Actionable insights — Portfolio Update: Mining & Sand

1. Sell a parabola in tranches — dollar-cost-average out through the "hockey stick phase"

1. Identify when a position or group has entered the hockey stick phase — the near-vertical segment of the chart where the rate of ascent itself is accelerating. That, not a price target, is the trigger.
2. Do not try to pick the top and do not sell all at once. Start a scheduled series of sell-downs — a dollar-cost-average in reverse — while the move is still going up.
3. Accept that this looks wrong in real time: "it often looks overly cautious as the lift-off hits its maximum acceleration." Judge the tactic on the average disposition price after the correction, not on whether any single sale was the high.
4. After the exit, impose a quiet period: do not re-endorse the group until price and fundamentals give a fresh, specific entry number — the discipline that keeps the round trip from becoming a round trip back to flat.
5. Re-enter on an explicit valuation trigger, and say so plainly ("at $29 this is an attractive buy point"), rather than on a feeling that the selling is "done."

• A chart whose slope is steepening week over week; the urge to "let it run" at maximum acceleration; and, on the way back down, a pre-set re-entry price rather than an open-ended wait.

2. Trade index inclusion and removal against the fundamentals — the flow is mechanical, the business is not

1. When a name is added to an index, recognize the resulting move as forced buying, not a verdict on the business. Ask separately: are fundamentals improving or deteriorating into that lift?
2. Inclusion + deteriorating fundamentals = sell. The flow has manufactured an artificial price you would not otherwise get — "a great time to ring the cash register."
3. When a name is removed, run the mirror test: index funds must sell regardless of outlook. Removal + improving fundamentals = buy. The decline has a known, non-fundamental cause and a known end date (once the rebalance clears).
4. Size the opportunity by how far the mechanical move has overshot: measure the drop from the pre-announcement price and check that the operating outlook has genuinely turned (industry cycle, volumes, pricing) before committing.
5. Enter partially when the fundamental upswing is real but "tentative" — full size is for when the cycle confirms.

• Index add/delete announcements and rebalance dates; a price move that has no operating explanation; and the direction of the underlying cycle at the moment the flow hits — that pairing is the whole trade.

3. Price a natural disaster as a timing discount, not a thesis break

1. When an operator is hit by a physical, non-recurring event (earthquake, flood, fire, mine collapse), quantify the actual damage in units — how much production was lost this quarter, how much guidance came down for the year.
2. Test whether the event touches the asset base or only a period's output. If the reserves, grades and mine plan survive intact, the hit is a cash-flow timing issue, not an impairment of value.
3. Check the forward estimate line: if next year's consensus steps back up (and the multiple on it is low), the market is pricing the accident as permanent when the industry treats it as routine — "common in the mining industry, but they tend to fade into insignificance over time."
4. Buy the discount only with the long-term asset metrics confirming quality (growing reserves, improving grades) — a one-off on a low-quality asset is not the same trade.

• The quarter-vs-year split in the guidance cut; an unchanged reserve/mine-plan disclosure; a forward EPS line that recovers; and a multiple computed on the recovered year, not the damaged one.

4. Use retail flows in a sector as a contrarian gauge — apathy after a great year is the buy

1. Pull the retail fund-flow data for the sector (broker/bank flow reports) rather than relying on anecdote or price action.
2. Compare the flow to the sector's recent realized returns. The signal is the mismatch: heavy retail outflows or negativity after a spectacular year is the bullish extreme; heavy inflows after a spectacular year is the warning.
3. Justify the contrarian reading with the base rate — retail's "recurring tendency to be on the wrong side of the bull and bear moves" in that specific group; this works where retail is a large, momentum-driven marginal buyer (miners, commodities, thematics).
4. Combine with a valuation trigger before acting; sentiment alone sets the odds, price sets the entry.

• Sector flow data diverging from trailing returns; retail selling a group that just made record profits; the same tell inverted (inflows into a spent rally) as the exit signal.

5. Separate tax-loss selling from business deterioration — and buy the November washout

1. In a name already down sharply on genuinely poor quarters, ask in November whether the final leg of weakness is being aggravated by tax-loss selling — investors dumping losers to book losses before year-end.
2. If so, treat the December low as artificial and expect a January rebound once the pressure abates; the trade is to accumulate into the forced selling, not to wait for good news.
3. Verify afterwards: if the stock snaps back with no fundamental change, the diagnosis was right — then step the rating down as the price runs (buy → hold → trim) rather than riding it round trip.

• A big year-to-date loser making new lows in November on no news; the turn of the calendar; and a pre-planned ladder of downgrades on the way back up.

6. Own the physical bottleneck, and price the moat that competitors cannot copy

1. Trace the commodity's production chain to the input that is non-optional — the thing without which output simply cannot happen ("you simply can't produce shale crude without frac sand").
2. Confirm the supplier's position in the single most important geography for that output, and check the market structure: a consolidation to near-duopoly implies pricing power in the next up-cycle even while spot pricing is currently soft.
3. Identify a physical infrastructure asset no competitor has and translate it into unit economics (here, a conveyor that removes trucking cost from the delivered price) — that is the durable spread, not the commodity price itself.
4. Separate the two signals: volumes recovering tells you the cycle has turned; pricing still soft tells you it is early. Buy the early stage partially.
5. Add optionality the market is ignoring (an adjacent, growing business line) as free upside rather than the reason to own it.

• Volumes turning before price; industry consolidation to two or three players; a capital asset that structurally lowers delivered cost; and a valuation that credits none of it.

7. Ask whether a peer's bad quarter is already in the price before treating it as a reason to wait

1. When a close competitor misses, note that the market extrapolates the miss across the group ahead of the name's own report.
2. Measure what has already been discounted: the size and speed of the decline since the peer's print (or since the position's own downgrade) is the market's implied bad-quarter charge.
3. Take the position anyway — partially — if the collapse plausibly covers the miss; keep the remaining capital for the report itself. The stance is "that's certainly possible, but it appears well-discounted," not a denial that the miss could repeat.

• A peer print that resets the whole group; the magnitude of the sympathy sell-off; and the target's own earnings date as the moment to complete (or abandon) the position.

8. Underwrite the commodity and the operator as two separate decisions

1. For any producer, state explicitly that the underlying commodity price is "often THE most critical" variable — then form a view on it on its own evidence, independent of the stock.
2. Decide whether the commodity's decline is a correction within a bull market or a regime change, using structural buyers and monetary drivers rather than price momentum (here: central-bank buying, rising sovereign-bond-crisis odds, ongoing fiat debasement).
3. Only then layer the equity-specific work — valuation, reserves, grades, one-off events — on top. A cheap operator under a broken commodity is a value trap; a cheap operator under a still-intact bull market is the setup.

• Structural, price-insensitive buyers of the commodity; the debasement/fiscal backdrop; and whether the equity's discount is being driven by the commodity or by something company-specific you can diagnose separately.