A quick note on timing before we start.
As most of you might know, due to travel, I am writing this article deliberately ahead of earnings for all three companies I cover here. Royal Gold (RGLD) reported yesterday. Wheaton (WPM) reports today. Franco-Nevada (FNV) reports on August 11.
But that's OK, as it doesn't make this article any less valid - I obviously made sure of that.
This piece is about business structure, not about a quarter. The things I care about here are contract terms, asset concentration, cost per ounce, and where growth actually comes from. Those move on a five-year clock. A single quarter will not change any of them.
I will follow up once all three have reported and we can see what the prints actually showed. For now, treat this as the framework you need to read those prints properly and to have a better understanding of the industry.
One more thing upfront. I do not own any of these three companies. I will explain why at the end, and I will tell you exactly what would change my mind. I think that is more useful than pretending I have a position I do not have.
Let's get into it.
Why I Keep Coming Back To Gold
My gold view has not changed in years, and it has very little to do with charts.
It comes down to a single structural tension.
On one side, artificial intelligence is the most powerful deflationary force the global economy has met in generations. It automates cognitive labor. It drives the marginal cost of intelligence toward zero. It scales revenue without scaling headcount.
On the other side, heavily indebted sovereign governments cannot survive a prolonged deflationary spiral. Deflation makes debt heavier in real terms. It is catastrophic for their ability to service what they already owe.
So they spend. They inject capital through national security budgets, industrial subsidies, and emergency fiscal mechanisms. They accept higher inflation because inflation quietly erodes the real value of the debt.
That is the "run it hot" thesis I keep writing about.
Gold sits completely outside that system. It cannot be printed. It cannot be diluted by a policy decision. It has survived every monetary regime in recorded history, which is a longer track record than any currency currently in circulation can claim.
Here is my problem, though.
I travel constantly between the EU and Albania. I have a public profile (I'm obviously not famous, but I write using my own name, and I'm blessed to have a large audience). Storing meaningful amounts of physical metal is genuinely impractical (and potentially dangerous) for me. And gold produces nothing while it sits there.
So, while I would love to own physical gold, I have always looked for a business that gives me gold exposure with a cash flow attached.
That search has always led me to a few select high-quality miners and, above all, the streamers.
Why I Want The Royalty, Not The Mine
Mining is a brutal business. You raise billions, you dig a hole in a country that might change its mind about you, you fight labor costs and diesel costs and grade decline, and if metal prices fall you eat the entire difference because your costs are fixed and your revenue is not. This applies to almost everyone, not just gold and silver miners.
The royalty and streaming model removes almost all of that.
These companies write a cheque upfront. In exchange, they get a permanent claim on a percentage of what comes out of someone else's mine, funded by someone else's balance sheet, operated by someone else's workforce. And, to be honest, just writing this gets me excited about this industry.
Royal Gold has a great overview of this in its investor presentation:
I think the scale of that efficiency is genuinely difficult to believe until you see the numbers.
The chart below helps me make that point even better:
To add another few numbers, Wheaton runs its entire global business with 46 employees. Royal Gold runs its business with 39 people across four offices. Royal Gold's own materials show enterprise value per employee at roughly $568 million, which is more than twenty times Apple's (AAPL) figure and more than thirty times Alphabet's (GOOGL). It doesn't mean these companies are better, but the comparison is interesting nonetheless.
Together with Franco-Nevada, these three companies are worth around $107 billion.
Royalties And Streams Are Not The Same Instrument
So far, the basics. Now, we need to dig deeper.
A royalty is a percentage of gross production or gross revenue. There is no ongoing cost. If the mine produces, you get paid. Your margin is essentially the metal price minus a rounding error.
A stream is a right to buy metal at a fixed, deeply discounted price. You pay upfront, and then you keep paying a small ongoing amount for every single ounce delivered. Franco-Nevada's Cascabel stream, for example, carries an ongoing payment of 20% of spot price for each gold ounce delivered.
The consequence matters enormously.
