Southern Copper

IThe business

Southern Copper digs copper out of the ground, turns it into metal and sells it. It is a Delaware company with its head office in Phoenix, Arizona, and its listing in New York and Lima. It was incorporated in 1952 and has mined copper since 1960.⁴ Its mines are in two countries. In Peru it runs Toquepala and Cuajone and the Ilo smelter and refinery. In Mexico it runs Buenavista, La Caridad and a group of underground mines called IMMSA, with smelters and refineries there too.⁴ It had 16,617 employees at the end of 2025, 64.4% of them under collective agreements.⁴

Figures are in US dollars, as the company reports them under US GAAP. Per-share values use 853.3 million shares: the 844.3 million outstanding on 29 October plus the 8.95 million to be issued on 19 November as a stock dividend, because the closing price on 4 November already excludes it.⁵¹³ TTM means the twelve months to 30 September 2026. A pound of copper is 0.4536 kilogram; LME is the London Metal Exchange price and COMEX the New York one. The third-quarter results were released on 20 October and the 10-Q was filed on 30 October.³⁵

In the nine months to September, sales were US$13,237M, up 38.6%, and net income was US$5,048.6M, up 66.8%.³ By metal, copper was 72% of sales, molybdenum 11%, silver 9%, zinc 4% and other products, mostly sulfuric acid, 4% C (calculated).⁵ Nothing about the volume explains the jump. Copper sold fell 2.3% to 685,760 tonnes. The LME price averaged US$6.09 a pound against US$4.33 a year before, up 40.6%, and silver doubled.³ This is a business whose profit follows the metals.

Where the copper comes from

Mined copper was 2,108.2 million pounds in 2025, or about 956,000 tonnes, down 1.8% on 2024.⁴ In 2026 the company expects 918,000 tonnes, 4% lower, and the nine-month figure was 693,602 tonnes, down 2.9%, because Peruvian ore grades are lower.⁵³ It also expects 29,000 tonnes of molybdenum, 163,400 tonnes of zinc and 24 million ounces of silver.⁵

Buenavista alone is 43% of the copper. It is an open pit near Cananea in Sonora that the company says can reach 500,000 tonnes a year, and it holds the zinc concentrator that produced 257.4 million pounds of zinc in 2025.⁴ Peru’s Toquepala and Cuajone together give another 43%, and they are where grades are falling: the company says Peruvian output was down 6.1% in the third quarter.³

What it is building

Growth is part of the case for the stock, so here is the list. The company aims for 1.6 million tonnes a year, against 918,000 now.³ It says Tia Maria and El Pilar would together add about 156,000 tonnes of refined copper and take it past one million tonnes, 85% refined.³

A plain reading: only Tia Maria and El Pilar are funded and under way, and between them they add about 17% to 2026 output. The jump to 1.6 million tonnes needs Los Chancas, Michiquillay, the Cuajone expansion and the Mexican pipeline, which are still studies or talks in two countries. My base includes Tia Maria, El Pilar, Cuajone and Los Chancas, the last delayed to 2034, and reaches about 1.2 million tonnes by 2036. My bull includes Michiquillay and reaches 1.4 million.

Funding and the dividend

In June the company sold US$1.25B of 5.350% notes due 2036, with net proceeds of about US$1,241M for Tia Maria and Peruvian capital spending.¹⁸ Debt at 30 September was US$7,996M against US$8,063M of cash and short-term investments.³ The board pays a quarterly dividend in two parts. On 15 October it declared US$1.30 in cash and 0.0106 of a new share per share, payable on 19 November to holders of record on 3 November.¹³ The company values the stock part at US$2.10, using the 15 October average price of US$197.83, and calls the total US$3.40 a share.⁵ The cash part is the money; the stock part is a small split. A holder owns 1.06% more shares, each worth correspondingly less. At the price, cash alone is US$5.20 a year, or 2.9% C (calculated), not the 7.7% that the headline suggests.

Why issue shares at all? The 10-K says the board has approved stock dividends every quarter since the second quarter of 2024, using shares it bought back before 2016.⁴ Those 119.5 million repurchased shares cost US$2.9B, and the treasury balance is now down to about 40 million shares C (calculated).⁴⁵ At about 9 million shares a quarter, that supply lasts about five more quarters. After that the company must issue new shares or end the stock part. Shares outstanding rose from 773.1 million at the start of 2024 to 844.3 million, up 9.2% in under three years C (calculated).²

IIThe moat

What keeps others from taking the profit? In copper the honest answer begins with what does not protect it. Southern Copper sells a commodity at a price it doesn’t set. There is no brand, no switching cost and no network. What it has is a low position on the cost curve and a large ore body that can’t be copied.

