Warren Irwin has an easy line on gold and a harder one on mines. The easy line is the price. Governments spend. They print. Fiat buys less each year. On that path, he told Paul Harris at the 2026 Precious Metals Summit in Beaver Creek, $10,000 gold is a long-term call he is willing to make. He will not say when. The short run, he said, is anyone's guess.
The harder line is the one that pays, or fails to. A higher gold price covers mistakes. It does not create the value that is not the gold price. That value, in his view, is made by taking a development project and building a mine to a standard. The companies that wait until the first ounce is poured, then pay a premium on a price that has already multiplied, are buying someone else's work and someone else's errors. This piece is his argument, not a recommendation to buy or sell any metal, any share, or any project.
The price call is the weather
Irwin is president and chief investment officer of Rosseau Asset Management, which he founded in 1998. Harris asked him where gold is, after a high earlier in the year and a pullback he called a consolidation. Irwin did not draw a chart. He drew a habit. Canadian and American governments spend more than they take in. The habit shows up as money that is worth less. Gold, priced in that money, goes up over time. Trust, Harris added, is the other half. He pointed to the Netherlands, France, and Germany moving gold out of American vaults and back home. Irwin said he would do the same. He does not trust Washington, and he said so as a Canadian watching a trade fight and talk of a 51st state.
Take the repatriation as his and Harris's reading of trust, not as a fresh customs tally. Some of those shipments were decided years ago. The feeling they describe is current. It is still not a mine plan. A bar moved from New York to Frankfurt does not pour an ounce in Nevada or the Yukon. It tells you the people who store the bar are less sure of the custodian. Irwin's $10,000 is the same kind of sentence. It is a direction. It is not a date, and it is not a reason to pay any price for a developer this month.
Equities lag because the money is elsewhere
Harris noted that gold shares have risen and still look cheap against old ratios. He asked if the buyers have left for good, or if a catch-up is coming. Irwin's answer was funds flow. Big money, he said, is content in chips and other tech names. He has seen the turn before. When the tech froth broke in 2000, people looked at what they owned and moved toward real assets. What followed, in his memory, was a long run in metals.
He does not think that turn is finished. He thinks it has not really started. The S&P 500, as he described it at the summit, was at highs. The 10-year Treasury, as he put it, yielded about 5 percent, high enough to compete with stocks. His blind spot for the crowd is the bond. If fewer private buyers want the paper, and yields rise, the broad market can take what he called a kick in the teeth. Then the question he remembers from the early 2000s comes back. Do you want the hype, or do you want something solid? Weakness in a few tech names, he said, is not that moment. You will know the moment when it arrives, because it will not be subtle.
Until then, gold shares can lag the metal for a boring reason. The buyers who set the multiple are busy. A lag is not a proof that the shares are wrong. It is a proof that the flow has not rotated. Treating the lag as a permanent discount is how you miss the rotation. Treating it as a promise that every gold share must catch up is how you buy the ones that cannot.
Thirty percent is the floor
The summit was loud with studies. Harris said a lot of them used $3,500 gold as the base case and showed internal rates of return around 30 or 35 percent. The better projects, he said, were showing 60, 70, 80, even 100 percent. He asked if 30 percent is now a marginal mine.
Irwin said yes, it is the low end. Gold at the summit was, in his mouth, roughly $4,200. Harris later spoke of a price near $4,400. A $700 move in a year, Irwin said, is no longer a shock. Volatility is the new fact. His old goal was a return of 30 percent or better. He still would not go below that, and he would rather own the higher figures if he wants room for a downside spike in the metal. Thirty percent at a gold price $700 under the spot tape is not a cushion. It is a hope that the tape does not revisit the study.
He does not think the industry is in the manic phase yet. He has watched cycles in which mines were built that should never have been built. He hates the waste. He does not see that waste as out of control at a 30 percent study. The danger he named is later. If gold is $6,000 or $7,000 and studies are rewritten upward, and then the price gives back $1,000 or $2,000, the mines built on the high case are the ones that break. That is an overhyped market. He does not think this is it. He also said he would not envy the executive who has to pick $3,500 or $3,000 as the number in the model. The volatility is extraordinary. The model is a still photo of a moving price.
Costs eat the photo
Harris added inflation. If real costs are rising anywhere near the rate people feel, a margin that looks wide at $3,500 can vanish. Irwin's version was blunter. Nothing wrecks a valuation like a capital-cost overrun. Operating costs can do it too. Harris pointed at fuel and the Gulf. Irwin pointed at his own truck. Diesel that once priced under gasoline now costs him 80 cents a liter more. A fill-up ran near $300. That is one man's pump, not a national index. It is still the right unit. A mine burns diesel every day. A study that froze the fuel price is a study of a world that already left.