A royalty is linear to the metal price. A stream is levered to it. When gold runs, the streamer's fixed cost stays fixed while revenue climbs, so margins expand faster. When gold falls, that leverage runs the other way.
You can see this in the actual cost data.
Franco-Nevada, which is royalty-heavy, reported a cash cost of $341 per GEO in 1Q26. Wheaton, which is almost entirely streams, reported $681 per GEO in the same quarter, up from $514 for full-year 2025 and $438 in 2024. Royal Gold, at roughly two-thirds streams, ran $677 per GEO in 2025.
Note that GEO stands for "gold equivalent ounces."
Wheaton's cost is rising by design. That is what a stream book does as it grows. Wheaton guides to $650 per ounce through 2030 at an 84% cash operating margin, and it markets the leverage as a feature, which it honestly is.
On a side note, this difference is also visible in the two company charts I just used.
Keep that mechanic in mind. It shows up in the next part of this article.
The Scorecard
Here's the full TOLL scorecard:
Yes, my top pick loses on TOLL. I will explain that.
The Framework
TOLL is my standard screen, as it focuses on Tangible assets, Oligopoly power, Low incremental capital intensity, Long-duration cash flows. Often, I apply twenty-five points to each segment.
For this sector, I added four tiebreakers because TOLL was built to identify durable businesses, and it does not capture things that only matter when you are financing someone else's mine:
Structure, meaning the royalty versus stream mix.
Concentration, meaning single-asset and jurisdictional exposure.
Capital allocation, because this business shrinks unless management keeps redeploying, which makes deal discipline the actual skill.
Valuation, on multiples that suit an asset-light royalty book.
One warning before the individual sections. Every one of these companies publishes a scorecard it wins on. Franco-Nevada leads with return on invested capital, where it shows itself at roughly 20% against peers near 10%. Wheaton leads with precious metals purity, at 99% of forecast revenue. Royal Gold leads with price to net asset value, showing itself at 1.21x against a historical range of 1.5x to 2.5x.
All three are accurate. All three are chosen. That is exactly why I score them myself.
Franco-Nevada: The Cleanest Business, The Messiest Situation
Franco-Nevada is the highest-quality royalty book of the three, and it is not close.
The portfolio holds 121 cash-flowing assets across 14 countries, with no individual asset above 15% of the total. Cash cost was $341 per GEO in 1Q26 against a margin of $4,534 per GEO, which works out to a 91% adjusted EBITDA margin. G&A ran 2.5% of revenue in 2025, the lowest of the three. The company carries zero debt and roughly $3.4 billion of available capital.
It has raised its dividend 19 consecutive years and has paid out more than $2.9 billion since its IPO.
I have always liked Franco-Nevada for a reason most gold investors overlook. It is not purely a gold company. In addition to precious metals, it holds energy royalties on producing oil and gas acreage, which generated $59.4 million in 1Q26, plus iron ore and other mining interests at $23.2 million. Those are structurally identical to the land and mineral royalties I write about constantly.
Now comes the problem.
Cobre Panama has been in preservation and safe management since November 2023, when Panama's government ordered the mine to stop. Franco-Nevada took roughly a billion dollars of impairment. The asset was one of its largest.
Here is why that story is changing.
Both units of the site power plant have been recommissioned and synchronized to the national grid, with coal shipments received and all systems tested under nominal operating conditions. Panama approved stockpile processing on April 7, 2026. Franco-Nevada has already booked a partial impairment reversal, recognized in 3Q26, with a further reversal indicator following that April approval.
The arbitration hearing is scheduled for October 2026. Franco-Nevada values its claim above $5 billion.
If the mine restarts at full capacity, the stream could contribute 150,000 to 175,000 GEOs annually (see below). Against a 2025 base of 519,106 GEOs, that is roughly 30% of current production. Think about that for a second.
This is why Franco-Nevada scores poorly on the second L in TOLL. Long-duration cash flows should not depend on the political calendar of a country you have no influence over.