What protects it

• The ore. At 31 December 2025 its reserves held 108,955 million pounds of copper, estimated at a copper price of US$3.30 a pound: 45,527 million in Peru, 44,154 million in Mexican open pits, and 19,274 million in development projects.⁴ That is 49 million tonnes, or about 52 years of 2025 mining C (calculated). The company says it believes it holds the largest copper reserves in the industry.⁴ I can’t check that claim against rivals’ reserves, which are set at other prices.
• Integration. It smelts and refines much of what it mines. In the nine months, smelter output rose 15.8% and refinery output 10.1%.³ That captures margin on its own concentrate and sells cathodes and rod.
• By-products. Molybdenum, silver and zinc paid for US$3.47B of costs in the nine months, which is why the net cost was US$17M, or US$0.01 a pound.³ Molybdenum alone was US$1,491.7M of sales.⁵
• Balance sheet. With US$8.06B in cash and short-term investments against US$8.00B of debt, it is not a forced seller and it pays for its own projects.³
• The take. It is high and fairly steady. Income tax was 36.1% of pretax profit in the nine months, workers’ profit share another 5.1% of operating profit before it, and royalties sit inside both C (calculated).³

Where it is thin

The first weakness is the price taker’s: the same US$0.10 that adds US$120M of net income takes it away. The company gives that sensitivity itself, for 2026 volumes and 2025 tax rates.⁴ In 2016, with copper at US$2.21, net income was US$776.5M; in the nine months to September 2026, with copper at US$6.09, it was US$5,048.6M C (calculated).²³ Net income rose about 6.5 times, nine months against a full year, and the business did not change by that much. The price did.

The second is the cost line. Before by-product revenue, the company’s operating cash cost per pound went from US$1.44 in 2016 to US$2.17 in 2025 and US$2.35 in the nine months of 2026, up 63% C (calculated).¹⁶⁴³ Net cost only looks flat because by-products moved the other way: they paid US$0.49 a pound in 2016 and US$2.34 in the 2026 nine months.¹⁶³ If molybdenum went from US$29 back to US$18 and silver from US$73 to US$32, that credit would fall from US$2.34 to about US$1.50 a pound on my numbers, and the company-defined net cost would rise from US$0.01 to about US$0.90 E (estimate).

The third is the ground itself. Production in Peru is falling on lower grades, and the whole company is mining less copper this year than last. The reserves are very large, but ore at depth is thinner, and the growth that matters comes from new pits.³ The fourth is that the ground is in two countries whose governments decide whether new pits open. That has its own section below. Last, the dividend structure and the owner make the minority’s claim weaker than the reserve figure suggests. That is covered under Management.

I read the protection as real but narrow. A new entrant can’t copy the ore body, and the one rival I compared has a higher cost before by-products. A low cost doesn’t protect against a lower price, though, and at US$4.70 a pound in my base, the business earns well but not exceptionally: operating cash flow of about US$5.2B in 2031 against a market value of US$151B E (estimate).

IIIThe competitive advantage

The edge I can show is cost and reserves, set against a rival that reports differently. I have only one clean comparison, Freeport-McMoRan, and I opened only its third-quarter release.

Southern Copper’s cost before by-products was US$2.46 a pound in the third quarter. Freeport’s site production and delivery cost was US$3.24.³⁹ The gap is 24%, but the definitions differ: Freeport’s figure includes items Southern Copper leaves out of its cost, such as royalties and purchased concentrates. Net of by-products the comparison flips the other way in Southern Copper’s favour: US$0.09 against US$2.07. Both are real. They answer different questions. The gross figure says what it costs to move the rock. The net figure says what is left after molybdenum and silver pay their share, and they are paying a lot this year.

Operating cash flow less capital spending was US$4.53B in the nine months for Southern Copper against US$2.29B for Freeport, on a market value I did not check for Freeport C (calculated).³⁹ Southern Copper spent 27% of net income on capital in the nine months.³ Freeport spent about 59% of its operating cash flow C (calculated).⁹ One reading is that Southern Copper has spent less because it has had less to spend on: its growth projects are slower. The other is that it is lean. The evidence is thin, so I take neither.

Three things Southern Copper does that I can see

It keeps its capital spending in check until a project is ready. Capital was near US$1B a year from 2016 to 2024, apart from 2019 to 2020 at US$592M to US$708M, and US$1,325M in 2025.² It is US$1,380M in nine months of 2026, a rise of 52.9%, because Tia Maria is now under construction.³

It pays out a lot. Cash dividends were US$2,563M in the nine months, 51% of net income and 57% of operating cash flow less capital spending C (calculated).³ That is the company’s own policy. It also goes mostly to the parent: 88.9% of each dividend.

It borrows in New York to build in Peru. The June 2026 notes, US$1.25B at 5.350% to 2036, were sold to fund Tia Maria and Peruvian capital spending, with BofA Securities, Morgan Stanley, Barclays and Santander as underwriters.¹⁸ The 10-year Treasury yield was 4.90% on 4 November.⁷ I did not compare the spread with Freeport’s debt.

