Franco-Nevada

IThe business

Franco-Nevada doesn’t mine anything. It pays a miner money today and gets, in return, a slice of what the mine sells for as long as the mine runs. There are two main forms. A royalty is a percentage of a mine’s revenue or profit. A stream is the right to buy a fixed share of the gold or silver at a low price set in advance, often 20% of the spot price, and resell it at the market price.¹⁴ Either way, the operator carries the cost of digging, power, wages and steel, and Franco-Nevada gets paid. At 11 August 2026 it held 447 interests: 257 in precious metals, 107 in other minerals and 83 in oil and gas. Of those, 121 are producing, 47 are at an advanced stage, and 279 are exploration.³

Company figures are in US dollars, as the company reports them, under IFRS. Per-share values are in US dollars against the NYSE close. The Toronto close was C$314.41; at the Bank of Canada rate of 1.3960 that is US$225.22, within 0.2% of the New York close.¹¹⁵¹⁶ GEO means gold equivalent ounce: silver, platinum, oil and the rest are turned into ounces of gold by dividing their revenue by a gold price, so ounces from different metals can be added. TTM means the twelve months to 30 June 2026. The third-quarter report is due on 11 November, after the date limit of this page.¹⁷

In the second quarter, revenue was US$580.9M, up 57%, on 132,405 GEOs, up 18%. Precious metals were 86% of it: 70% gold, 14% silver and 2% platinum group metals.⁸ By place, 40% came from South America, 25% from Canada, 16% from the United States, 7% from Central America and Mexico and the rest from elsewhere.⁸ By type, streams brought in US$287.1M, revenue royalties US$211.5M, and profit-based royalties US$70.5M.³ In the first half, 88% of revenue came from royalties and streams that pay on production with no adjustment for what it cost the operator to dig it, and the rest came from interests that pay on profit.³

The size of the machine

The company had 44 full-time employees and five part-time contractors in March.⁴ Revenue in 2025 was US$1.82B, which is about US$41M per employee C (calculated). General and administrative expense, including stock pay, was US$19.7M in the first half, which is 1.6% of revenue.³ Cash costs, the ongoing payments it makes under streams, were US$92.4M in the half, or US$344 a GEO.³ That is why the adjusted EBITDA margin was 91.1%, and 90.9% for all of 2025.³² A business like this has almost nothing to buy and almost no one to pay.

What it earns on the money invested

The interests cost US$11.6B in total, before US$5.6B of depletion, and carry a net book value of US$6.0B.¹⁴ That cost includes Cobre Panamá, which was written down by US$1,169M in 2023 and sold only 11,208 GEOs in 2025.¹⁸² 2025 adjusted EBITDA of US$1,656M is 14.3% of the gross cost, at an average gold price of US$3,435 C (calculated).²¹⁴ On the first half of 2026, at about US$4,600, the same measure would be near 19% C (calculated). These are pre-tax returns on what the company paid, which is a useful picture of what a dollar spent on royalties has brought back. It is also a picture that moves with the gold price almost one for one.

What is coming from the existing book

The company’s 2026 guidance is 510,000 to 570,000 GEOs, and it says it is tracking the upper half, helped by processing of stockpiles at Cobre Panamá.⁸ That guidance assumes gold at US$4,500, silver US$75, platinum US$2,000, palladium US$1,650, iron ore US$100 a tonne, WTI oil US$70 a barrel and Henry Hub gas US$3.00.⁸ Gold futures closed at US$3,899 on 30 October, below the US$4,500 in the guidance.⁶ Oil is the other way: second-quarter WTI averaged US$92.79 against the US$70 assumed, and the company says each US$10 on its assumption adds about 12% to oil revenue.³⁸

For 2030 the company expects 555,000 to 615,000 GEOs at the same prices, with nothing from Cobre Panamá and with no acquisitions to come.² Two known steps down fall inside that window. Candelaria’s stream falls from 68% to 40% of gold and silver in the second half of 2027, and Antapaccay’s moves from copper-indexed to 30% of production in the second half of 2028.² Against them: Côté Gold and Porcupine, bought in 2025, are in their first full years, Valentine and Greenstone are ramping, and Cascabel in 2028, Eskay Creek in 2027, Stibnite in 2029, Castle Mountain and a lift at Detour Lake to 1 million ounces a year by 2031 are in the company’s plan.²⁸ The portfolio also holds more than 300 exploration and development royalties that the guidance does not count.² That is the right to share in any discovery, with no further payment.

