12:19 1. Replace single-name shorts with a put overlay — and trade it against the tape
The repeatable method
- Accept the structural problem first: a short single name has capped gains and uncapped losses, and post-2021 that arithmetic broke most short books. "It is hard. It is hard."
- Keep the equity book net long on your best fundamental ideas — do not neutralise it stock by stock.
- Buy index puts (plus a small number of single-name puts on the most extended leaders) sized so that, if the index falls, total net exposure is below zero, not merely reduced.
- Trade the overlay counter-cyclically: increase put exposure after a big rally (protection is cheapest and least wanted then), sell puts into declines (when it is dearest and everyone wants it). "We buy on strength and we sell on weakness."
- Budget it as an ongoing premium, not a timing bet: "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."
Here: equities net long — "most of the net long comes from gold and oil" — with S&P 500 puts (
SPX) plus
NVDA puts taking the book "well below zero on a net basis" (
27:48). His summary: "we're definitely ready if something really bad happens quick."
Watch for
- Sharp index rallies as the moment to add protection, not to reduce it; selling puts into a drawdown as the source of cash for buying stock.
25:16 2. Compute a short's real carry — the cash rebate, not just the dividend
The repeatable method
- When sizing a short, do not stop at "I have to pay the dividend." Shorting delivers cash to the account, and that cash earns the prevailing short rate.
- Net carry = (cash rate on short proceeds) − (dividend yield paid away) − (borrow fee, where the name is hard to borrow).
- Compare that number by jurisdiction before choosing where to express a macro short — the same trade can be free in one currency and paid in another.
- Only then decide whether the position needs to fall to be worth holding, or whether flat is already a positive return.
Here: Canadian banks — cash ~2.5%, dividend ~2.5%, so "you're flush on that." In the US at ~5% rates: "when you're short you actually make 5% on your cash. So the stock stays flat, your net return is actually positive 5%."
Watch for
- The spread between short-rate rebates and dividend yields by market; high-rate jurisdictions make shorting low-yield, expensive names close to a free option.
25:42 3. Classify every short before you put it on — four types, four exit rules
The repeatable method
- Tactical: a stock or sector that has just rallied hard; target a 5-10% pullback, "which happens all the time." Exit on the pullback, not on a thesis.
- Valuation: "things that I think are massively overvalued and overappreciated." Exit when the multiple normalises.
- Macro/sector: an expression of a top-down view. Exit when the macro view changes, regardless of price.
- Exposure management: index shorts sized purely to dial net exposure up or down. Exit is continuous, not event-driven.
- Write down which one each position is, because the exit rule differs — and never let a tactical short quietly become a valuation short.
Here: SOXX is an exposure/valuation short managed against the tape — "if it moved up from here I probably short more. If it went down 10, 15% here I'll probably cover half of it" (
26:27) — while the gold shorts are pure hedges against his own longs: "if we had a strong rally I short more. If we had a big pullback then I cover" (
11:40).
Watch for
- Any short you cannot classify — that is usually a directional bet you have not admitted to.
46:09 4. The dead-price screen — a growing company at a fire sale
The repeatable method
- Screen for stocks that have gone nowhere for three to five years — the flat chart is the filter, not a disqualifier. "There's a little kind of a style to my picks."
- Test the business, not the chart: is revenue still growing and is it still generating free cash flow every quarter? If yes, "eventually the stock should rise."
- Diagnose the specific reason the price is dead and check whether it has an end date — a debt build-up now paid down, an acquisition pause now over, a cyclical end-market at a trough.
- Demand a valuation that pays you to wait (single-digit PE, mid-single-digit EV/EBITDA, real free cash flow) so the holding period costs nothing.
- Identify the catalyst but refuse to time it: "I don't know if they will make an acquisition tomorrow morning, next week, next month, but eventually they will."
- Explicitly list what you are not paying for, and keep it as upside rather than as thesis.
Here: ADEN.TO — flat for five years, 8.5x PE / 6x EBITDA, balance sheet repaired, next acquisition pending, with a housing recovery deliberately excluded: "that's the extra kicker. No, I'm not betting on that." The precedent he cites is
TOT.TO, which "had not done much for a while" before tripling (
41:46).
Watch for
- Balance-sheet repair completing at a serial acquirer; a fragmented industry where the consolidator has a track record; margin expansion left as unpriced optionality.
48:03 5. Buy the multiple compression that has no earnings compression behind it
The repeatable method
- Screen for a stock down heavily over a year where earnings and consensus estimates went up over the same period. The gap is entirely a change in what people will pay.
- Name the narrative driving the de-rate in one sentence. If you cannot, you do not understand the setup.
- Test the narrative against a historical analog rather than arguing it in the abstract — find the last technology that was supposed to eliminate this activity and check what actually happened to volumes.
- Recompute the return with the narrative assumed wrong but unresolved: what free-cash-flow yield do you collect while nothing changes, and what does management do with it?
- Treat a resolution of the fear, or the cyclical recovery, as a second and third source of return — never as the base case.
Here: CIGI down 42% while profits and estimates rose, on the fear that AI eliminates brokerage, property management and engineering. His analog is CAD software: computers were supposed to end engineering firms and "engineering work just continued to grow and grow and grow" (
49:16) — leaving 12x earnings, an ~8% free-cash-flow yield reinvested into acquisitions, and a return-to-office recovery as the kicker.
Watch for
- Sectors being repriced on an AI-disruption story while reporting rising earnings; the prior-technology analog (CAD, the internet, spreadsheets) as the cheapest available test of the claim.
51:08 6. Price an "expensive" grower by peer bracket and doubling time
The repeatable method
- Decompose the growth into independent engines and add them — unit/store count growth plus same-unit sales growth — so the headline rate is built, not quoted.
- Check for a leg not yet in the numbers (a new product line only partly rolled out) and note it separately as unpriced.
- Bracket the multiple against three peers: one no-growth incumbent, and two higher-multiple growers in the same category. If the incumbent trades near your name with none of the growth, the multiple is not the objection people think it is.
- Convert the multiple to a doubling time: at g% compound growth, earnings double in roughly 72/g years, so today's PE halves on that horizon with no re-rating at all.
- Say out loud that it is expensive. If the position only works when you avoid that sentence, it does not work.
Here: BROS — 17% store-count growth + 5-8% same-store sales = 20%+, food only now rolling out, at 42x. Bracketed against SBUX 34x and not growing, CAVA 92x, SHAK 52x. "If it grows at 20% a year, it means it will double in about three and a half years. So the 42 in three and a half will be 21."
Watch for
- A no-growth incumbent trading within ~25% of your grower's multiple; category peers at 2x your multiple with no better growth.
12:55 7. Borrow the technical diligence you cannot do yourself — then buy the catalysts
The repeatable method
- Admit which parts of the analysis you genuinely cannot perform. In mining that is geology, ore quality and metallurgy — "in mining there's a lot of issues that can come over time."
- Maintain a short list of named specialists whose own capital is committed and whose track record you can check, and screen for companies they have financed or joined.
- Use their participation as a pass on the technical questions only — "the fact that some of these people have done the work, I feel more comfortable" — never as a pass on valuation or on the catalyst path.
- Then apply your own screen: can you enumerate the catalysts ahead — permitting, PEAs, final investment decisions, new production areas — and do you know, or know someone who knows, the board and management?
- Prefer the small company where those catalysts move the whole equity, over the large producer where they do not.
Here: MLM.CN (McFarlane Lake Mining), a sub-$300m Ontario gold junior, pre-vetted by investors he follows — Michael Gentile, "Cohen at Scotia," Lassonde — became one of his largest recent positions instead of buying
AEM as the safe play. Same principle at
NICU, where $100m from an operating-miner investor group plus the Dundee group did the technical work (
39:42).
Watch for
- Financings led by specialist investors who own and operate mines themselves; a queue of dated catalysts you can list in advance rather than a story about ounces in the ground.
34:09 8. Insider selling as a veto, not a data point
The repeatable method
- Pull the insider transaction history before the management meeting, so the meeting cannot talk you out of it. "This is something we do at Timelo quite a bit. We follow insider buying and selling very closely."
- Distinguish one-off sales (tax, a house, diversification) from a pattern — "insiders sell regularly" is the disqualifying observation, not any single filing.
- Weight it hardest where the insiders' informational edge is largest: lenders (loan-book quality), miners (project reality), and any fast-growing business whose accounting runs ahead of cash.
- Let it override an impressive meeting. Enthusiasm from management is expected; selling by the same people is information.
- Apply it symmetrically to founders — a founder who built a good company and is now selling is still selling.
Here: PRL.TO — he met management for the first time, was impressed by the growth and the enthusiasm, and passed solely because "I see insiders there… I'm just not comfortable because of that." The same screen is the third and decisive reason he is staying away from
CIA.TO, whose founder "has also been selling shares" (
35:07).
Watch for
- Sustained insider distribution at a fast-growing lender or a founder-led resource company; treat a management meeting scheduled after a selling run as confirmation work, not discovery.
10:36 9. Express a theme as a relative-value pair, not a direction
The repeatable method
- When a theme has already been discovered (here: "if you like the AI trade, buy copper"), stop asking whether it is right and start asking what has been repriced for it.
- Compare two sub-sectors of the same asset class on the same three measures — price/NAV, price/cash flow, PE — rather than forecasting either commodity.
- Go long the cheaper sub-sector and short the dearer one, so the position pays on convergence and does not require the commodity call to be right.
- Confirm the crowding empirically: look for days when the "cheap" side and the "expensive" side decouple with the theme rather than with their own fundamentals.
- State the trigger and refuse to state the date: "it could be in 27, maybe it's in 28. That's a real debate. In my mind it's not if."
Here: long
GDX-type gold equities ("extremely cheap"), short some copper equities (
COPX) on much higher multiples — with the evidence of crowding taken straight off the tape: on a day 75% of stocks fell with the AI trade on, "copper stocks, many of them up 10%. That's kind of an AI stock, AI area" (
9:10).
Watch for
- A commodity sub-sector starting to trade on AI-trade days rather than on its own inventory data — that is the tell it has been re-rated as a theme, not as a business.
18:42 10. In a contested deal, separate the vote from the trade
The repeatable method
- Judge the governance question on its merits — is the price a take-under, is there a conflict (management or a founder buying assets from the company they run), are there tax consequences for holders? Read the independent sell-side work before the circular arrives.
- Decide your vote from that, and say so publicly if you are a holder.
- Then price the security separately under both outcomes: what do you receive if the deal closes at today's price, and where does the stock trade if holders reject it?
- Only take the position if both branches are acceptable at the current price. "Either way, I'm comfortable either scenario in terms of the current price."
- Look through the consideration: if part of the payment is stock in the acquirer, value the acquirer independently — a discount to a deal that itself pays you in a cheap security is a double discount.
- Be honest about the vote's odds. Heavy retail ownership usually means a deal passes regardless of how the institutions feel.
Here: HR.UN — "it's really a take under," he expects to vote against, and he has been buying since the announcement because the stock at ~9.75 against ~12 of stated value hands him units in GO.U, which he separately thinks is cheap: "GO is cheap and then you get it under, even cheaper going through."
Watch for
- Break-up deals where management or the founder's family takes assets; the spread to stated deal value versus the standalone recovery price; the independent-broker report (here, TD's) published before the circular.
29:23 11. Size to the weaker of business conviction and timing conviction
The repeatable method
- Score the business and the timing as two separate questions. A near-monopoly with a new, impressive management team can still be two years from an inflection.
- When a new CEO is "reviewing every aspect of the business," assume the earnings are flat, not rising, while that runs: "you can't do everything at once."
- Take a real but small position — enough to keep you doing the work and to own the eventual re-rate — and state explicitly why it is not larger.
- Set the trigger to size up: evidence the restructuring is finished and earnings growth has resumed, not a further fall in the price.
Here: CAE.TO — "I really believe it's really really well managed and it's almost a monopoly… So it's a good business to own long term," but "it's not a big position yet because I'm not sure it's ready to take off soon because of all these restructuring and changes."
Watch for
- The quarter where a turnaround's restructuring charges stop and revenue growth resumes; conversely, a full-size position in a company whose timing you would not defend.
31:12 12. Expensive is not the same as shortable — and a macro view is a sector veto, not a stock call
The repeatable method
- Separate "I won't buy this" from "I'd sell this short." Quality, a clean balance sheet, excited buyers and a live acquisition pipeline are each reasons a rich stock keeps going up.
- When you decline a name on price alone, say whether you are short it. If the answer is no, that is a discipline worth recording, not an inconsistency.
- Run the top-down view as a sector filter applied before the bottom-up work, so you are not talked into a good company in a sector you have already vetoed.
- State the veto in one line — "I'm a little bearish on the economy so I don't want to own a lender" — and apply it consistently, including to the names that screen cheapest.
Here: TIH.TO — "it's a great company. I just find it's too expensive… I'll reveal that we're not short it either." And the economy veto knocks out the whole lending sector regardless of the individual case:
GSY.TO (
33:19) and
PRL.TO, with "lending money is easy. Getting it back is a little harder" as the standing reminder.
Watch for
- A cheap lender screening well into a slowdown — the classic value trap the macro veto exists to catch; and any short whose only argument is the multiple.
39:17 13. Do the primary call — find the person building it, not the person selling it
The repeatable method
- Before forming a view on a small company, check your own network against its staff list — not the CEO and IR, but the operating people: development, engineering, mine planning.
- Call the one you know. Half an hour with someone who has no stock to sell you is worth more than the deck.
- Use the call to establish the sequence of production/development areas and their timing — the catalyst queue — rather than the story.
- Cross-check with who has already put capital in and what they know: an investor group that operates mines can judge the asset technically in a way a generalist fund cannot.
- Do not buy the same week. He liked it and still said "we don't own it now because we just talked to them last week."
Here: NICU — he remembered a Bay Street contact, found him working as Magna's VP development, "chatted for like half an hour," and came away with the second/third/fourth/fifth production areas as the catalyst path, plus $100m from operating miners and the Dundee group as validation.
Watch for
- Operating-company insiders inside your own network; financings led by people who run mines rather than by generalist funds.