The risk in the partners

The main mines are owned by the company and its subsidiaries, with no operating partner I could find, so the risks sit elsewhere: the communities around Tia Maria, the courts, the Mexican state power company that has to connect El Arco, and the group that owns 88.9%. It buys power, rail, construction and ore from companies that Grupo Mexico also owns. Those payments were about US$477M in 2025 C (calculated).⁶ Section IV covers that.

Rivals in waiting

A new rival can’t build 956,000 tonnes a year in a few years, but the field is not static. Goldman says the Grasberg mine in Indonesia and Kamoa-Kakula in Congo, both hit by major incidents last year, won’t be at full capacity before 2028; Freeport says Grasberg’s ramp-up is meeting expectations, with significant growth expected in 2027 and 2028.¹⁹⁹ When those return, supply rises, and prices lose some of the shortage premium they now carry. Ore bodies take a decade to open, which is the industry’s protection, and the reason Southern Copper’s growth list matters.

IVManagement

The people who run Southern Copper answer to one family. Germán Larrea Mota-Velasco is chairman of the board. He is shown as the beneficial owner of 1,508,070,283 shares of Grupo Mexico, 19.03% of that company, and the Larrea family controls the holding company that is Grupo Mexico’s largest shareholder and a majority of its votes.⁶ Grupo Mexico owns 88.9% of Southern Copper through Americas Mining Corporation.⁶ So the question for an owner here is as much what the controlling family wants as what the managers do.

The people

Oscar González Rocha was president and chief executive for years and a director, and he died unexpectedly on 7 April 2026. The company called his 50-year career defining.¹⁰¹¹ On 23 April the board named Leonardo Contreras Lerdo de Tejada chief executive. He is 40, has been a director since May 2021, joined Americas Mining in September 2018, was president of Asarco in 2019, president of IMMSA from 2020, chief financial officer of Americas Mining from January 2022 and its general director from April 2024. Before that he founded a private investment vehicle, Murano Capital, in 2015, after private equity and investment banking. He studied industrial engineering at Universidad Anahuac and did an MBA at Chicago Booth. The proxy states that he is the son-in-law of Mr. Larrea.⁶¹¹ The board made no change to his pay.¹¹

I can’t judge him. He has run the group’s underground mines, so he knows part of the operation, and he has been inside the parent for eight years. The proxy lists no role of his in Peru, which is 43% of the copper, and his appointment came 16 days after the death of the man who ran it. A successor chosen from the family will not be a stranger to the chairman. It is also a succession that nobody outside the family voted on.

How they are paid, and what that pays them to do

The disclosed pay is low and not tied to the share price. Mr. González Rocha’s 2025 total was US$1,298,630: base salary US$458,171, a discretionary cash bonus of US$103,089, and US$386,526 as his share of the 8% of Peruvian pretax profit that Peruvian law requires the company to pay out to its workers, plus holiday and other bonuses required by law.⁶ The CEO-to-median-employee pay ratio was 27.49 to 1; the median employee earned US$47,246.⁶ There are no stock options and no equity awards for the named officers; the plan expired in 2006 and none was granted to them after 2000.⁶ Officers can join a Grupo Mexico stock plan that buys shares of the parent at fair value through payroll deductions.⁶

The incentive, then, is a salary, a bonus the committee sets by discretion and a profit share that grows with profit in Peru. Nothing pays a manager more if the shares of Southern Copper do well, or less if they do badly. In a business where the owner is the boss, that is not strange. It does mean that what management does is steered by one holder’s priorities and not by a pay formula I can read. The compensation committee in 2025 was Mr. Larrea, the late Mr. González Rocha and others, and the company relies on the controlled-company exemptions that waive independent compensation and nominating committees.⁶

How much of their own money is in

Directly, not much. All 17 directors and officers held 602,270 shares on 2 April, worth about US$106M at the price, and Mr. Larrea held 404,126.⁶ The new chief executive held 600.⁶ The money that matters is held through Grupo Mexico: looking through, Mr. Larrea’s reported 19.03% of Grupo Mexico is about 17% of Southern Copper, or about US$25B at the price C (calculated). That is a very large stake, held at one remove, and it makes his interests those of an owner of 17% (and much more with the family’s other holdings), not those of a manager with a bonus.

What they did with capital over a full cycle

The ten years to 2025 included the low of 2016 and the 2021 and 2025 highs. Operating cash flow came to US$29.7B, capital investments to US$9.8B and cash dividends to US$16.5B C (calculated).² Net debt went from US$5.4B at the end of 2016 to US$1.8B at the end of 2025 and then to a small net cash position at 30 September 2026 C (calculated).²³

My reading is that this team returns cash and builds slowly. Capital was held near US$1B a year for a decade, even as the ore got worse, and the growth list is long on paper and short on construction. The cost of that caution was missed production: copper produced was 2,036 million pounds in 2025, below 2019’s 2,133 million. The credit is that cash did not go into deals at peak prices. The doubt is that the same few people decide what is a growth project and how the proceeds are shared.