Oil, gas and iron ore

Diversified assets earned US$164.8M in the first half. Energy was US$129.8M of that: US$89.2M from United States interests in the Marcellus, Haynesville, SCOOP/STACK and Permian, and US$28.4M from the Weyburn unit in Saskatchewan, with the rest from other Canadian interests.³ Iron ore royalties from Vale and Labrador Iron Ore were US$26.2M.⁸ This is about 13% of revenue, so it matters less than gold. It does mean the stock has some exposure to the oil price. In the second quarter Edmonton Light crude averaged C$131.28 a barrel, 53% above a year earlier, and the Iran war was cited as a reason.³

I have seen two kinds of mistake with a company like this. One is to treat it as a gold bar with a ticker: it isn’t, because its ounces change with operators’ decisions, and the price you pay for each new ounce is set in a market. The other is to treat it as a growth business: it isn’t, because what it buys is mostly someone else’s depleting mine. It is closest to a gold-linked bond with a share in the upside, plus a standing offer to buy more.

IIThe moat

What keeps others from taking the profit? For a royalty company the question is odd, because there’s no product to defend. What I look for instead is whether the cash that comes in can be earned again at a good price, and whether anything stops a rival from doing the same.

What protects it

• Cost structure. Eighty-eight per cent of first-half revenue came from interests that pay on production, with no deduction for the operator’s cost.³ When a mine’s power bill or wages rise, the operator’s margin shrinks and Franco-Nevada’s doesn’t. Where a stream has a fixed price per ounce, such as the US$418.27 on part of Cobre Panamá, that price is only adjusted for inflation.¹⁴
• No capital calls. The company says it has no capital requirements beyond the initial commitment and limited cash calls on a few working interests.³ It pays once, and the mine owner builds.
• The book is wide. Of 121 producing interests, no one asset was above 11.3% of first-half revenue C (calculated).³ One accident, one tax change or one flood at a mine hits one slice.
• A balance sheet that miners don’t have. There is no debt, US$1.0B of cash at 30 June and US$4.3B of available capital, including shares in other companies and US$1.5B of revolving credit with a US$750M accordion.³²² When gold falls and a miner’s lenders pull back, that money is worth more. In 2025 the company deployed US$1.8B.¹³
• Perpetual optionality. Over 300 exploration and advanced royalties cost nothing further and cover large areas around operating mines.²

Where it is thin

The cost structure protects only against operating cost. It does nothing against the price of buying new ounces. Royalties are bought in an auction among a few buyers, and when gold rises, sellers want more. In 2025 the company paid US$1,050M for a 7.5% gross margin royalty on Côté Gold, US$300M for a royalty at Porcupine, and US$500M for the Western Limb stream.⁴ Those three produced about 48,000 GEOs a year at the second-quarter rate C (calculated), so the price was about US$38,000 for each annual ounce, or about 8.5 times that quarter’s average gold price of US$4,517 C (calculated).³ A buyer who pays that has to hope gold stays high, or that the mines grow.

Operators can also take the royalty back. In March, after Jiangxi Copper bought SolGold, the buyer exercised its right to buy back half of the Cascabel stream and royalty, and Franco-Nevada received US$138.2M, a gain of US$63.8M C (calculated).³ That is a fair price on paper, and it ends a stream that would have run for decades. Buyback rights, step-downs and mine lives all mean that part of each year’s cash is a return of capital, not a return on it.

The record over a decade shows what that costs. From 2016 to 2025 the company produced US$8.13B of operating cash flow and put US$6.95B, 85% of it, into new interests, equity stakes and loans C (calculated).⁵ Weighted shares went from 175.2M to 192.7M. Precious-metal GEOs were 457,474 in 2016 and 440,140 in 2025.⁹² The two counts use different conversion prices, and Cobre Panamá, which sold 43,554 GEOs in 2019 and then halted in 2023, explains much of the gap.²⁰¹⁸ But the lesson stands. Reinvesting 85% of operating cash flow kept the ounces per share about where they were. The average gold price rose 2.75 times, and operating cash flow per share rose 2.88 times C (calculated).⁵² In other words, the owners’ cash per share followed the gold price.

