Jean-François Tardif is short Canadian banks, semiconductors and the S&P 500 through puts, and he says the fund is "definitely ready if something really bad happen quick."
Most of the conversation about AI spending treats the bull and bear cases as opposites. Tardif treats them as sequential. Nvidia's own guidance points to nearly $1.7 trillion of sector spending if growth continues at its current pace, which he doesn't think is sustainable — and he isn't betting on when the top arrives so much as insuring against what happens after it.
"I think we're very much coming to the top the peak of spending."
Tardif is president of Timelo Investment Management, which he founded in 2012 after leaving Sprott, and specializes in idiosyncratic small-cap ideas with a commodity tilt; two of his three stock picks from his last appearance on the show have roughly tripled and risen more than 50% since.
I listened to the full episode so you can skip it. 55 minutes of audio, 15 minutes of reading.
Here are the 11 takeaways that matter.
👤 Guest: Jean-François Tardif, president of Timelo Investment Management and a former Sprott portfolio manager
🎙️ Host: Amber Kanwar, a former BNN Bloomberg anchor with over a decade covering Canadian markets, who now hosts In the Money
📰 Published: 10 September 2026, on YouTube
🔴 YouTube | 🟣 Apple Podcasts | ⏱️ 55 min | ✅ Time saved: 40 min
Key Takeaways
He's positioned for a hard landing in AI spending, without a timeline for when it hits
Nvidia's guided 70% growth implies nearly $1.7 trillion of sector spending, which he calls unsustainable
Gold stocks are extremely cheap relative to copper stocks right now, on his numbers
He's short some copper names and buying small-cap gold on other analysts' homework
He switched from outright shorts to index puts in 2021 to lower the cost of being wrong
The position costs little monthly if he's wrong, and pays off fast if markets fall
He expects to vote against the H&R REIT "take-under," and bought more of the stock after it was announced
TD's own research flagged the deal as undervalued with a conflict-of-interest angle
Canadian banks are trading at their most expensive multiples in decades on trading and wealth income that is itself cyclical
Two of his three prior stock picks roughly tripled or gained over 50% since his last appearance
Total Energy Services and Calian Group; the third, H&R REIT, is still working out
His three new ideas are all cheap-and-growing names trading at fire-sale multiples
Adentra, Colliers International and Dutch Bros
1. NAFTA Is the Real Risk Line
Kanwar opened on the deteriorating tone between Canada and the US and asked how quickly tariffs had moved up Tardif's concern list.
He isn't worried yet because the scale of the announcements is small relative to GDP, and NAFTA itself is still intact. "So I would say as long as NAFTA still stays... that would be the time to really worry."
A full NAFTA breakdown is the actual risk line, and he doesn't see it as likely. "If NAFTA goes away completely, that would be a disaster for all our Canadian exporters companies," though he added that with the two countries not talking, the tariff fight is already "causing real pain" for auto-parts makers such as Linamar and Magna
He isn't repositioning the portfolio around tariffs specifically. Most companies he holds that would be hurt by a real NAFTA breakdown already operate their people and offices in the US, which insulates them regardless of nominal incorporation
He also pointed out that the US actually runs a trade deficit with Canada once oil is excluded — a detail he joked the president should hear
2. Betting Against the AI Peak
Kanwar asked how the AI trade shows up in a small-cap, commodity-leaning portfolio given the sector's swings between euphoria and fear.
He frames the trade as two camps, both defensible on their own terms. The bulls point to real earnings growth, and Nvidia guided 70% growth for next year. Tardif did that arithmetic on air: "Well, that will make about close to $1.7 trillion of spending overall if the whole sector grows another 70%." Micron's earnings are booming, he said, so owning the growth is not wrong on its face
The bear case is that the spending curve eventually turns, and he's in that camp. "I really don't believe that 1.7 trillion spending is sustainable," and when spending turns down, he expects a hard fall in the stocks tied to it
He's already positioned for the turn through index-level shorts rather than single names. He is short the SOX semiconductor index and holds puts on the S&P 500 and Nvidia specifically, adjusting the size as the market moves
The same logic extends to the copper trade some investors treat as an AI hedge. "Copper stocks are likely to go down once we go down the other side of the mountain and capex starts to go down," which he expects in 2027 or 2028 — "That's a real debate when in my mind it's not if"
3. Gold Over Copper
He said gold stocks are extremely cheap versus copper stocks right now on a price-to-NAV, price-to-cash-flow or P/E basis. The fund owns a small copper position but is also short some copper names on the relative-value view
His gold buys skew toward smaller names with visible catalysts — permitting decisions, final investment decisions and similar events — rather than the safer large-cap route other guests on the show have recommended
His highest-conviction current gold name is McFarlane Gold, a sub-$300 million market cap he bought recently on the strength of other people's diligence. He named Michael Gentile and an analyst at Scotia among the people whose prior work on the name gave him comfort: "They already know the engineering is fine. The ore is good."