The controlling holder, and what it means for minority holders

Grupo Mexico’s stake is 750.6 million of the 844.3 million shares outstanding, about US$133B at the price C (calculated). The other 11.1%, about 94 million shares and US$17B, is held by the public.⁶⁵ This is what that structure means in practice.

• The control can’t be contested. The company is a controlled company under NYSE rules, so it is exempt from having a majority of independent directors or independent compensation and nominating committees.⁶ Its charter does require special independent directors, three of the eight, chosen by a committee of two special designees and one board designee, and the board added three more independent members.⁶ But the 2026 vote shows the limit. Germán Larrea got 764.0 million votes for and 25.1 million withheld; Leonardo Contreras got 761.1 million for and 27.9 million withheld.²² If Grupo Mexico voted all its roughly 734 million eligible shares in favour, which is my assumption, about 46% of the minority shares that voted withheld from Mr. Larrea and 51% from Mr. Contreras E (estimate). They lost the vote, and nothing in how the vote works lets them win it.
• Cash goes mostly up the chain. Of every US$1.30 cash dividend, 88.9% goes to the parent. The parent’s interest is in cash, so the payout is high, and the minority gets pro rata. That is a fair outcome, and the minority has no say in the amount.
• The group is also a supplier. Southern Copper paid companies in the Grupo Mexico group about US$477M in 2025, mostly for power, ore, construction and rail, and received about US$102M.⁶ The proxy says these transactions are reviewed and approved by senior management, as all related-party transactions are; I found no separate vote by the independent directors in the text I read.⁶ I can’t say whether the terms were at market. A related party is not a bad partner, but it is a price I can’t check.
• There is a precedent. In 2005 the company bought Minera Mexico from the parent. On 9 October 2012 it received US$2.1B from the parent in satisfaction of a Delaware Court of Chancery judgment that it had paid an excessive price in that merger.²⁰ A related-party deal was challenged, and the parent paid the company; the 2015 10-K cash-flow table lists the same US$2,108.2M on a line called "SCC shareholder derivative lawsuit".²¹ It is also a record of what the parent was willing to do in 2005.
• Any sale of the company is the parent’s decision. Nobody else can bid for control. If Grupo Mexico wanted the 11.1% it would pay about US$17B at the current price, about 5.7 years of the cash dividends it receives at the current rate C (calculated). I am not predicting that. It is a fact about the exits: they belong to one holder.

There is a second reading of the same facts, and it favours the minority. The parent is also the owner of nearly nine-tenths of the profit. It gains most from a rising value and from cash paid out. Its dividend policy has been steady and large. A holder who sits beside a patient controlling owner usually does fine if the owner’s interests and his own run the same way, and on dividends they do.

Grupo Mexico’s own shares give a different view of the price. Its Mexican-listed B shares closed at MXN 225.21, or US$11.97 at 18.8175 pesos to the dollar, and the 19.03% in the proxy implies about 7.92 billion shares and a market value near US$95B C (calculated).²³⁶ Its stake in Southern Copper alone is worth about US$133B, so the market values the parent at 29% less than that one holding, before its railway, its other mines and its debt C (calculated). I have not opened Grupo Mexico’s filings, so I can’t say what is in the gap. The pattern is clear, though: the public pays more for the minority piece of Southern Copper than for the same copper inside its parent. I do not put a discount for governance into my values. If you want one, take 10% off my base and it is $57.

VFinancials

The statements are simple to read and hard to compare across years, because the copper price swings every line. Three things need explaining before the table.

• Revenue includes provisional prices. Copper, molybdenum and zinc sales are booked at an estimated price and adjusted to the average price of a later month, so a quarter’s revenue includes the final price on earlier sales.⁵
• Cost per pound is the company’s own measure. It leaves out workers’ profit share, royalties, tax and some purchased concentrate. I show it because the company reports it each year, not because it is the full cost.
• Per-share figures move with stock dividends. Earnings per share is restated each time shares are paid out. I show net income and shares separately instead.

The third quarter

Sales were a record US$4,696.6M, up 39.1%, on copper at US$6.41 LME. Net income was US$1,801.8M, up 62.7%, adjusted EBITDA US$3,040.2M with a margin of 64.7%, and operating cash flow US$2,224.3M, up 42.6%.³ Mined copper was 232,396 tonnes, down 1.1% on a year earlier; mined silver, zinc and molybdenum were lower too.³ Operating cash cost before by-products was US$2.46 a pound and US$0.09 net.³ Capital investments were US$515.8M, 28.6% of net income.³ The tax charge was US$990.9M on pretax profit of US$2,780.1M, or 35.6%.³ The company says the world copper market will be in slight surplus in 2026, with 1,105,000 tonnes in warehouses on 9 October, about 14 days of use.⁵ That is the company’s own view and it is cooler than the price.

What an owner can really take out

Operating cash flow is real cash, and no stock pay needs to come off it: the officers have no equity awards, and the stock dividend is a payment to all holders, not to staff.⁶ What does come off is the capital needed to stand still. A mine at 2,000 million pounds a year loses ore every year, and replacing equipment, tailings capacity and stripping costs about what the company spent in 2024, US$1,027M, when growth projects were small.² I use US$1.0B in 2027 money, rising 2.5% a year E (estimate).