I read this as follows. The business model works like a conveyor belt, and the belt has to be fed. Most of what is bought replaces what is mined out. The record on price paid is mixed. The one large writedown I found in the annual tables was Cobre Panamá in 2023, which was a government decision, and the Côté royalty is still ramping. What the company can’t do is get around the price of buying ounces, because there is no way around it.

IIIThe competitive advantage

The honest answer is that Franco-Nevada’s edge is size and balance sheet, not a different product. Royal Gold and Wheaton offer the same thing. What I can check is how the three are priced.

On 2025 cash flow the three are within two turns of each other. On the first half of 2026, Franco-Nevada is priced at 21.6 times twice its half-year cash flow, and Royal Gold at 15.4 C (calculated). A reader needs care here. Royal Gold’s share count jumped from 65.8M to 84.4M between its June and September 2025 reports, so its 2025 cash flow includes only part of a year of whatever it bought.¹⁰ I did not open its deal records. The half-year numbers are cleaner, and they say the market charges Franco-Nevada about 40% more per dollar of cash flow C (calculated). The reason may be quality, or the Cobre Panamá option, or the energy interests, or the no-debt balance sheet. The price paid is a fact, and the reason is a guess.

Three things Franco-Nevada does that I can see

It lends and buys across the structure. The circular says the company has become a financial backer, with lead orders in the equity raises of Tocantinzinho and Porcupine, and now royalty or stream packages for Casa Berardi, i-80 and Bullabulling.¹³ On Bullabulling the total is A$420M, about US$300M, for a royalty of 3.90% on a project with a pre-feasibility study, expected to start production in 2028, and a Franco-Nevada right of first refusal on future streams and royalties.²² I can’t judge the project. What the structure shows is that it gets more than a royalty: it gets a seat in the operator’s next financing.

It earns on assets it doesn’t operate and doesn’t have to defend. When Lundin Mining expands the Candelaria underground to 22,000 tonnes a day by 2030, or Agnico Eagle drills below the Detour Lake pit, the operator pays for it and Franco-Nevada’s share of the extra ounces costs it only the per-ounce price in the contract.⁸

It stays small. Forty-four people can’t run a mine. They also can’t get tangled in one. That is part of why a US$1.8B year in 2025 didn’t change the cost line.¹³ But it means the company relies on operators’ public statements for almost all of its production data, and its own guidance says so.⁸

The risk in the partners

The weak spot is the partner. Antamina’s silver is 11.3% of revenue, and the company says third-quarter silver deliveries will be lower because of lower throughput there.⁸ Hemlo revenue fell in the quarter because mining went through lower-grade stopes.⁸ Candelaria produced less than a year earlier.⁸ In each case Franco-Nevada learns about it when the operator reports. The company can ask, but it can’t fix.

Rivals in waiting

A new rival can enter by buying a royalty book. Royal Gold’s share count rose 28% between its June and September 2025 reports, which suggests it bought something large; I did not check what.¹⁰ Larger mining companies also part with royalties when they need cash, and any buyer with a lower return requirement can outbid Franco-Nevada. I think this is the main reason the company’s edge shows up as a premium multiple and not as a higher return on each deal.

IVManagement

Paul Brink, 58, became chief executive in May 2020. He has been with the company since its IPO in 2007, led business development from 2008, and was president and chief operating officer from 2018 to 2020. Before that he worked in corporate development at Newmont, in investment banking at BMO and in project finance at UBS. He is an engineer from Witwatersrand with an Oxford master’s degree in management studies.¹³ Sandip Rana, the chief financial officer, joined in 2010; Eaun Gray is chief investment officer; Lloyd Hong is chief legal officer, and he led the Cobre Panamá arbitration.¹³ It is a stable team, and the people who built the deal book are the ones who run it.

David Harquail founded the company, was its chief executive from the IPO and has been non-executive chair since 2020. He said in March that he would not stand for re-election at the May 2026 meeting. The board said it intended to appoint Tom Albanese, a former chief executive of Rio Tinto and Vedanta Resources and the board’s lead independent director, as chair.¹³ Harquail’s holding fell from 928,505 shares to 773,505 in the year to 19 March 2026, a drop of 155,000, and the rest is worth C$247.2M at the circular’s price.¹³ The circular doesn’t say why it fell. A drop can be a sale, a gift or a transfer, and I did not open an insider filing to see which. It is still a data point, and the only change in a holding that I saw in the circular.

