Actionable insights — Portfolio Update: harvesting Hudbay (HBM), the cyclical-top trim discipline, the energy valuation-elevation screen

1. "Harvest" a cyclical winner near its top — trim to a core, don't let it round-trip

1. When a cyclical commodity producer is near all-time highs after a large gain, flip the question from "how much more upside?" to "how do I avoid giving a year's gains back?" The asset you own at the top is a different proposition than the one you bought at the bottom.
2. Take partial profits — trim enough to bank a meaningful chunk and de-risk a position that has grown large and cyclically extended — but keep a core stake for any genuine long-term (structural) thesis that survives a cyclical pullback.
3. Explicitly decide not to add at the highs, and pre-commit to re-adding only on a substantial pullback. The mental model: be "the patient buyer of the dip," never "the holder who rode the top all the way down."

• A big winner at/near all-time highs on a commodity at record levels; a consensus narrative that's gone from contrarian to crowded; gains large enough that a normal cyclical correction erases a year of work.

2. Value a cyclical on Price/Sales and mid-cycle earnings — not trailing P/E

1. Distrust a "reasonable" trailing P/E on a cyclical: the denominator is peak-cycle earnings (record output prices + record margins), so the multiple looks cheaper than the business really is.
2. Cross-check with Price/Sales, which is far less distorted by lofty, potentially unsustainable margins — a P/S near a multi-year high is a classic late-cyclical warning even when the P/E looks fine.
3. Re-run the valuation on mid-cycle metal/commodity prices; if it's "fully valued" there, the current cheapness is an artifact of the peak.

• Trailing P/E vs P/S divergence on a cyclical; P/S percentile vs its own 5-year range; what the multiple becomes on normalized (mid-cycle) prices.

3. Find the by-product cost trap — a "low-cost" miner that's secretly a bet on a second commodity

1. When a miner reports an unusually low (or negative) cash cost, decompose it: is the low cost operational, or is it driven by by-product credits from a second metal at record prices?
2. If a by-product is doing the heavy lifting, the "cost story" is really a leveraged bet on that second commodity — a decline there doesn't just dent revenue, it guts the cost narrative and the headline margins.
3. Treat that double-cyclicality as concentrated risk on the way up (it amplified the win) and watch the by-product metal's price as the leading risk indicator.

• Negative/anomalously-low cash costs; the share of profitability riding on a by-product metal; that metal's price starting to retreat.

4. Read the consensus-target gap — a "Strong Buy" with no upside is a sell-side tell

1. Don't stop at the rating label; compare the average price target to the current price. Analysts are slow to downgrade a winner, so the rating lags the opportunity.
2. A Strong-Buy consensus whose average target sits only modestly above the price is the Street quietly signaling the upside is gone, even as the ratings stay bullish.

• Rating vs target-implied upside divergence; average target compressing toward the price after a big run.

5. Separate the structural (late-decade) thesis from the next-12-months tape

1. Hold two timeframes at once: a real multi-year structural deficit (you can't manufacture copper ore; 15+ year supply lag; AI-power/EV/grid demand) can be genuine and the next 12 months can be a surplus.
2. Let the structural thesis justify the core hold; let the near-term cyclical risk govern the trim and the sizing. Don't use a 2029 story to defend a position that's a peak-cycle bet today.
3. Weigh credible near-term caution (e.g. a major desk arguing the move has run ahead of fundamentals) against the long-term scarcity case rather than dismissing it.

• A structural-deficit date vs the next-12-month supply/demand balance; desk research calling a move overextended; demand softness + returning supply.

6. Pre-define the re-add triggers before you trim

1. When you trim, write down the specific conditions that would make you a buyer again — concrete price levels on the driving commodities, not vibes — so the dip is an opportunity you've already decided to take.
2. Tie the triggers to the thesis-breakers: the level at which the near-term case breaks (surplus confirmed, China softening) even though the long-term case survives is exactly the dip you want to buy.

• Your pre-set commodity levels; a "confirmed" (not rumored) surplus; the by-product metal breaking down — the conditions that make the patient-buyer plan actionable.

7. The valuation-elevation screen — buy cheap, hated sectors before the re-rating

1. Target a sector trading at "extremely undemanding" multiples on both P/E and Price/Sales — deep-value levels inconsistent with the business quality — where sentiment is washed out.
2. Anchor the contrarian case on the underlying commodity's inflation-adjusted price (not the nominal): a benchmark back near a decade-ago level in real terms, against crisis-low inventories, is the mispricing.
3. Use chart structure as the timing overlay — names that have broken a long-term downtrend (or are poised to) — and accept that some haven't broken out yet ("could be wrong"). The thesis is a gradual multi-year "valuation elevation" toward growth-stock multiples, not a single catalyst.

• P/E and P/S both at deep-value levels; the commodity's inflation-adjusted price vs history; a broken (or about-to-break) multi-year downtrend; names still above last year's throw-away lows but far below fair multiples.

8. Play a commodity recovery one step removed — the oil-service proxy

1. To express an oil/gas-recovery view, consider the oil-service names alongside the producers: a service company's stock closely tracks producer performance, giving correlated upside with a different operating profile.
2. Apply the same chart + undemanding-valuation filter; a new addition still has to clear the cheapness bar, not just the thematic one.

• Service names whose charts track the producers; valuation as cheap as the drillers; a not-yet-broken-out setup with a credible breakout case.