On the last twelve months that leaves about US$6.4B of owner cash, which is 4.2% of market value E (estimate). That is about the same as net income of US$6.4B, because capital spending is close to depreciation. The number hides the key fact: it is earned at copper near US$6 a pound. At my base long-run price of US$4.70 in 2031 dollars the same business makes about US$4.1B of owner cash in 2031, 36% less E (estimate). So profit earned at today’s prices overstates what an owner can take out in a normal year by more than a third.

Balance sheet

At 30 September the company had US$4,823.7M of cash and US$3,239.2M of short-term investments, up from US$604.6M at the end of 2025 because it sold the new notes and invested the proceeds.³ Long-term debt was US$7,995.9M, lease liabilities US$512.6M and the asset retirement obligation US$495.8M.³ Income taxes payable were US$1,015.5M, because of the high profit.³ Equity was US$13,516.9M. After the cash dividend of about US$1.1B due on 19 November and my estimate of fourth-quarter cash flow, I count net cash and leases at about nil at the end of 2026.³ I treat the asset retirement obligation and the Peruvian labor-share claim as small against a 50-year reserve life, and I test the second in the risks.

How the shares have traded

The shares closed at US$176.84 in Q4 2026. In November 2016 they were US$25.87 on the same adjusted basis, so they rose 6.8 times in ten years, with cash dividends on top C (calculated).¹²¹ Copper futures rose 2.8 times C (calculated).⁸ The shares did more than the metal. Profit moves faster than the price when costs are fixed, and the multiple paid on profit also matters: at the price, the shares are on 23.7 times the last twelve months of net income C (calculated). The 52-week range is US$114.47 to US$218.46.¹ At US$176.84 the price is 19% under that high and 54% over the low.

VIValuation

First, how this goes badly wrong

Before the numbers, I asked what would most likely ruin this investment, and worked back. Three things stand out.

• Copper reverts. Copper averaged US$2.21 in 2016, US$4.51 in 2025 and US$6.09 in the first nine months of 2026, and the front-month futures price set a record close of US$6.80 on 9 September.⁴³⁸ The 2026 average is 47% above 2024. Goldman’s own "fundamental fair value" is US$12,800 a tonne, US$5.81 a pound, and it says a no-tariff outcome could flip 2027 into surplus.²⁴ The company itself calls 2026 a slight surplus.⁵ If copper settles at US$3.40, my bear, the value is $22. Molybdenum and silver would fall with it.
• Peru or Mexico changes the rules. Peru has had six governments in little over a decade.¹⁷ A tax rise, a stalled permit, a blockade or a court order falls on assets that cannot move. Mexico has changed its mining law, and the state power company has to link El Arco to the grid before it can start.³⁵ I put numbers on these in Section VIII.
• The ground gets worse and the owner takes the cash. Grades are falling in Peru, output is down this year, and net cost looks lower than it is because by-products are priced at records. And the controlling holder decides what is spent and paid. Cash costs growing 3.5% a year instead of 2.5% cut my base to $61.

None of these needs a surprise. No full year before 2026 averaged above US$4.51, and the price of the share is built on the first nine months of this one. They set my bear case, and they are why I start from the cash an owner can take out and not from a multiple of EBITDA.

How I built it

I model the mines as one business with a copper price deck, a by-product deck and a volume path. The price deck is the heart of it, so I show it first. Today copper is US$6.40 on COMEX, and the LME average was US$6.41 in the third quarter.⁸³ Bank forecasts are well above my base: Goldman has US$13,800 a tonne for 2027, US$6.26 a pound; Citi has said US$15,000 within a year; BMI has US$13,500 in 2028, US$15,000 in 2029 and US$15,800 in 2030, against Bloomberg consensus of US$12,500, US$13,772 and US$13,500.²⁴¹⁹²⁵ I do not use them for the long run. Forecasters have followed spot all year, and I treat that as information about the next two years and not about the next twenty. My base starts at US$6.00 in 2027 and slides to US$5.10 in 2030, which is at or under consensus for those years, then settles at US$4.70 in 2031 dollars and rises 2% a year with costs. That long-run figure is above every annual average up to 2024 in the dollars of the day; with 2% a year of inflation taken out it is about US$4.34 in 2027 dollars, 4% under the 2025 average of US$4.51.⁴

Volumes and by-products. Copper produced starts at 1,965 million pounds in 2027, drifts down 0.3% a year as Peruvian grades fall, and adds Tia Maria (265 million pounds, full in 2031), El Pilar (79), the Cuajone expansion (121) and Los Chancas (287, full in 2036), from the company’s own project figures.³ Molybdenum is 64 million pounds, zinc 360 and silver 23 million ounces, in line with the 2026 plan, scaled gently with copper.⁵ Other sales (mostly sulfuric acid) are US$630M a year, growing 2%.