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

For 2025 Brink was awarded C$4.69M against a target of C$3.9M: a cash bonus of C$1.7675M against a target of C$975,000, time-based units of C$975,000, and performance units of C$1.95M. For 2026 his salary rose 3.5% to C$1.009M.¹³ The committee gave above-target bonuses because the company deployed US$1.8B, three-year average growth in net asset value per share beat its internal benchmark by a wide margin, one-year total return was 74%, and the tax dispute was settled. It also noted that return relative to peers was below the median.¹³

The performance units are the largest part. They vest after three years, and the number that vests moves with a factor. Under the earlier design the factor was 1.98 for the 2018 grant, 1.968 for 2019, and zero for 2020.¹³ The committee looked at that swing and decided vesting shouldn’t rest on a formula alone. It now uses a new comparator group and a strategic review. The two stated measures are relative total return and net asset value growth per share.¹³ I have two thoughts. First, stock options were stopped in 2020, which removes the cheapest way to make money from a rising gold price.¹³ Second, growth in net asset value per share pays management to buy interests that raise the estimated value of the book. That’s close to right. But the estimate depends on a gold price assumption and on which projects count, and the circular doesn’t give the formula. A team paid on that number has a reason to do deals when sellers are happy, which is when gold is high. I can’t see from the filing whether the benchmark discounts for that.

How much of their own money is in

A lot. Brink holds 252,655 shares, worth C$80.1M, plus performance and time-based units of C$12.2M and options worth about C$30.4M C (calculated), for C$122.8M at risk. The circular says the chief executive’s holdings are more than 14.1 times his total direct pay.¹³ Rana holds 58,464 shares worth C$18.5M and Hong 16,600 worth C$5.3M.¹³ This is the kind of ownership that makes a paid manager think like an owner. Harquail still owns 773,505 shares, and FMR LLC reported 21,597,096 shares, 11.2%, as of 31 March 2026, and BlackRock 11,933,281, 6.2%, as of 30 September.¹³²³²⁴ Those are holdings. They aren’t statements of reasons.

What they did with capital over a full cycle

The ten years to 2025 included a bust in gold and the boom. The company produced US$8.13B of operating cash flow, invested US$6.95B, paid US$1.80B in cash dividends, and raised its dividend every year C (calculated).⁵ Weighted shares rose 10%, from 175.2M to 192.7M, and in 2026 financing was dividends and option proceeds with no buyback.⁵³ There was no debt at any year end except 2018 (US$207.6M) and 2019 (US$80.0M), and the company drew US$175M on its revolver in July 2025 to help buy Arthur and repaid it in the same quarter.²⁰²

On the whole this is a competent team that did the harder thing in the cycle, which is to keep buying. Two caveats. A stream or royalty bought from a miner in need can look cheap in a downturn and costly if it is bought at the top, and 2025 and 2026 are not the bottom. And the pay design rewards deals. Over ten years the dollars went into assets and not into the shares, so we don’t know how this team would treat a large pile of surplus cash when there is nothing good to buy. The US$4.3B of available capital is the first real test of that.

VFinancials

The statements are simpler than most. There is almost no inventory, no plant and no debt. Three things still need explaining.

• Revenue is gross. Streams are booked at the full sale price of the metal, and the payment to the operator comes through as a cost of sales. Revenue therefore moves one for one with the gold price on streams.
• GEOs are not comparable over time. From the fourth quarter of 2021 oil and gas are counted in GEOs; before that they were not. In 2026 the company fixes conversion at its guidance prices, so GEOs no longer move with commodity prices. I use them only where the method is clear.³¹⁹
• Adjusted EBITDA is the company’s own measure. It leaves out depletion, impairments and gains on the sale of interests, such as the US$63.8M on Cascabel.³

The second quarter

Revenue was US$580.9M, up 57%. The average gold price was US$4,517 against US$3,279 a year before, silver US$73.44 against US$33.64, and the number of GEOs rose 18%.³ Adjusted EBITDA was US$529.7M, up 45%, and net income US$354.0M, up 43%.⁸ Operating cash flow was US$482.5M, up 12%, for two reasons: more tax was paid, and less gold and silver bullion was sold than a year earlier.³ The first half of 2026 set records for revenue (US$1,231.6M), operating cash flow (US$1,002.9M) and adjusted EBITDA (US$1,121.6M).⁸ Gold futures at the end of October were about 14% below the second-quarter average price C (calculated), so the third quarter will report lower prices, though the company’s Cobre Panamá deliveries start then.³⁶

What an owner can really take out

Operating cash flow is real cash, and it already bears the company’s small stock pay. In the first half, stock-based expense was US$2.7M.³ So there’s nothing to add back. But operating cash flow isn’t what an owner can take out, because a royalty company that stops buying shrinks. Mines run down, step-downs bite and buyback rights are used. To keep the ounces flat the company has to buy replacements.