It is also why I rank it first anyway. I will come back to that.
Wheaton: The Growth Is Already Signed
Wheaton has the best growth story in the sector.
The company produced roughly 803,000 GEOs in 2025 and guides to 860,000 to 940,000 for 2026. It targets 1.2 million GEOs by 2030 and expects to hold that level through 2035. That is roughly 50% growth, and the streams behind it are already signed. Development projects not included in guidance could add another 200,000 GEOs on top.
Reserve life on proven and probable reserves is 23 years. Around 80% of production comes from mines operating in the lower half of their cost curves.
And, to add another detail you may like, Wheaton also has the deepest pockets in the industry. In 2026, it closed the largest precious metals streaming transaction ever recorded, paying BHP $4.3 billion for the Antamina silver stream.
Two things bother me.
First, concentration. Wheaton's 2025 production mix was Salobo at 37%, Peñasquito at 16%, Antamina at 12%, and Constancia at 10%. A single asset at 37% is a lot of eggs in one Brazilian basket. Management expects that to improve to 26% by 2030 as Antamina ramps, which helps, though it takes four years.
It's not a risk that keeps me up at night, but it needs to be mentioned, as assessing geopolitical risk is always important.
Second, the balance sheet. Funding Antamina involved a $1.5 billion two-year term loan, and Wheaton moved from a net cash position to roughly $2.1 billion of pro forma net debt at around 0.7x EBITDA. That leverage is manageable (it's still very healthy). But then, it is still a company borrowing to buy growth, which is exactly the reinvestment treadmill I described earlier.
Also, revenue mix is 62% gold, 36% silver, with palladium and cobalt rounding it out. The silver weighting gives you more volatility than the other two, as silver is simply more volatile than gold (and more cyclical).
Royal Gold: The Cheapest, And The Newest
Royal Gold is the only U.S.-based name here, which matters for withholding tax. Canadian dividends face 25% statutory withholding, reduced to 15% under most treaties. Franco-Nevada and Wheaton both also carry disclosed Canada Revenue Agency audit exposure around the tax treatment of offshore streams. Royal Gold has neither issue.
It also has the best dividend record in the group. Twenty-five consecutive years of growth, a 15% dividend CAGR since 2000, and the highest yield of the three at roughly 1%. I know, I know, 1% isn't really a dividend, but the streak is worth mentioning.
And it is the cheapest. Royal Gold trades around 22.2x EV/EBITDA against 23.3x for Wheaton and 23.4x for Franco-Nevada. Its own materials put price to net asset value at 1.21x versus a 1.5x to 2.5x historical range.
Revenue splits 67% streams and 33% royalties, with almost 370 total properties (most aren't producing yet), and 94% exposure to gold and silver (dominated by gold).
Moreover, the reservations are real.
Concentration is meaningful, with the top five properties at 53% of revenue and Mount Milligan alone at 22%. Royal Gold is also the only one of the three carrying net debt, at roughly $360 million.
Most importantly, this is a substantially new company. Royal Gold closed Sandstorm Gold and Horizon Copper for $4.1 billion in October 2025, which was paid mostly in equity, plus a $1 billion Kansanshi gold stream.
The Uncomfortable Part
Here is the section I suspect nobody wants to read. And it doesn't give me pleasure to write it.
Over the ten years ending July 28, 2026, with dividends reinvested, only WPM beat the S&P 500.
Note that my model didn't include max drawdown data for the S&P 500 (it says N/A). However, for the purpose of this article, that's no issue.
Royal Gold underperformed the metal itself and underperformed the miners it was supposed to be a safer alternative to, while suffering a drawdown just as deep. And the drawdown numbers demolish the idea that royalty companies give you gold exposure gently. Physical gold's worst peak-to-trough was 26.4%. Wheaton's was 48.6%, and Royal Gold's was 49.6%.
It gets worse. The 2025 calendar year alone returned 154.7% for GDX, 110.5% for Wheaton, and 70.5% for Royal Gold. For the miners, that single year accounts for essentially the entire decade of gains.