Costs. I took all cash costs before depreciation, workers’ profit share and tax, including administration and exploration: US$4,215M in the nine months, or US$2.84 a pound produced, up from US$2.44 a year before C (calculated).³ I split that into US$2.52 a pound that grows with inflation and 3.5% of revenue that moves with price, to reflect purchased concentrate and royalties. New SX-EW and project pounds cost a quarter less. Profit share is 5.05% of operating profit, and tax and royalties 36.1% of what is left, the nine-month rates.³ Working capital is 12% of any change in revenue. Capital to stand still is US$1.0B in 2026 money, rising 2.5% a year, plus the growth projects above.

The life. The model runs year by year until the copper in the reserve is used up, or to 2071. Reserves are 108,955 million pounds contained, at a US$3.30 price; I count 85% as recoverable, and 1.25 times that in my base because US$4.70 is above US$3.30. The base mines for 45 years.⁴ At 9% anything after 2050 is worth almost nothing, so the life matters little to the value and a lot to what the company is.

The required return. I use 9%: the 10-year Treasury at 4.90% plus 3.5%, plus 0.6 points for two countries where the rules can change.⁷ That compares with 8.5% on my Franco-Nevada and Costco pages and 10.0% on Cameco. A rate of 7% lifts the base to $81; the grid below shows the rest. No rate I can defend reaches the price. The value is each year’s free cash flow at year end, discounted from 1 January 2027, with my estimate of net cash and leases at the end of 2026 added.

What the price implies

At $176.84 the market value is US$150.9B, which is 23.7 times the last twelve months’ net income and 20.4 times operating cash flow C (calculated).¹²³ Enterprise value is about 13.8 times operating income plus depreciation C (calculated). Free cash flow was 3.7% of market value. These are the figures of a business at a record copper price. The question is what the price needs. I solved it several ways.

This is the table to read. To reach $176.84 on my base volumes and costs, copper has to settle at US$11.70 a pound for good, 1.7 times the September record. If copper held today’s US$6.40 for the whole life of the reserve, the value would be $73; if it rose 2% a year as well, $100; and if my bull volumes and cost growth are added, $128. Still short of the price. Or the required return has to fall to 2.6%, below the 4.90% on a Treasury bond. The market may have reasons. It may expect electrification and AI to keep copper in deficit, since banks forecast US$13,000 to US$15,800 a tonne through 2030.²⁴²⁵ It may expect the company to reach 1.6 million tonnes, which would roughly double my base volume gain. It may give weight to 108,955 million pounds of reserves that could last 50 years. And it may price the free float of 11% as scarce. All of these are possible. I can’t put a number on the last, and I did not tune any input to reach the price.

The market’s case does not need one extreme input. It needs several moderate ones together. Hold copper at today’s US$6.40 and let it rise 2% a year, give the company my bull volumes and cost growth, and accept a 7% return instead of 9%, and the value is $172, close to $176.84. Each step is one a thoughtful buyer could take. My base differs on every one, and I think a 9% return is fair for two countries where the rules can change, but 8% is not unreasonable and gives $71 at my base. Near-term bank forecasts are not the issue: if I use Bloomberg consensus for 2028 to 2030 (US$5.67, US$6.25 and US$6.12), my base moves only from $63 to $65, because the years after 2030 carry the value.

What I can’t judge

Three inputs decide this page and I can’t know any of them. I don’t know where copper will trade in 2032. I don’t know what the next pit will cost or whether it will be allowed. I don’t know what the controlling holder will do with the cash. So here is the grid. Read my $63 as one cell in it.

A US$0.50 move in long-run copper is worth about US$8 a share at the middle of the grid. A 1-point change in return is worth about US$5 to US$9. The first is a bet on a price. The second is a bet on what an owner is willing to accept. They can move together: if rates fall, metals tend to rise. That is the strongest version of the market’s case, and I can’t rule it out.

Other inputs

Three cases

Other uses of the same dollar

The ten-year Treasury pays 4.90% with no risk to the money R (reported).⁷ My base returns −4.4% a year at this price over ten years, my bull 4.0% and my bear loses 16.0% a year E (estimate). On this site the gap between price and base is 181% here, against 97% for Cameco, 73% for Franco-Nevada and 39% for Exxon Mobil. Different commodities, different cases. It is still the same test, and Southern Copper is furthest from passing it. The parent offers another route to the same copper: Grupo Mexico’s market value, about US$95B on my arithmetic, is 29% below the value of its Southern Copper stake alone. I have not valued Grupo Mexico and it carries a railway, other mines and its own debt, so I can say only that it costs less per dollar of Southern Copper held.²³⁶

The margin of safety

• Price$176.84Close, Q4 2026 [1]
• Base value$63My estimate E (estimate)
• Margin of safety below$42a third under base E (estimate)

I set the margin of safety at a third under the base, about 67% of it, close to the 70% on my Costco and Exxon Mobil pages. At $42 the base is 50% above the price paid, and my US$22 bear value is 48% below it. So it is protection against being wrong by a third, and not against the bear case: if copper fell to my bear deck, a holder at $42 would still lose about half. The shares were last at or below $42 on 30 September 2022, when copper was near US$3.40; they were last at or below my base of $63 on 31 May 2023 C (calculated).¹² I don’t expect either and I’m not forecasting them.