4. Puts Instead of Shorts
Kanwar asked about the mechanics and risk of shorting in this environment.
He reduced his number of outright shorts starting in 2021 and now leans on index puts instead, buying more protection after a rally and selling into a pullback — "we buy on strengths and we sell on weakness just to help us lower our risk and our net exposure to the market"
The strategy is asymmetric by design. "I have very low cost because if we're wrong on the options it's a very small cost every month but if we're right and faster the market goes down better it is for options then potentially we could make a lot of money"
In Canada the cost of carrying a short is roughly offset by interest on the cash proceeds — in the US it can be a net positive. With Canadian cash yielding around 2.5%, matching the roughly 2.5% dividend on the banks he might short, "you're flush on that." In the US, where rates run near 5% against a lower dividend yield, the carry works in the short seller's favor: "When you're short you actually make 5% on your cash. So the stock stays flat your net return is actually positive 5%."
He runs four distinct kinds of short, and said the current one is an index bet rather than a single-stock call. A short-term short against a sector that has rallied and where he expects a 5-10% pullback; a short against something he thinks is massively overvalued; a sector short expressing a macro view; and an index short to change his overall market exposure. The current semiconductor position is the last kind: he is short the SOX index rather than named chipmakers, and would short more on a rally and cover half on a 10-15% fall
The book overall stays net long, mostly through gold and oil positions, with the options overlay taking the net exposure down without making the fund outright bearish. "We're definitely ready if something really bad happen quick. We're definitely protected"
5. The H&R "Take-Under"
Kanwar revisited Tardif's prior H&R REIT pick, which fell 12% into a contested sale after activists failed to find a single buyer over two years.
He called the deal a "take-under" rather than a takeover, in which GO Residential bought the residential portfolio, Blackstone took the industrial assets, and founder Tom Hofstedter kept the non-core pieces
He had already trimmed the position last year on valuation concerns before the deal was announced. "I sold not all but I did reduce my position at H&R at the time," citing a prior podcast appearance by real estate specialists who didn't see much upside above $12 a share
He flagged a conflict-of-interest concern around the founder's participation in the deal, and pointed listeners to a TD research report that he said reached the same conclusion he did — that the assets look undervalued and the structure raises questions
He expects to vote against the deal as it stands, pending the formal circular, but said he's comfortable either way: buying more after the announcement because GO Residential's own shares look cheap and the takeover price adds an extra discount. "I don't think there's a lot of risk here"
6. Banks: Harvest the Gains
A viewer question on the Canadian banks prompted a broader discussion of how expensive the sector has become.
The banks are trading at their most expensive multiples in multiple decades on both price-to-earnings and price-to-book, despite a weak housing market and a slowing economy. He credited the strength to non-lending revenue — insurance, portfolio management and trading — which he called cyclical rather than structural
His advice for a tax-sheltered holder is to harvest the gain rather than hold through a downturn. "If you get really nervous 6 month a year from now not owning the banks, they'll probably be at the same price or lower. So buy them back later"
He pushed back on the idea that wealth management has made the banks structurally less cyclical. A bear market cuts assets under management directly, and lower trading volume cuts the other leg of that "non-cyclical" income too
He is short some of the banks but declined to name which ones, framing it the same way as his broader index shorts — a bet that today's valuation premium isn't being compensated for the risk
7. Mailbag: CAE and Toromont
Two viewer questions on industrials turned into a contrast between a name he owns and one he thinks is priced for perfection.