The 2030 outlook already nets known step-downs, so I charge no replacement before 2031. After that I assume the book loses 3% of its ounces a year, and the company buys them back at 8.5 times the gold price per annual ounce, the price of the 2025 deals E (estimate). That is about US$517M a year at today’s gold price and ounces, and US$571M in 2031 in my base. The charge matches the record: operating cash flow less investing outflows was US$1.2B over ten years, so the company kept about 15 cents of each dollar before dividends C (calculated), and in that time the ounces per share held flat or fell C (calculated). I can’t pin the rate down. At 6% the base value falls from $130 to $97. At zero it rises to $163.

That leaves about US$1.26B of owner cash on the last twelve months of US$1.78B, which is 2.9% of market value E (estimate). Dividends of US$1.76 a share, about US$340M a year, are 19% of operating cash flow and 0.8% of the price C (calculated). The rest is kept to buy interests. An owner doesn’t get it as cash. He gets it as more royalties, if the company buys at a fair price.

Balance sheet

At 30 June the company had US$1,014.2M in cash, US$1,215.1M of equity investments, no debt, and total equity of US$8.24B.³ Of the investments, US$1,154.0M are listed shares and US$124.3M is a stake in Labrador Iron Ore Royalty, which the company counts as a royalty.³ The rest of its available capital of US$4.3B is two undrawn revolving credit lines of US$1.5B, with a US$750M accordion, maturing in 2031 and 2029.³ Current liabilities include US$32.8M of Barbados minimum top-up tax for 2024 that was unpaid at 30 June. The authority said no penalties or interest apply if paid by 30 September.³

The tax dispute with the Canada Revenue Agency ended with the September 2025 settlement. It had assessed US$204.7M for 2013 to 2019, and the company had posted security of US$44.7M in cash and US$48.1M of letters of credit. In the first quarter of 2026 the letters were released and US$44.1M plus US$5.4M of interest came back.¹⁴³ The settlement doesn’t bind the CRA after 2019, and Canada changed its transfer-pricing rules from 2026.³ I leave that open in my base and test it in the geopolitical section.

How the shares have traded

The shares closed at US$65.46 on 31 October 2016 and at US$224.95 on 30 October 2026, which is 3.4 times, plus dividends C (calculated).¹ Gold futures rose 3.1 times over the same ten years C (calculated). The stock hit US$280.82 on 26 February 2026, a month after gold’s record, and has fallen 20% since.¹⁶ The company’s own letter says the share price has appreciated 1,414% since the 2007 IPO, that returns have compounded at about 17% a year, and that this is double the return on bullion.¹³ That claim is about 18 years. The last ten look like gold plus a small premium.

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.

• Gold falls and stays down. Gold has risen from an average of US$1,248 in 2016 to US$3,435 in 2025, and the futures price reached US$5,318 on 29 January 2026 before falling 27%.²⁶ Nothing in the company’s costs falls with it, because the costs are a share of the price. If gold went back to US$2,500, my base falls to $93, a 28% drop.
• The treadmill. A royalty book is a melting asset. Over ten years the company spent 85% of its operating cash flow on new interests and its ounces per share didn’t grow. If the cost of staying level doubles, my base falls to $97. The danger is that the company pays high prices for ounces when gold is high, and then gold falls.
• One country, one mine, one decision. Cobre Panamá cost a US$1,169M writedown, and at full capacity the company says it could add 150,000 to 175,000 GEOs a year.¹⁸² It is the example of how a government can turn a stream off. Peru, Chile and Mexico were 42% of first-half revenue C (calculated).³

None of these needs anything strange to happen. They set my bear case, and they explain 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 royalty book as one business, except Cobre Panamá. Revenue is precious-metal GEOs times the gold price, plus diversified revenue. This treats silver and platinum group metals as moving with gold at the ratios in the company’s guidance (gold 4,500 to silver 75, 60 to 1). In the second quarter the ratio was 61.5 to 1.⁸³ Diversified revenue is the 2026 guide of US$245M to US$285M at WTI of US$70, so I use US$265M, growing 2% a year. Cash costs are 7.5% of revenue, as in the first half. General and administrative expense is US$46M, and growing 3%. Cash tax is 19.5% of adjusted EBITDA less depletion of US$335M, the middle of the company’s guide, growing 2%. The company’s own rate is 20% to 23%, partly deferred.³

The owner cash is operating cash flow less the replacement bill defined above. I count the ten years 2027 to 2036, each year’s cash at year end, from 1 January 2027, then growth of 2% for ever, which is inflation and no more. Cobre Panamá is valued apart, at two points more discount. The stream gives 150 ounces of gold and 1,720 ounces of silver for each million pounds of copper, bought at US$418.27 an ounce for 120 of them (indexed for inflation) and at 20% of spot for the other 30.¹⁴ I take the payment as 15% of revenue, tax at 15%, and a finite life, because the commission’s first recommendation is talks toward a project with an “orderly closure” and no extension.²⁵ Cash and equity investments come in at 30 June marks, which are probably stale, since the shares of gold companies tend to move with gold. I count all 193.5M diluted shares, including options and unvested units.³

The discount rate. I use 8.5%: the ten-year Treasury at 4.94% plus 3.5%.⁷ That is the rate on my Costco and LVMH pages, and 1.5 points under the 10.0% on Cameco, because Franco-Nevada has no debt, no construction and no single mine above 12% of revenue. A rate of 6.5% in the grid below lifts the base to $176. Nothing within a reasonable band reaches the price.

What the price implies

At $224.95 the market value is US$43.4B, which is 24.4 times TTM operating cash flow C (calculated).¹³ On my numbers it is a business that pays about 2.9% of its price in owner cash today, less than the 10-year Treasury, and promises growth through gold and deals. The question is what the price needs. I solved it two ways.

This is the most useful table on the page. To reach $224.95 on my base cash flows gold has to sit near US$7,700 for good, which is about twice the 30 October futures price and 45% above the January record. Or owners have to accept 5.4% a year, which is less than half a point over the 10-year Treasury, which pays 4.94% with no risk to the money. The market may have good reasons. It may expect gold to keep rising as central banks buy, since the company itself describes heavy buying by central banks in the first half.³ It may put a high value on the 300 exploration and development royalties, which the company’s outlook leaves out and which I credit only through the ounces I assume. It may think a debt-free company that has done US$2.5B of deals since 2025 will keep finding good ones. All three are possible. I can put numbers on the first only if I assume it, and so I have not tuned any input to reach the price.

What I can’t judge

Two inputs decide this page, and I can’t know either. I don’t know where gold will trade in 2032. I don’t know what the next ounce will cost. So here is the grid. Read my $130 as one cell in it.

A US$1,000 move in gold at the middle of the grid is worth about US$25 to US$30 a share. A 1-point change in rate is worth about US$14 to US$19. 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 real rates fall, gold tends to rise and the rate owners accept falls. That is the strongest version of the market’s case, and I can’t rule it out.

Sensitivity

Three cases

Other uses of the same dollar

The ten-year Treasury pays 4.94% with no risk to the money R (reported).⁷ My base returns 0.6% a year at this price, my bull 7.9% and my bear loses 7.9% a year E (estimate). The closest alternative in the same field is Royal Gold, which is priced at 15.4 times twice its half-year cash flow against 21.6 for Franco-Nevada C (calculated). I have not valued Royal Gold, so I can’t say it is better value, only that it costs less per dollar of cash flow.¹⁰¹² There’s no other gold company on this site yet.

The margin of safety

• Price$224.95Close, Q4 2026 [1]
• Base value$130My estimate E (estimate)
• Margin of safety below$80set just above the bear case E (estimate)

I set the margin of safety just above the bear case, because with a price this far from my values a discount to base is no protection. Below $80, which is 38% under my base, my $72 bear value is about nine-tenths of the price paid. That describes a market that has given up on gold: the shares last closed at or below US$80 in June 2019, when gold futures were US$1,347 C (calculated).¹⁶ I don’t expect that, and I’m not forecasting it. It is simply the price at which I could be wrong about gold and not lose much.