Two different questions

How good is the business, and how good is the price? On the first, I find a low-cost producer with the largest reserve I can compare, a net cash balance sheet, a 38% net margin at spot and a dividend that is mostly cash. It is a good business by any test I know, with two conditions: copper must stay above about US$3.50 and two governments must keep the rules. On the second, at $176.84 the price is above my bull case. The cash an owner can take out is 4.2% of the price at spot and about 2.7% in a normal year E (estimate). A good business at a price that needs copper at US$12 is a different thing from a good business. The first does not decide the second. Which of these this is, I leave to the reader to decide from the numbers above.

Where this could be in 5 and 10 years

Value at a date is the present value then, at 9%, of the free cash flow after it, per today’s 853.3M shares, in that year’s dollars. It treats each year’s free cash flow as paid out, and today’s net cash as paid in 2027. The return is the yearly rate that turns today’s price into the 2036 value plus the cash paid on the way. A holder who keeps the stock dividends owns more shares, each worth proportionally less, so I leave them out.

Why, in plain words. The base is worth US$66 in five years because the company is still paying out about US$4B a year while building Tia Maria and Los Chancas, and about US$72 in ten. At today’s price a holder gets back about US$50 of cash and US$72 of value in ten years, US$122 in all against $176.84. Even my bull returns 4.0% a year, below the 4.90% on a Treasury bond. For the price to be earned back at 9% a year, the shares would need to reach about US$343 in 2036 on my base cash flows C (calculated).

The inputs that drive it are the copper price, the cost line and the required return. The most likely way my base is wrong is copper. Each US$0.50 on the long-run price is worth about US$8 a share, and at US$6.29 in 2031 dollars, which carries Goldman’s "fundamental fair value" of US$5.81 for 2027 forward at 2% a year, the base is $89.²⁴ I don’t know where copper will be in 2032.

VIIThe case for the price

What most people miss

Three things cut in the company’s favour. First, by-products pay for the mine: molybdenum, silver, zinc and other products brought in US$3,467.7M in the nine months against an operating cash cost of US$3,484.6M before them, so the copper cost US$17M net.³ Second, the money for growth is already in the bank: US$8.06B of cash and short-term investments against a US$1.8B Tia Maria project, US$1,101M of it committed.³¹⁷ Third, the reserve is priced low. It is estimated at US$3.30 a pound, and the company’s impairment test at US$2.80 copper shows cash flows 2.8 to 7.5 times book value.⁴ I credit a little of the upside by letting my base mine 1.25 times the reserve, which adds about US$2 a share.

The bull case

In plain words: copper is short, demand is rising, and Southern Copper has what it says is the largest reserve in the industry, a cost below Freeport’s, no net debt and a funded plan to go from 918,000 to 1.6 million tonnes. In the third-quarter release the chairman, Germán Larrea, called the results records and said the company was advancing toward 1.6 million tonnes.³ Banks are bullish on copper, not on this company, and I have these only as press reports: Goldman had US$13,800 a tonne for 2027 on 8 June (and called US$12,800 fair value), Citi US$15,000 within a year on 2 June, BMI US$15,800 in 2030 on 16 July.²⁴¹⁹²⁵ I found no dated analyst note on Southern Copper that I could open, and I did not check anyone’s holdings.

Against reality

Sales, profit and cash flow are records and the balance sheet is strong.³ Volume is not there: copper mined is down 4% this year, Tia Maria is about half built, and no larger project has permits and funding together.³ If copper holds US$6.50 for good, Tia Maria, El Pilar, Los Chancas and Michiquillay are built, and costs rise 2% a year, my bull value is $121. Add today’s US$6.40 rising 2% a year and it is $128. The price is $176.84. Multiply it out, with odds that are only my guesses: copper at US$6.50 for good, 15%; 1.4 million tonnes by 2036, 20%; costs up only 2% a year when they rose 16% this year, 40%. That is about 1% if independent E (estimate), and I say 3% because they move together. The first guess is the one the market would dispute most. Even the bull value falls short of the price, and only spot copper held for good, bull volumes and a 7% return together come close ($172).

What a 10x would take. Ten times $176.84 is $1,768, about US$1.51 trillion. At today’s 20.4 times operating cash flow that needs about US$74B a year against US$7.4B now, and sales near US$171B C (calculated). At the company’s 1.6 million tonnes that is US$49 of sales per pound, or copper near US$35, more than five times today’s price; at today’s price it is ten times today’s output, 9.2 million tonnes. The shares rose 6.8 times in ten years from a company one-seventh the size, with copper up 2.8 times.¹²⁸ I know of no mining company that has come near a US$1.5 trillion value, and I did not search systematically. It is close to impossible.