CAE's new CEO impressed him enough to start a position, even without an earnings catalyst yet. Matthew Bromberg, who came from large US companies, is "reviewing every aspect of the business," and Tardif called the flight-simulation business "almost a monopoly" worth owning long-term even if earnings stay flat for a couple of years while the turnaround plays out
He isn't worried about pilot demand or airline budget pressure derailing that view. Pilot training demand grows with global air travel and is mandated by law, and CAE's smaller military-training business benefits as Europe and Canada raise defense spending
Toromont, a Caterpillar dealership he called a decent AI-buildout proxy, is a name he likes but won't buy at today's price. The balance sheet is strong — close to net cash — and the company has been signaling acquisitions for a while, but the multiple has already re-rated on AI-linked demand for its second business line. "I just find it's too expensive. So me personally, we don't own it"
8. Lenders and Cannabis
On the subprime consumer lenders goeasy and Propel, he stayed away from both — one on valuation, the other on insider selling. goeasy's core lending book is fine, he said, but its move into used-auto lending caused real problems, and with the stock now above book value and his own caution on the economy, "it's not for me it's complicated." Propel's management, whom he met for the first time recently, sounded "extremely enthused" about the business, but insiders there sell regularly — something Timelo tracks closely — and that alone made him uncomfortable
In cannabis, he owns Village Farms for its balance sheet and growing European business, not for a sector recovery bet. The company has net cash and is already profitable, at 25-30 cents of expected earnings next year against a roughly $3 US stock price. It just finished a production expansion in British Columbia as the lowest-cost Canadian producer and is growing sales into Germany's expanding cannabis market, where domestic growers currently just import raw material
He also sees Village Farms as the sector's likeliest takeover target, given its combined US, Canadian and European asset base, pointing to the pending Curaleaf-Aurora Cannabis deal as evidence consolidation is underway. A second, smaller cannabis holding, Trulieve, is held purely as an index-inclusion arbitrage rather than a fundamental pick
9. Iron Ore and Magna Mining
He's watching Champion Iron and Labrador Iron Ore rather than owning either, with both stocks, in his words, "kind of in the toilet." A big new mine coming into production in Africa and his own caution on the broader economy both argue for staying on the sidelines, and he flagged that Champion Iron's founder — who built the company early and made a great deal of money doing it — has also been selling shares. He wants to see where iron ore bottoms before buying. Between the two he called Labrador Iron Ore the less risky vehicle, because about half its long-run income comes from a royalty rather than from running a mine, and neither is in the fund today
He was more constructive on Magna Mining after a call with someone he knows inside the company. The company is producing copper and nickel from one area and has three more areas coming into production, a growth pipeline he called a source of catalysts. He noted a $100 million investment from a group that owns and operates mines itself and places its surplus cash flow elsewhere, and pointed to a second group of long-time mining investors on the register, as reasons the setup "looks promising." The company is run by founder and CEO Jason Jessup, and Tardif's fund does not own it yet
10. Two Old Picks Revisited
Kanwar checked in on two of Tardif's three picks from his prior appearance, both of which have performed strongly.
Total Energy Services has roughly tripled since he named it at $11 a share, and he still holds it. His original call was for earnings to roughly double to $4 within three to five years; the company's earnings are already around $3, a year into that window, on capacity expansion in US compression and service-rig growth that has made it the largest operator of its kind in Australia. He thinks the earnings target could arrive sooner than originally expected, helped by a strong balance sheet and a founder-CEO he called disciplined and very acquisitive
Calian Group is up more than 50%, largely on rising government training and defense spending as Canada and Europe increase military budgets. The stock traded around 10 times earnings when he first pitched it; it's closer to 14 times now, but he still holds it as earnings keep growing and the company has signaled further acquisitions
11. Three Cheap Growth Names
Tardif closed with three new picks, all growing businesses he says the market is pricing as if they aren't.