Two different questions

How good is the business, and how good is the price? On the first, I find a debt-free company with 44 employees, a 91% margin, 447 interests, a record of dividend growth, and no single asset above 12% of revenue. It is a good business by any test I know. On the second, at $224.95 the price is above my bull, and the cash an owner can take out is about 2.9% of the price. A great business bought at a price that needs gold to double is a different thing from a great 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 8.5%, of the owner cash after it, per today’s 193.5M shares, in that year’s dollars. It includes the royalty book and Cobre Panamá as in my base. It leaves out today’s cash and shares of US$13.6 a share, which I treat as paid out in 2027 and count in the owner cash column. The return is the yearly rate that turns today’s price into the 2036 value plus the cash paid on the way.

Why, in plain words. For the base to stand at five years, gold has to stay near today’s futures price after inflation, the ounces have to reach about 515,000 by 2030 as the company outlines, and buying them back has to cost about 8.5 times the gold price per ounce. By 2036 it also needs the replacement bill to stay at about 3% of ounces a year. The base value rises only from $120 to $128 over those five years, because owners are paid about 6% of value each year along the way. The bear differs from the base on gold (US$2,300 in 2029), on ounces (down to 450,000), and on Cobre Panamá. The bull differs on gold (US$5,800), on ounces (590,000), and on a 160,000-ounce Cobre Panamá.

The numbers that drive it are gold, the replacement bill and the discount rate. The single most likely way my base is wrong is gold. At US$2,500 the base falls to $93, and at the January record of US$5,318 it rises to $164. I do not know where gold will be.

A reader’s guide. These are estimates of value, not forecasts of the share price, and the price can sit above or below value for years; today it sits above even my bull.

VIIThe case for the price

What most people miss

Three things cut in the company’s favour and get less attention than gold. First, the tax dispute is over. In September 2025 the CRA agreed that no Canadian tax is due on the foreign earnings of the Barbados and Mexico subsidiaries for 2013 to 2019, and in 2026 the company got back US$44.1M of cash plus US$5.4M of interest.¹⁴³ Second, the margin is steady. Cash costs were US$344 a GEO in the first half against revenue of about US$4,600, so each extra US$100 on gold adds about US$92 to EBITDA C (calculated).³ Third, oil. Energy interests earned US$129.8M in the first half, and the company says each US$10 on WTI adds about 12% to oil revenue.³⁸ That is an exposure a gold investor may not expect.

The bull case

In plain words: gold is a hedge that central banks and investors want, Franco-Nevada keeps about 91 cents of each extra dollar of revenue, because its costs don’t follow, and it has US$4.3B to spend while miners are short of money. It also holds over 300 exploration royalties that cost it nothing more.³²

Management makes the case in its own words. Brink said on 11 August that, with US$4.3B available, the company was positioned for a strong pipeline of deals.⁸ The March letter to shareholders says the ambition is to be the “Go-to Gold Stock” for generalist investors, notes a return of about 17% a year since the 2007 IPO, expects about US$2.0B of operating cash flow in 2026 at the prices then, and names two catalysts: a final decision on Glencore’s Coroccohuayco project, and a Cobre Panamá restart that would add 30% to the growth outlook.¹³ Among holders, FMR LLC held 21,597,096 shares, 11.2%, at 31 March and BlackRock 11,933,281, 6.2%, at 30 September.²³²⁴ Those are holdings and not reasons. I did not find a dated analyst note I could verify, so I name none.

Against reality

Today’s numbers support the first half of that case: first-half operating cash flow of US$1.00B was a record, and the balance sheet is as strong as claimed.⁸ What is not there yet is the price. Gold futures are 27% below the January record and the 2026 guide of US$4,500 is 15% above the futures price C (calculated).⁶⁸ Cobre Panamá is processing stockpiled ore while the government is asked to open talks, with an audit compliance rate of 87.7% as the company states it (it also says 361 of 370 commitments were met, which works out nearer 98%, and I do not know which is right) and a commission that wants a project with an end date.³²⁵

If gold stays at the January record, if the 2030 outlook is raised above its top, if Cobre Panamá restarts and runs near full capacity, and if the company keeps buying at today’s prices, then $224.95 is possible. My bull case gets close, at $213. Today, against that list, the score is: the balance sheet yes, the other four not yet.

