cost base that only works at $5,400, gets hurt. The miner who returns cash, and keeps a cost line that still clears $4,000, is the one this framework can still like.
The bank's strategy line, in that October 8 account, is blunt. It expects more mergers. It also says not every deal is received badly. Its stance in the large caps is to avoid the companies that might do a deal. It would rather back sustained cash returns than hunt the mid-size and small names that might be bought. That is a choice. It is the opposite of a list built to guess the next takeover. Gold stocks to watch, in this note, are stocks the bank thinks are less likely to spend the year explaining a bid.
Why costs are the plot
Gold mining industry trends in 2026 have been a lesson in leverage. When the metal ran to $5,400, margins opened. When it fell toward $4,000, the same mines looked ordinary in a hurry, and the equities moved harder than the ounce. Energy is a large part of why. Diesel, power and reagents do not care what a slide deck said in January. Labour does not reset because a quarter was good. Contractors reprice when they can. A bank that sees those pressures lasting into 2027 is saying the easy part of the margin, the part that arrived free with the gold price, is the part that can leave.
A higher gold price still helps. UBS is not arguing that $4,000 is a bad price. It is arguing that $4,000 with a rising cost curve is a different business from $4,000 with last year's costs. Senior gold producers feel that in dollars per ounce across a very large book. So do the other top gold producers. A single-asset mine feels it as a question of whether the quarter still funds the plan. Gold development companies feel it as a capital budget that was drawn at one cost and will be built at another. Gold exploration companies feel it last and worst, because they do not have a margin at all until someone else funds the next hole.
That is why the published list is short, and why it leans to seniors, one mid-tier, one resetting producer, one build and one royalty. Selective exposure into third-quarter results and into 2027 guidance, which is the bank's timing, means the next set of cost numbers can change the argument. Guidance is where a miner admits what it thinks an ounce will cost next year. A preferred list that ignores that print is a list from last month. Readers should treat October's names as a snapshot of a view, not as a lock.
Newmont, the senior they would rather own
UBS calls Newmont its preferred senior gold producer. The reasons, as reported, are plain. Clarity on cash returns. A low probability of a material merger or acquisition. A modest improvement in the operating outlook. Execution, the bank says, has improved after earlier disappointments. None of those is a claim that Newmont has the best rocks on earth. They are claims about behaviour. A senior that knows what it will do with cash, and is unlikely to spend it on a large deal, fits the strategy of avoiding the companies that might transact.
The Nevada overhang is the concrete piece. Newmont and Barrick, through Nevada Gold Mines, run the largest gold complex in the United States. Barrick operates it and holds 61.5 percent. Newmont holds 38.5 percent. For years, some of the best nearby projects sat outside the venture, and the partners were in dispute. In 2026 they agreed to put them in. Barrick contributes Fourmile. Newmont contributes the Fiberline and Mike projects. Newmont pays Barrick $1.95 billion in cash. The ownership split stays the same. Disputes are resolved. The path is also cleared for a Barrick plan to list or separate North American gold assets, because Newmont's consent was part of the bargain.
UBS, in the October 9 account, treats Newmont as the clear winner of that bargain. The payment removes an overhang. It adds longer-term optionality inside a venture Newmont does not operate. A reader can disagree. Newmont is writing a very large cheque. Barrick keeps the operator's chair, keeps 61.5 percent, and receives the cash. "Winner" here is the bank's judgment about strategic position, not a court ruling. The useful point for the list is narrower. UBS thinks the fight is over, the assets have a home, and Newmont can go back to running a business that returns cash instead of arguing about a district. That is a senior gold producer thesis. It is not a junior story and it is not an exploration story.
Barrick, the partner who lost the headline and kept the job
Barrick is still on the list. UBS says so while admitting Newmont was the clear beneficiary of Nevada. The bank's answer is that Barrick's share price and its valuation already reflect that outcome. You do not have to discover the disappointment. The market, in this telling, has priced it. What UBS still likes is an operation that appears to be improving. Cash returns help in the near term. The valuation is one it calls attractive. The probability of a material deal looks low.
Hold the low-deal line next to the North American separation. A spin-off or an initial public offering of a regional gold business is not the same act as buying someone. It can still absorb management time, and it can still change what a shareholder owns. UBS's "low probability of material mergers and acquisitions" should not be stretched into "nothing corporate will happen." The Nevada agreement itself exists so a corporate step can happen. The distinction the bank seems to want is between a company that goes shopping with the cash and a company that hands cash back. Barrick can do the second while still rearranging the first. Investors who need a quiet holding company should read the separation language before they treat "low M&A" as "no headlines."