Put the two threats in one line. The gold price can fall toward the case you used. The costs can rise away from the case you used. Either one can take a 30 percent return to zero. Both can happen in the same year. That is why Irwin wants the higher return if he is going to live with the volatility. The extra percent is not greed. It is the gap between a study and a pour.
Where a miner actually adds value
This is the idea. Harris was surprised, at a summit full of news, by how little talk there was of takeovers. Juniors have raised money. They have drilled. They are less desperate to sell. Majors, he said, seem to want a project fully de-risked, which can mean built, before they bid. He asked when the right moment is. Once the first ounce is out, the value often jumps. A 40 percent premium on a company that has already doubled or tripled is a different check from a 40 percent premium before the first spade.
Irwin asked the question he thinks a mining company should ask itself. Where do we add value? His answer is the build. Taking a development asset and bringing it to production is a large part of the value that is not just the gold price. A rising price, from a few thousand dollars toward $5,000, covers a lot of mistakes. That, he said, is one of the biggest problems in the gold business. People can look clever on a spreadsheet while the metal does the work. A mining company, if it is one, builds and runs mines.
Buying a mine that is already built means you inherit the build, including the errors. You may run it a bit better. Scale might save you something in the low single digits. You do not get the three- or fourfold gain he associates with buying earlier, at the development stage, permitted or not, and designing the mine to your own standard. Decades of building, applied from the start, are the product. A junior's finished plant can be a plant you then spend years fixing, while you pay a multiple of what it would have cost a few years before.
He allowed one exception, and he called it rare. A project the street had talked about in the teens, as a developer, later changed hands nearer $8 after it was built. Waiting paid. He does not think that is the pattern. The pattern, he said, is the opposite. He is, in his words, 100 percent behind the view that a mining company's non-price value is the build.
Goose, and the hood
Harris offered B2Gold and the Goose mine as a recent picture. B2Gold bought the project, and the company that held it, before the build was finished. When it looked under the hood, Harris said, it found a lot it did not like, and spent time and money putting the plant right. The surprise, for him, was why the buyer did not move earlier and start the build on its own plan.
Irwin's reply was about the moment of the decision, not about the rear-view mirror. Everyone now says they should have bought developers, because the treasuries are full. When some of those decisions were made, gold was not at $5,000, and the cash was not there. The risk he cares about is the team. A major that has built mines for 20 or 30 years has a crew that has gelled. A junior that hires one person from Rio Tinto, one from BHP, and others from somewhere else, then sends them to a place none of them know, is assembling a first day, not a craft. That is where mistakes get made.
He would not, as a major, want an underground mine a junior had built, unless the job was simple. An open pit might be. Underground, he said, no. Harris then flipped the example. Fuerte Metals and the Coffee project in the Yukon came the other way. Newmont had advanced it. Tim Warman, Fuerte's chief executive, has described the test work as heavy, the kind a major does when it is not trying to save a season. Harris called the metallurgy bulletproof and said Fuerte inherits that work. Irwin agreed that the largest companies can gold-plate an asset past what a junior would spend. He has also seen junior builds he called disastrous. The inheritance is the point. The standard is set by who held the pencil when the plant was drawn, not by who holds the shares when the ribbon is cut.
Lazy cash is the cycle talking
If majors are not buying developers, Harris asked, what are they doing with the money? He put all-in sustaining costs near $1,900 an ounce and the gold price in the mid-$4,000s, a margin he called roughly $2,200 to $2,500 an ounce. Debt has been paid down. Dividends and buybacks are up. Cash is still piling up. He said at least 10 or 12 companies hold a billion dollars or more. That morning he had heard John Hathaway, a keynote at the same summit, call the balance sheets lazy, and ask why a gold miner does not hold some of the cash as bullion.
Irwin's version of the bullion idea was operational. Hold back some production. Keep it in a vault. He has seen an executive, over many years, keep a real share of retained earnings in metal. If you mine gold and you believe the Irwin line on fiat, holding the product is not foolish. He did not say every board should do it tomorrow. He said it is not a dumb use of a belief.
The cash pile, he said, is the cycle, and the cycle lies. Investors who wanted focus and sales at the bottom will want deals at the top. They will call unused cash lazy. Companies will buy the projects that are for sale, which are often the ones that do not matter, at prices that only make sense while the tape is hot. Then the downturn comes, and the same investors will ask why the company owns all these stupid little projects. His advice to the majors was to ignore that pressure. Have a long view. Buy when it makes sense. Sell when it makes sense. He included himself in the indictment. Investors, he said, are manic. He is one, and he still thinks a lot of what comes out of their mouths is crazy.