Where I come down

The balance sheet, the reserve and by-product revenue that wipes out the net cost moved me; without them my bear case would be below $22. The price did not move me, because it needs copper far above my base for decades, or spot copper, bull volumes and a 7% return all at once. Permits for Los Chancas and Michiquillay and copper above US$6.50 for years would raise my values most, though the price would still need roughly twice that.

VIIIRisks, and what would change my view

The three ways this goes badly wrong are at the start of Valuation. These are the rest.

• The price of copper. The company’s sensitivity is US$120M of net income for each US$0.10 a pound at 2026 volumes.⁴ From US$6.09 to my base long run of US$4.70 is US$1.39, or about US$1.7B of net income a year C (calculated). Tariffs are the swing: Goldman says a tariff in January 2027 would pull imports forward and a clear no-tariff outcome could send copper toward US$12,800 a tonne.²⁴
• Grades and volume. Mined copper is down 2.9% in nine months because Peruvian ore grades are lower, and 2026 is guided 4% under 2025.³⁵ The base needs the new pits to more than replace that.
• Costs. Operating cash cost before by-products rose 63% in ten years, and total cash opex per pound produced rose from US$2.44 to US$2.84 in a year C (calculated).³ A large share of costs is in Peruvian soles and Mexican pesos, so local inflation or a stronger currency raises the dollar cost; the proxy shows Peruvian executive salaries up 6% in soles but 11.5% in dollars in 2025 for that reason.⁴⁶
• The owner. Grupo Mexico decides the dividend, the capital plan, the related-party contracts and who runs the company; the new chief executive is the chairman’s son-in-law and has had the job since April.⁶¹¹ A holder who disagrees has no remedy except to sell.
• The share supply. The treasury holds about 40 million shares and the stock dividend uses about 9 million a quarter. After that the company must issue new shares or stop.⁴⁵ Nothing says which, and the result for a holder is neutral to a split or a cash cut.
• Money and ground. The Buenavista spill, the Guaymas terminal, the tailings dam claim and the labor-share lien are open. I put no number on them.⁵

Peru and Mexico: permitting and political risk, and what it means for the value

All of the copper is in two countries, 43% of it in Peru and 57% in Mexico in 2025, and the reserves split about the same way: 45,527 million pounds in Peru, 44,154 million in Mexican open pits, and 19,274 million in development projects in both.⁴ Here is what I found, in the order of the risk.

Peru. Keiko Fujimori took office on 28 July 2026 after a two-round election; her plan, as SMM read it on 8 July, speeds up permits, offers tax incentives and does not raise mining taxes.²⁶ At the time of that report no mining minister had been named, and I did not find a later report on her government or its minister. Peru has had six governments in little over a decade, and an estimated US$7 billion of copper projects are stalled by illegal mining.¹⁷ For Southern Copper the sharpest items are specific. Tia Maria: seven lawsuits were filed; two are closed; the Supreme Court ruled in 2022 that the environmental approval was lawful; the Ministry of Energy and Mines reviewed the permit in April and reauthorized it on about 20 April.⁵¹⁷ Los Chancas: the company says it is working to regain control of the project area from illegal miners.³ The tailings dam: since 2012 a regional project has sought the demolition of the dam used for tailings from Toquepala and Cuajone since 1995; the company says the claim has no merit and the amount cannot be estimated.⁵ Labor shares: a court-ordered lien covers 10,501,857 investment shares and the Peruvian branch has challenged the value; the case is suspended.⁵ None of these is in my base as a loss. They are the kind of event that falls on assets that cannot move.

Mexico. In May 2023 Congress cut mining concession terms from 50 to 30 years, restricted water use, required closure guarantees and added a 5% contribution to indigenous communities on new projects; the changes are before the Supreme Court and the company says it expects no negative effect.⁵ In February the government asked for the voluntary return of more than 200 concessions that produce nothing.²⁷ The company is in talks with the administration about US$10.2B of investments, including El Arco, which also depends on the state power company linking Baja California to the national grid.³ El Pilar has its permits.³ The 2014 Buenavista spill still has criminal complaints, three collective actions and dozens of civil suits pending, and a partial shutdown of the acid terminal at Guaymas dates from 2019.⁵ The Mexican mining royalty is 7.5% of earnings before tax, depreciation and interest, and it was US$137.9M in 2025.⁴

I priced each as a separate cut on my base case, with my guess at the odds. The 0.6 points I add to the required return is the general allowance for this kind of risk. These are named events on top.

Weighted by my odds, the five named events cost about $2 a share, or 4% of the base E (estimate). I could double every probability and still be under $5. That is the point of this table. Peru and Mexico can hurt, and they are real. They are not what separates $63 from $176.84. The price carries the copper bet, and these risks come on top of it. At a price that already assumes the best, any one of them is a loss, and none of them is in the price as a discount.