I am writing a positive piece about this sector immediately after the best year it has had in a generation. You should hold that fact in your head while reading everything above.
Where I Land, And Why I Own Nothing
My TOLL model ranks Wheaton at 88 points. I rank Franco-Nevada first anyway, at 85.
Here is my reasoning, and you are obviously welcome to disagree as I'm ignoring my own model.
The framework penalizes Franco-Nevada for Cobre Panama under long-duration cash flows, and that penalty is correct. What a scoring system cannot do is hold the same fact two ways. Cobre Panama is a scored negative and an unpriced call option at the same time. Roughly 30% of production, a claim above $5 billion, a power plant already back on the grid, stockpile processing approved, impairment reversals already booked, and a hearing in October.
Underneath that, I am buying the better business. Half the cash cost per GEO. The lowest G&A. No debt. No single asset above 15%. Nineteen years of dividend increases. Plus energy and iron ore royalties that give me exposure I actually want for reasons that have nothing to do with gold.
Wheaton is second, and it is genuinely close. If you want the contracted growth and you can live with 37% in one mine, take it.
Royal Gold is third, and third here is not an insult. It is the cheapest, has the longest dividend record, and avoids the Canadian withholding drag. I want more evidence on the Sandstorm deal first.
So why do I own none of them?
Two reasons. My capital is committed to assets I understand better and hold larger positions in. And gold has run extremely hard, which the table above makes uncomfortably clear. Also, for another 1-2 months, my liquidity is tight.
Either way, given my view on gold and my royalty focus, either FNV or WPM would look good in my portfolio.
If You Want Income From Gold
None of these three solves the income problem. Franco-Nevada yields around 0.8%, Wheaton around 0.7%, Royal Gold around 1%.
The instrument people ask me about is NEOS Gold High Income ETF (IAUI), so let me explain how it works and what it costs.
IAUI does not simply hold gold. It holds roughly 75% in Treasury bills, uses those as collateral for synthetic gold exposure through options, holds about 24% in a physical gold ETP, and runs an active covered call overlay on top. Its most recent monthly distribution was $0.4855 against a share price near $48, which annualizes to roughly 12%. The expense ratio is 0.78%. So far, the raw numbers.
That 12% is not a dividend, as gold produces no cash flow. The distribution is option premium and return of capital, which means part of what you receive is your own money coming back to you. That's why I only like this ETF to generate income from gold volatility. It's not a tool you should prefer if you are very bullish on gold.
Now the cost.
As I just implied, selling calls means selling away upside. Since launching in June 2025 through early 2026, IAUI delivered roughly half the price return that GLD did over the same window, and the gap was entirely the calls.
So think carefully about this. I have just spent 3,000 words arguing that gold is structurally supported by a multi-year debasement dynamic. An instrument that monetizes gold's upside is in direct tension with that argument.
IAUI makes sense if you need monthly cash and you expect gold to move sideways. It works against you in exactly the scenario I have described.
I would rather own the royalty companies and accept a 1% yield with dividend growth behind it. But if you need the income today, at least now you know precisely what you are paying for it. That's why I included it today.
Takeaway
The royalty model is the best way I have found to own gold as a business rather than as a rock.
Today, we discussed three companies with a combined market cap of roughly $107 billion, run by fewer people than a mid-size restaurant, with permanent claims on other people's mines.
Franco-Nevada is my top pick despite scoring second on my own framework, because the Panama situation is both its biggest weakness and its largest unpriced catalyst, and because the underlying business is cleaner on every cost and concentration metric that matters.
Wheaton has the growth, already signed, with concentration risk attached. Royal Gold is the cheapest and the newest in its current form.
But please read that ten-year return table again before you act on any of this. One of three beat the index. The metal itself had a shallower drawdown than either of the big streamers. And a single calendar year did most of the heavy lifting.
Needless to say, I will continue to cover this industry, as it's fascinating, especially in light of my Big Picture view on gold.
Stay tuned!