Harris drew the moral in one line. Investors are never satisfied. Irwin agreed. A management team that can tell them to relax is doing the job. The mining cycle is the mining cycle. It goes up and down. Standing in front of it with a slogan does not flatten it.
The junior cannot say the quiet part
The same courage, Irwin said, is harder in a junior. Harris asked why you so rarely see a developer stand up and say: we should not build this, we are the wrong team, the time is wrong, the project is marginal. Irwin has been the shareholder in that room. His answer was uncomfortable. They have to fake it until they make it. They have to look as if they can build, while hoping they never have to, because they do not want to build. If they hang a for-sale sign on a project and then stop, the majors wait them out. Three years of no permitting and no environmental work, and the seller is dead in the water. The buyer knows it.
So the junior that lacks the skill cannot admit the lack. It has to keep moving the project inside the skill it does have. That is not noble. It is the trap. The major that waits for a finished mine, and the junior that pretends it can finish one, are the same mistake seen from two chairs. One overpays for a plant it should have drawn. The other draws a plant it should have sold. The gold price, while it is high, lets both of them look right for a while.
What he would actually own
Harris closed by asking how to split a precious-metals portfolio. Bullion, royalties, large producers, developers, explorers. Irwin said he lives at the high-risk end. The meetings that excite him at a summit are the ones where someone says they are starting a company around a property they just got. That is where he has made his money. In early. In size. A retail investor, he said, probably wants some of everything. Royalties and seniors for cash flow, and maybe a dividend. Exploration for the upside that can be large. He does not think he adds value by buying Barrick or Agnico. He adds it, in his own book, by owning the early names that become large and, if it works, get bought.
Read that as a description of one fund's edge, not as a menu. Early and big is also early and wrong, often. A new company with a great story at a summit is the shape of both his winners and the sector's losses. The diversification he offered the retail listener is the admission that his own method is a job, not a default. If you cannot tell a build team that has gelled from a build team that met last month, you are not in his trade. You are in the gold-price trade, and he already told you that trade is the easy one.
The idea, once
Irwin will take $10,000 gold over time because he thinks the money will be worth less, and he will not tell you the year. He thinks gold shares lag because the big money is still in tech, and that a bond yield near 5 percent can be the thing that knocks that market down and sends people looking for real assets. He thinks 30 percent at $3,500 is the bottom of what he would own, because the metal moves too fast and costs do not sit still. He does not think the bad mines are being built yet. He thinks they will be if the studies are rewritten at $6,000 and $7,000.
None of that is the return he is pointing at. The return that is not the gold price is the build. A major that buys the finished mine pays for the wait and inherits the errors. A major that buys the project and builds it to its own standard is doing the thing he thinks the business is. A junior that cannot build, and will not say so, is hoping to be bought before the truth is due. Cash on a balance sheet will be called lazy until it is spent on the wrong assets, and then it will be called stupid. Ignore the shout. The cycle does not care that the gold price made everyone look smart on the way up.
A higher price covers mistakes. It does not create the value. If you remember one line from Beaver Creek, remember that one.
A note on sources and limits
This account follows a Kitco Mining interview at the Precious Metals Summit in Beaver Creek, Colorado, in September 2026, with Paul Harris and Warren Irwin, president and chief investment officer of Rosseau Asset Management. The $10,000 figure, the 30 percent return floor, the $3,500 study price, the gold prices near $4,200 and $4,400, the 10-year yield near 5 percent, the $1,900 sustaining cost, and the margin near $2,200 to $2,500 are the speakers' figures from that conversation, not a live tape re-checked for this piece. Irwin's diesel fill-up is his own anecdote.
B2Gold's Goose mine, acquired with Sabina Gold and Silver before construction was finished, is the example Harris raised. Public reports since have described a delayed ramp and extra spending to fix crushing and other plant issues. This piece does not audit that spending. Fuerte Metals, led by Tim Warman, holds the Coffee project in Yukon after a purchase from Newmont. Harris described Newmont's test work as the asset Fuerte inherits. A feasibility study and a construction decision were still ahead in earlier company comments. Do not read the interview as a pour.
John Hathaway was a keynote at the 2026 summit. Harris attributed the "lazy balance sheet" line to a conversation with him that morning. European gold repatriation is Harris's example of trust. It is not a 2026 inventory. Irwin's rare case of a project talked about in the teens and bought nearer $8 after it was built is left unnamed here, because the recording does not make the company clear.
Nothing here is investment advice or a solicitation to buy or sell any security or metal. A $10,000 call can be early by a decade. A 30 percent study can fail. Build risk is risk. Readers should read the studies and speak with a licensed adviser before any decision.