Barrick is also a Canadian market name in the way the Toronto listing makes it one, and a global miner in every way that matters to a cash flow. Its costs, its geopolitical spread and its copper as well as its gold sit outside a simple "senior gold producer" label. The October note, as summarized, does not walk those assets one by one. It walks behaviour and valuation. That is a thinner note than a mine model. It is still the note that put the name on the list. Gold stock performance from here, in the bank's frame, depends on whether the operating improvement is real and whether the cash actually comes back. A cheap stock that starts doing deals, or that misses costs, leaves the frame.
Endeavour, where the yield is the compensation
Endeavour Mining stock is the mid-tier name, and UBS does not dress it up as a safe senior. Higher country risk, the bank says, is already in the relative valuation. The compensation it points to is a spot free-cash-flow yield above 10 percent. It expects a further step up in cash returns in 2026. It thinks a healthy distribution yield can be kept up over the medium term, even while capital spending rises. The growth it cites is about 40 percent from 2025 to 2029, tied to building the Assafou project in Ivory Coast.
That pairing is the whole Endeavour argument. You are paid, in this model, to own a West African producer while it spends to get larger. Interactive investor's account of the same bank's work put operating assets in Senegal, Ivory Coast and Burkina Faso, plus development and exploration along the Birimian belt. A price target of 4,600 pence was set against a share price then near 4,004 pence. A target is the bank's opinion of value. It is not a forecast of the path, and it is not a reason to ignore the countries. Burkina Faso, in particular, has been a hard place to operate a foreign mine. A yield above 10 percent is often the market's way of saying the cash might not arrive as modeled, or might not be allowed to leave.
Capital spending that drives 40 percent growth is also a cost story, not only a growth story. Assafou has to be built at whatever diesel, steel and labour cost in 2027, not at the cost assumed when the study was signed. If unit costs rise across the industry, a project in build feels it twice. The operating mines earn less per ounce. The new mine costs more to finish. UBS says the distribution can survive that. Survival is a claim about the balance sheet and the gold price deck, not a claim that the build will be boring. Endeavour Mining stock is on the list because the bank thinks the yield pays for the risk. It is not on the list because the risk has been removed.
SSR Mining, a pile of cash and a smaller life
SSR Mining is the reset. UBS says the Turkey disposals are done, and that the case has moved on to delivery, to extending mine life, and to putting more than $2 billion of cash to work. The portfolio, in the bank's estimate, can sustain just under 500,000 ounces a year of gold-equivalent production. That is not a giant. It is a mid-size book with a very large cash balance relative to the story it has left.
Cash is not a strategy until someone says what it is for. The October account says life extensions and delivery. It does not say a buying spree. That fits the bank's wider bias against deals that the market may punish. A company with $2 billion and a production base under 500,000 ounces will be asked, every quarter, why the cash is not a bigger mine, a dividend, a buyback, or a bid. Each answer has a cost. A life extension is the quiet answer. It is also the one that can disappoint if the geology does not stretch. Delivery means the mines that remain have to do what the guidance says, in a year when the bank expects unit costs to rise.
The Turkey exit is the scar that makes the cash possible and the trust fragile. This article will not retell it as a thriller. Investors who own SSR, or who are reading UBS as a reason to look, should read the company's own account of what was sold, what liabilities remain, and what the cash is actually free to do. "More than $2 billion" is the bank's figure in a news summary. It is not a substitute for the balance sheet. A cash number that cannot be spent, because of tax, or claims, or a board that will not move, is not the yield it looks like on a slide.
Skeena, the one name that is still becoming a mine
Skeena Resources stock is the odd name on a list built around cash returns. Skeena is not a senior. It is a Canadian gold development company whose main asset is Eskay Creek in British Columbia. UBS, as reported on October 9, says key permits were secured in February. Construction is on track. The catalysts ahead are updates on the main parameters of Eskay Creek, and visible progress toward the commissioning the company is targeting. That is a build thesis. It is not a free-cash-flow thesis. There is no 10 percent yield to collect while the concrete is poured.
A development company in a cost-up year is a specific bet. The gold price can be fine and the equity can still suffer if capital costs rise, if the schedule slips, or if commissioning takes longer than the target. Permits in hand remove one class of surprise. They do not remove weather, contractors, or the difference between a study and a mill. UBS putting Skeena on a preferred list means the bank thinks those updates are worth owning into. It does not mean the bank has abolished construction risk. Junior gold mining stocks and gold development companies borrow the word "gold" from the seniors and then add a calendar the seniors have already survived. Skeena is closer to the second group than the first.
It is also the Canadian project on the list that is not yet a producer. Canadian gold mining stocks run from Newmont's Canadian assets and Barrick's listing, through real producers, down to a venture market full of stories. Skeena sits in the developer slot. It is not a TSXV gold stock in the sense of a prospect with a drill and a dream. Treating it as one will make the risk look smaller than it is, or larger, depending on the mood. The right size of the risk is a mine in construction in a jurisdiction that can permit, in a year when building things costs more. Canadian gold producers, the ones already pouring, are a different line on the same national map. Skeena is trying to join that line. Joining is the work. The list does not do the work for you.
Franco-Nevada, the royalty that skips the shovel
Franco-Nevada is the gold royalty company on the list, and it is there for a reason the diggers cannot copy. A royalty or a stream is a claim on revenue or on metal, not a claim on the diesel, the workforce and the mill. When unit costs rise, the miner absorbs them. The royalty, in the simple case, does not. That is why gold royalty companies can look calm in a year when gold mining companies look sweaty. The calm is not free. The royalty holder does not control the mine. If the mine stops, the cheque stops. If the mine is seized, restarted, or renegotiated, the cheque changes.
UBS's specific point, as reported, is Cobre Panama. The restart, the bank says, is mostly not in the price. Franco-Nevada trades at about 15 times spot 2028 enterprise value to EBITDA, against a five-year average of about 21.5 times. Those multiples are the bank's comparison, not a law of value. A stock at 15 times can be cheap because the market is wrong about a restart. It can also be at 15 times because the market is right to doubt the restart, the timing, or the terms. Cobre Panama is First Quantum's mine. It has been through a shutdown and a political fight in Panama. "Mostly unpriced" means UBS thinks a positive outcome is not capitalized. It does not mean the outcome has been secured. A royalty on a silent mine is a legal document plus a hope.
Even so, the royalty is the cleanest expression of the cost theme. If 2027 is a year of higher unit costs, a business that does not pay those unit costs has a structural edge, provided the ounces still flow. Franco-Nevada's edge is that structure. Its risk is concentration in a small number of large assets, of which Cobre Panama has been the loud one. Undervalued gold stocks, if that phrase is going to be used here at all, should be used the way UBS used the multiple: against the company's own history, and against a restart that may not happen on the schedule a model prefers. Cheap versus a five-year average is not the same as cheap versus a future that includes the mine coming back.
Who did not make this list
The absence is the rest of the argument. Gold exploration companies are not the core of an October list aimed at third-quarter results and 2027 cost guidance. An explorer does not have a unit cost in the sense a producer does. It has a burn rate. In a year when seniors are being told to hand cash back rather than buy growth, the explorer's usual buyer steps aside. That buyer is the mid-tier looking for a project. It is exactly the buyer this strategy avoids. That does not make exploration worthless. It makes it a different trade from the one UBS just described. TSXV gold stocks, which are where a great deal of Canadian exploration lives, are mostly outside this filter. A drill result can still reprice a venture share in an afternoon. It will not satisfy a test that asks whether 2027 cash returns survive higher diesel.
Junior gold mining stocks that are not yet building, and that are not Skeena, are in the same outer ring. Some will be the mid-size targets the bank says it is not trying to pick. If more mergers do arrive, a few of those targets will be the winners of a year whose official advice was to own the acquirers who do not acquire. That irony is normal. A strategy can be right on average and miss the single name that gets a bid. The cost of chasing every possible target is owning a pile of stories that do not get a bid and do not have cash. UBS, on the evidence of this list, would rather miss a bid than own the pile.
Other UBS work in 2026 named other favourites, and that should be said so this list is not treated as scripture. An August account included AngloGold and Genesis alongside Newmont, Endeavour, SSR and Franco-Nevada. A September pass through Australian coverage called Newmont the preferred large cap and named a long roster of local buys, with Catalyst as a smaller-cap preference. Lists change with the book, the region and the month. The October 9 international list is the one that matches this headline. It is six names. It is not every gold mining company the bank has ever rated Buy. Gold mining stocks to watch, if the phrase means "what did they actually put in this note," means these six. If it means "what might they add next quarter," the honest answer is that guidance season exists so the list can move.
How the six fail in different ways
A filter is not a promise that every name passes the year. Each one fails on a different rock.
Newmont fails if the operating improvement stalls, if cash returns are less clear than the bank thinks, or if the Nevada cheque buys optionality that takes a decade to show up in cash flow while costs rise now. A modest improvement is a modest claim. Modest claims are falsified by ordinary misses.
Barrick fails if the market was right to mark it down for Nevada and the operating improvement does not close the gap. It also fails if a North American separation becomes the distraction that "low M&A" was supposed to avoid. It also fails, like every producer, if 2027 unit costs eat the margin that looks fine above $4,000.
Endeavour fails if country risk stops being a valuation discount and becomes a stopped mine, a changed fiscal term, or a dividend that cannot be paid out. It fails if Assafou costs more and takes longer, and the 40 percent growth arrives late, while the yield the bank used to justify the risk does not. A free-cash-flow yield is a snapshot at a gold price. Move the gold price down the bank's own deck, toward $4,000 by 2030, and the yield is not a constant.
SSR fails if the cash is not as free as the summary sounds, or if life extensions do not replace the production that left with Turkey. A company can be rich and still shrink. Shrinking with a cash balance is a better problem than shrinking without one. It is still shrinking.
Skeena fails in the way builds fail. Capital cost. Schedule. Commissioning. A permit is a start. A pour is the test. If 2027 is the year costs rise, a project in construction is exposed to the thesis in its pure form. There is no operating margin yet to cushion a bad capital number.
Franco-Nevada fails if Cobre Panama does not restart. It fails if a restart comes on terms that do not match the stream. It fails if the multiple at 15 times is cheap for a reason that lasts. A royalty can also fail in the ordinary way, if the underlying mines produce less. The cost shield is real. It is not a shield against an empty pit.
What "selective" has to mean
Selective exposure is a phrase that sounds like advice and is really a refusal. UBS is not saying own the sector. It is saying own a few names whose behaviour matches a year of higher costs and possible deals. Gold stock performance in 2026 already showed what owning "the sector" felt like. Down 40 percent and up 50 percent, depending on the week you measured the GDX. A basket does that. A single name can do worse. Concentration in six ideas, or in one of them, is not safer than the basket if the one idea is the one that breaks. It is a different distribution of pain.
The gold price deck is the other restraint. Margins that need the price to stay above $4,000 are margins that the bank's own 2030 number only just clears. The 2027 number of $4,800, if it arrives, is a good year for a low-cost senior and a necessary year for a mine in build. It is not a number that makes every undeveloped ounce economic. Undervalued gold stocks, measured against a $5,400 January and never remeasured against $4,000 or against a rising cost, are a story about the past peak. The list is an attempt to remeasure. Readers who want a list of names that only work at the January price should look elsewhere. This list, read strictly, is for people who think $4,000 still matters and that costs will not sit still.
Cash returns are the behaviour the bank can see. A dividend and a buyback are not a geology report. They are a choice by a board that could have chosen a deal. UBS prefers the choice it can see, in the seniors, to the deal it might imagine. That preference will look wrong in any quarter where a refused deal would have been a good one. It will look right in any quarter where a completed deal destroys the cash return that was the reason to own the stock. Gold mining stocks to watch, under this rule, are stocks you can watch the cash leave the company toward the owner. A story you have to imagine into a buyout is a different sport.
The Canadian slice, without a flag
Canadian gold mining stocks are not a single asset. On this list the Canadian exposure is specific. Skeena is a Canadian developer building Eskay Creek. Franco-Nevada is a Canadian royalty company whose risk, on UBS's telling, is a Panamanian mine. Barrick is listed in Toronto and operates everywhere. SSR has a long Canadian market history and a cash-and-delivery problem that is no longer a Turkey problem. Newmont is American, with Canadian mines inside a global book. Endeavour is a London mid-tier in West Africa.
None of that is a reason to own a passport. A mine in British Columbia and a stream on a mine in Panama do not share a political risk, a cost curve, or a calendar. Canadian gold producers that are already pouring in Canada are barely the subject of this particular note. The note's Canadian names are a builder and a royalty, plus global miners who happen to report where Canadian shareholders can buy them. TSXV gold stocks remain the market where exploration risk is the product. They are not where UBS went looking for a 2027 cost answer. Confusing the venture exchange with this list is how a cash filter becomes a flyer.
How to read a preferred list without obeying it
A preferred list from a bank is a set of opinions with a letterhead. The letterhead is useful because it forces a reason next to each name. Cash returns. Low odds of a deal. A yield that pays for country risk. A cash balance after an exit. A build that has permits. A royalty multiple that assumes a restart is not in the price. You can reject any reason. You should not replace it with "they are top gold stocks" and then stop thinking. The reason is the part that can be checked.
Check the costs when the companies report the third quarter and when they give 2027 guidance. That is the window UBS itself pointed at. If unit costs are guided up by more than the gold deck can cover, the margin sentence breaks. If cash returns are cut to fund a surprise deal, the behaviour sentence breaks. If Assafou's budget moves, the Endeavour sentence breaks. If Eskay Creek's parameters move the wrong way, the Skeena sentence breaks. If Panama stays shut, the Franco-Nevada sentence breaks. A list that cannot be wrong is an advertisement. This one can be wrong name by name.
Size is the part the list will not do. Six names are not a portfolio unless you decide the weights, the currency, the time you can hold through another 40 percent drawdown, and the gold price you can live with. The bank's deck includes $4,000. A holder who needs $5,400 back in a straight line is not in this document. A holder who cannot sit through guidance that raises costs is not in it either. Gold mining profitability over the next year will be argued in those guidance tables, not in the headline that six companies "made the list."
The idea, once
UBS's October 2027 list, as reported on October 9, is Newmont, Barrick, Endeavour Mining, SSR Mining, Skeena Resources and Franco-Nevada. Newmont is the preferred senior, after a $1.95 billion Nevada payment that the bank thinks left it in the better seat. Barrick stays because the bank thinks the market has already priced that outcome, and because cash returns and a low chance of a big deal still fit. Endeavour is the yield that is supposed to pay for West African risk while Assafou is built. SSR is more than $2 billion of cash and a book the bank thinks can hold just under 500,000 gold-equivalent ounces, now that Turkey is behind it. Skeena is Eskay Creek, permitted and in construction. Franco-Nevada is the royalty, with Cobre Panama's restart treated as mostly not in a multiple that sits below the company's own five-year average.
The reason they are together is not that they are the best rocks. The reason is a year of higher unit costs, a gold price the bank will not assume stays at the January record, and a preference for cash returned over cash spent on a deal. Margins work, in this telling, if gold holds above $4,000. The bank's own deck has 2027 at $4,800 and later years stepping down toward that line. Exploration stories and most TSXV names do not answer that test. They answer a different one.
A list is a cash filter. It is not a treasure map. The companies on it can still fail, each in its own way, and the list can be rewritten after the next set of costs is published. That is what selective was supposed to mean.
A note on sources and limits
The six names and the reasons attached to them come from Investing.com's October 9, 2026 account of UBS's preferred gold mining stocks, written by Vahid Karaahmetovic. Gold's path from about $5,400 in January to about $4,000 in July, about $4,700 in August and about $4,170 at the end of September, the GDX swings of roughly minus 40 percent to plus 50 percent, the $4,000 margin line, and the expectation of higher unit costs in 2027 are from that account. This piece did not read UBS's original client note. Figures described as the bank's are the bank's as reported, not a new audit.
The Nevada structure is from company-related reports in September 2026. Barrick contributes Fourmile. Newmont contributes Fiberline and Mike. Newmont pays Barrick $1.95 billion. Barrick remains operator at 61.5 percent. Newmont stays at 38.5 percent. UBS's view that Newmont was the clear beneficiary is the bank's judgment, as carried on October 9. It is not a finding that the cash payment was small.
The $4,800, $4,500 and $4,000 gold deck, the $77.50 silver figure, the line about avoiding large-cap deals in favour of cash returns, Endeavour's 4,600 pence target against a price near 4,004 pence, and the West African asset sketch are from interactive investor's October 8, 2026 report on the bank's work. Endeavour's free-cash-flow yield above 10 percent, the 40 percent growth claim, Assafou, SSR's $2 billion and sub-500,000-ounce estimate, Skeena's February permits and construction comment, and Franco-Nevada's 15 times versus 21.5 times comparison are from the October 9 account. Earlier 2026 UBS lists that included other names are mentioned only to show that preferred lists change. They are not this list.
Nothing here is investment advice or a solicitation to buy or sell any security or commodity. Country risk, construction risk, cost inflation and a restart that is not agreed can all break a preferred name. A royalty can go quiet if the mine does. A cash balance can be less free than a headline. Mining equities can fall while gold rises. They can rise for a week and still be a poor purchase at the price you pay. Forecasts and price targets are opinions. Readers should read the company filings and the primary research, and should speak with a licensed adviser before any decision.