Adentra, a fragmented lumber and building-products consolidator, trades at 8.5 times earnings and 6 times EBITDA while generating consistent free cash flow. The company made acquisitions during the COVID housing boom and took on debt; with the balance sheet now repaired and housing weak, he's betting on the next acquisition rather than a housing recovery. "I'm not betting on that. I think the stock is cheap here with housing not performing"
Colliers International is down 42% over the past year despite growing earnings, which he called an unusually attractive setup for a well-managed, multi-decade compounder still run by founder Jay Hennick. At roughly 12 times earnings — one of the cheapest multiples in its history — he's skeptical of the market's fear that AI displaces property brokerage and engineering services, drawing a parallel to how CAD software made engineering firms grow rather than shrink. A recovery in office occupancy would be a further tailwind he isn't counting on to make the thesis work
Dutch Bros is the most expensive of the three at roughly 42 times earnings, but he argues the growth justifies more, not less. Store count is growing 17% a year with same-store sales up 5-8%, putting total growth above 20% even before a new food menu rolls out more broadly. He compared the multiple favorably to Starbucks at 34 times with no growth, and to more expensive fast-casual comparisons like Cava at 92 times and Shake Shack at 52 times, arguing Dutch Bros' growth rate would normally command a much richer US multiple than it gets
Bonus Insights
Kanwar disclosed at the top of the show that she owns two of the stocks discussed, CAE and GO Residential, and said later in the H&R segment that she holds GO directly
Tardif pointed to H&R's own Primaris spin-out as the template for what could still be done with the rest of the trust. When H&R spun the mall business out it installed an ex-Bay Street analyst, Alex Avery, as chief executive, which Tardif called a smart move and said has worked; one option, he said, would be to spin out another division under new management and grow it rather than sell everything now
Kanwar said Cole Smead of Smead Capital, a US-based minority holder who has appeared on the show, is also unhappy with the H&R deal, and put the practical problem to Tardif: after a two-year sale process with no single buyer, a vocal dissident has little to win
Tardif and Rick Rule never overlapped at Sprott, he said, correcting the host's timeline: he left before Rule arrived, and he founded Timelo in 2012 rather than 2009
On the show's closing question, he said he has golfed since he was a child and snowmobiles in winter, on a Ski-Doo. He once played a round with Mike Weir in his junior years — "he was better than me" — and puts his own handicap at about five. Kanwar said that when someone once asked her the same question, she answered "putting"
Tardif's bottom line is that the portfolio is built to be net long through gold and oil while carrying enough short and options protection through the banks, semiconductors and the index to make money if AI spending rolls over. He is not forecasting the date that happens; he is paying a small monthly cost to be paid when it does.
Products, Companies & Tools Mentioned
Nvidia (The guided 70% growth figure behind Tardif's ~$1.7 trillion sector-spending estimate, and one of the names he holds puts against)
H&R REIT (His prior pick, now being sold off to GO Residential, Blackstone and the founder in a deal he calls a "take-under")
GO Residential (Buying H&R's residential portfolio; Tardif also owns GO shares directly, which he calls cheap)
Total Energy Services (His prior top pick, up roughly threefold since $11, on US compression expansion and Australian service-rig growth)
Calian Group (Up more than 50% on rising government training and defense spending in Canada and Europe)
CAE (Flight-simulation company he started buying on new CEO Matthew Bromberg's turnaround, calling the business "almost a monopoly")
Toromont Industries (A Caterpillar dealership and AI-buildout proxy he likes but calls too expensive to own)
goeasy and Propel Holdings (Subprime consumer lenders he avoids — one on used-auto losses, the other on persistent insider selling)
Village Farms International (His main cannabis holding, on its balance sheet, European growth and takeover potential)
Curaleaf and Aurora Cannabis (Curaleaf's pending bid for Aurora, cited as evidence of cannabis-sector consolidation)
Trulieve (A smaller cannabis holding, owned as an index-inclusion arbitrage rather than a fundamental pick)
Champion Iron and Labrador Iron Ore Royalty (Two iron-ore names Tardif is watching but not yet buying, waiting to see where the metal bottoms)
Magna Mining (A copper-and-nickel growth story he's constructive on after speaking with someone inside the company)
Adentra (A fragmented building-products consolidator trading at 8.5 times earnings that he's betting will make its next acquisition)
Colliers International (Down 42% over a year despite growing earnings; his pushback on the market's AI-disruption fear for property services)
Dutch Bros (A fast-growing drive-thru coffee chain he says is cheap relative to its 20%-plus growth rate)
If this was worth your time, send it to someone who has to have a view on this.
Get the latest market chatter as it happens: