Triple Flag Picks Up the Leverage Key

A quiet 25,000-GEO quarter, a re-rated royalty book — and a CEO who spent two calls saying $4.3 billion was too big now floating a 2x EBITDA balance sheet.

TFPM.TO · Earnings Call · 2026-11-04

The quarter that didn't need a headline

Triple Flag's third quarter was, on the numbers, exactly what the model promises: 25,000 GEOs, putting the year in the top half of the raised 100,000–110,000 GEO range; revenue of $107 million, up 14%; adjusted EBITDA of $94 million, up 22%; and operating cash flow per share of $0.45, up 15% from $0.39. Eban Bari was explicit that this is the metric that matters: “Operating cash flow per share is the metric that most directly compounds shareholders over time.” — Sheldon Vanderkooy, Chief Executive Officer · 2026-11-04 The company repurchased another $20 million of stock in the quarter and paid down $50 million of revolver drawings, exiting with $185 million drawn against $1.1 billion of available liquidity. Note what management would not do: give asset-level guidance. That refusal is asset-by-asset boilerplate now, ranked near the top of the company's keyword list this quarter and in the third quarter of 2025 alike. It is the one part of the call that is pure muscle memory — analysts ask about Ravenswood's run rate, about the 2030 build, about Q4, and get the same polite no.

The value-surfacing machine, running on schedule

Where the call actually earned its keep was in the portfolio, and the two big news items of October both landed in Triple Flag's favour without a dollar of new capital. SSR Mining published an updated Marigold technical report — a 20-year life of mine, 4.6 million ounces of net reserve, and a maiden mine plan for Buffalo Valley (670,000 ounces of reserve, first production 2029). Concurrently SSR agreed to buy Rockwell Resources, whose Converse project sits six kilometres from the Marigold Complex, hosts over 2 million indicated and 3 million inferred ounces, and carries first gold expected in 2033 at 150,000–170,000 ounces a year. Triple Flag holds a 3.35% NSR on Buffalo Valley and a 5% NSR on Converse — both bought cheap. James Dendle's framing was the whole investment case in one line: "These are royalties that were acquired at a very modest valuation... At no incremental cost to Triple Flag." The Polo Sur royalty is the same trade in Chile, where Antofagasta is now evaluating folding oxide material into the Centinela plan. And Northparkes is the same trade at larger scale: E48 ramping, E22 block cave in development, E44 gold-dominant material with minimum deliveries from 2030, and a mill study at 10–11 million tonnes per annum against a resource base of more than 630 million tonnes including inferred resources. Sheldon Vanderkooy leaned on a Scotiabank study to make the point that this is not narrative but realised return:

At the time we announced our investments, The Street ascribed an average IRR of 4.9%, Now with the benefit of hindsight, Scotiabank ascribes a realized IRR of 12.8% This 12.8% IRR leads our peer group.
Sheldon Vanderkooy, Chief Executive Officer · 2026-11-04

That is a fair brag. It is also the setup for the interesting part.

The change: the size ceiling came off

In February, answering the same question about billion-dollar royalties that peers were then discussing, Sheldon Vanderkooy was unambiguous: “We have always said that our sweet spot is really in the $200,000,000 to $500,000,000 range... $4,300,000,000 is too big for Triple Flag Precious Metals Corp.” — Sheldon Vanderkooy, Chief Executive Officer · 2026-02-19 He repeated a variant of that in November 2025 and again in August. On this call, pressed by Bank of America's Lawson Winder on how much trailing leverage the company would tolerate "on a short term basis in order to complete a larger transaction," the answer changed shape:

I think if you are at 2x EBITDA, you are in a zone of safety there. And I think the market is comfortable with that. And quite frankly, the cash flow generation of a portfolio like this, that could be de-levered. And so when you start to run those numbers, that actually allows us to look for, you know, quite large transactions that really move the needle.
Sheldon Vanderkooy, Chief Executive Officer · 2026-11-04

He then confirmed to Fahad Tariq that “we would look at larger transactions as well” — Sheldon Vanderkooy, Chief Executive Officer · 2026-11-04 — larger than the mid-size pipeline, not instead of it. The keyword evidence corroborates the shift in framing: large transactions enters this quarter's top tier at number two, having been a background phrase in earlier periods. Management has not committed to anything, and Sheldon immediately hedged with prudence and "not either/or." But the frame moved from we don't swing that big to here is the leverage we could use if we did. With $550 million deployed over the past year against $350 million the year before, and a pipeline he called "very robust," that is a meaningful widening of the addressable deal set.

What stopped being said

Three silences are louder than anything in the prepared remarks. First, ATO. For at least two consecutive calls the Steppe Gold stream dispute was a set piece — arrears quantified, parent guarantee discussed, an arbitration started, and a standing assurance that guidance would be met "even if we don't receive a single ounce." On this call, not one word. The Steppe thread that ranked in the fourth quarter of 2025 has simply vanished. Either it is resolved or it is being de-emphasised; management did not say which. Second, Cerro Lindo, the company's single highest-ranked keyword just one quarter ago, has dropped out of the top tier entirely. That is arguably success — the anticipated stream step-down has now happened, and the second-largest contributor this quarter was Northparkes alongside it, with Ravenswood at 6% of revenue in its first full quarter. But the asset that anchored the silver story for years is no longer a talking point. Third, Australia, once described by the CEO as the company's "single highest country concentration," is fading in the keyword set and appears on this call mainly as the jurisdiction Northparkes and Ravenswood happen to sit in, while revenue leadership has shifted to Latin American and Australian assets jointly. There is also a small, telling schedule slip. The Prieska stream decision has moved three times. In November 2025: “it's a funding decision for Triple Flag in 2026.” — James Dendle, Chief Operating Officer · 2025-11-05 In August 2026, on the deeps: “when the company moves towards an investment decision on the deeps, which we expect to be next year, we'll look to do our valuation and presumably invest the stream at that time.” — James Dendle, Chief Operating Officer · 2026-08-06 On Wednesday, to Lawson Winder: “we would expect to be making that decision late in 2027. We would draw down in 2028.” — James Edward Dendle, Chief Operating Officer · 2026-11-04 No drama, no explanation — just the option drifting two years to the right while the upper levels get built with Glencore's money.

The verdict

This was a strong, unremarkable quarter wrapped around one genuinely new idea: Triple Flag is now willing to talk about the balance sheet as a deal tool rather than a rounding error. Everything else — the G&A hedging tied to share-price-linked comp, the cash flow generation pitch, the annual dividend increase, the opportunistic clear value read on its own stock — is continuity. The growth profile toward 150,000–160,000 GEOs by 2030 is reiterated, not revised, with a 2031 view promised for February. What an investor has to decide is whether a royalty model with ~80%+ margins and modest leverage should be using 2x EBITDA to buy a much bigger asset at a moment when gold and silver royalties are competitively priced. Management's own history suggests they underwrite well; the Scotiabank IRR number suggests they have been patient and right. But patience on price and appetite on size are different muscles, and this is the first quarter the CEO has flexed the second one. Watch the next large transaction — and watch whether ATO stays silent.