Multiply it out. Gold at or above US$5,300 for good: 20% E (estimate). Futures closed above US$5,000 on only 27 days in 2026, between 26 January and 17 March.⁶ The 2030 outlook raised above its top: 40%. Cobre Panamá full restart: 25%. Together about 2% if they were independent E (estimate). They are not, since high gold helps ounces and financing, so I would say 5%. Each looks reasonable alone; together they are a long shot. The 25% I give my bull case in the weighted value of $136 is, if anything, generous.

What a 10x would take. Ten times $224.95 is $2,250, about US$435B of market value. At today’s 24.4 times operating cash flow that needs about US$18B of operating cash flow, against US$1.78B now C (calculated). At today’s ratio of cash flow to revenue, that is revenue of about US$23B C (calculated). With about 540,000 GEOs a year, that is gold near US$43,000 an ounce. Or gold at US$6,000 and 3.9 million ounces a year, seven times today’s. The share price rose 3.4 times in ten years and the company says 15 times in 18, from a company that was far smaller C (calculated).¹¹³ Tenfold from US$43B has not happened to any company I know in this field, and I did not search systematically. Here it would need either a gold price no one has seen or ounces that no list of projects supports. It is close to impossible, and I would build nothing on it.

Where I come down

The settlement of the tax dispute and the 91% margin moved me; without them my bear case would be lower than $72. The price did not move me, because it already assumes more gold than any of my cases. A signed Cobre Panamá deal, or cheaper ounces on new deals, would raise my values most, though a full restart is worth only about US$8 a share more than my base.

VIIIRisks, and what would change my view

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

• Operators. Franco-Nevada runs nothing. If a mine floods, a pit wall moves or a manager changes the plan, the company finds out when the operator reports. In July, Agnico Eagle reported a rock movement along the Barnat pit wall at Canadian Malartic, which the royalty does not cover.⁸
• Ounces lost. Candelaria’s share steps down in 2027 and Antapaccay’s in 2028; Cascabel was half bought back; the Ahafo open pit finished in 2025.²³ Some of each year’s cash is a return of capital.
• The price of the next ounce. The company paid about 8.5 times the gold price per annual ounce in 2025. If it pays more, or if gold falls after it pays, returns on new money fall.
• Mark to market. US$1.2B of equity investments, mostly shares of mining companies it has financed, are marked at market and tend to fall with gold.³
• Tax. Guidance is a 20% to 23% effective rate, a Barbados top-up tax of US$32.8M for 2024 was unpaid at 30 June, and Canada changed its transfer-pricing law from 2026.³
• US taxpayers. The company lists among its risks whether it is found to be a passive foreign investment company. That matters for how a US holder is taxed, and I am not giving tax advice.⁸
• Thin management bench. Forty-four people, a chair who was due to change in May, and a CEO who has been in the job since 2020.⁴¹³
• Equity issuance. The weighted share count rose 10% in ten years, and a large deal could be paid for with new shares.⁵

Geopolitical risk, and what it means for the value

Franco-Nevada’s revenue is spread over many countries and mostly in the Americas: in the second quarter 40% came from South America, 25% from Canada, 16% from the United States, 7% from Central America and Mexico and 12% from the rest of the world C (calculated).⁸ The exposures that matter sit in a few places. Peru (Antamina, Antapaccay, Condestable, Yanacocha) was 22% of first-half revenue, Chile (Candelaria, Salares Norte) 12% and Mexico (Palmarejo) 8% C (calculated).³ Panama is the live case. I found no Russian or Iranian interest in the revenue table. China matters in a different way: Jiangxi Copper bought SolGold and used its right to buy back half of the Cascabel stream and royalty.³ Governments, courts and tax authorities decide most of this, not markets.

I ran each scenario through my model, holding everything else at base. The odds are my guesses, with the basis shown. The bigger point is what is already in the price. At $224.95 the price needs gold near US$7,700 for good. Wars and sanctions can lift gold. The company says gold hit record highs on central-bank buying early in 2026 and then reversed, with conflicts in the Middle East feeding inflation.³ So some of the market’s case is a bet on more geopolitical stress. What the price does not include is a country-specific loss or a higher tax rate: those would come on top.

Two further risks sit in the filings. The first is a new Canadian rule on transfer pricing in force from 2026; the company says it is still weighing the effect.³ The second is tariffs and trade retaliation, which the company lists among its risks.⁸ I have not modelled them beyond the tax case above. I have not forecast any event.