Actionable insights — The State of the Tech Trade as SK Hynix Opens for Trading

1. Re-rate a name the moment a feared cost becomes a sellable asset

1. Identify a stock the market has punished for one specific fear — here, "it spends huge sums on AI with no way to monetize it."
2. Watch for the catalyst that converts the feared cost into an asset: management articulating a concrete way to sell/rent the very thing that scared investors (excess capacity as a rentable product, like a cloud).
3. Buy on the reframe, not on new numbers — the multiple re-rates when the bear case is neutralized, before the revenue shows up. Confirm the base business still funds the wait (steady cash flows underneath).

• Management turning a criticized expense into a described revenue line; a durable core business still throwing off cash to fund the transition; a multi-year-low multiple on the old fear.

2. Read a counter-intuitive tape reaction as capitulation on a fear

1. When a single report sends money out of the names it should have helped and into the narrow beneficiary, don't take it at face value — ask what fear the crowd is expressing.
2. Read "sell the AI spenders / buy the memory that just printed" as capitulation on the CapEx fear — a sentiment extreme, i.e. a contrarian setup for the very names being dumped.
3. Act on the reversal (buy the washed-out spenders), while keeping the caveat that a genuinely higher-than-expected CapEx guide can still derail an individual name.

• Money leaving the logical beneficiaries on "good" news; a single-factor group being dumped on a fear rather than facts; flows confirming (record ~$19B/wk into tech).

3. The investable-vs-tradeable test — can you model three years of cash flow?

1. Before sizing a position, ask one question: can I see the next ~3 years of revenue/cash flow "with great clarity and great certainty"? If yes, it's investable; if no, it's only tradeable.
2. For a tradeable name, hand it to your charting/instinct process and keep it small — do not build a fundamental thesis on cash flows you admit are ambiguous.
3. Stress the ambiguity explicitly: competition, a possible efficiency breakthrough that cuts demand, and how much of a big move already prices the out-years' best case.

• A stock that's doubled on out-year estimates; a credible efficiency/substitution risk to demand; your own inability to defend a 3-year model — the cue to trade small, not invest.

4. Blowout-or-plateau — judge a beat against what was already expected and already priced

1. Anchor to the gap between realized and expected growth last period (the "how badly did consensus lag?" number), because that resets the bar for this period.
2. Ask the two-outcome question: does a beat blow it out and re-rate the index higher, or does an in-line print just confirm a plateau because prices already ran to it?
3. Size risk to the fact that the good news (earnings, rates, regulation) may already be in the tape — a plateau is the base case when everything is "supportive."

• The prior realized-vs-expected gap; how much the index already ran into the print; guidance over the headline beat (Samsung "got slammed" on guidance this week).

5. The walked-back-hawkishness playbook — fade an over-hawkish Fed into the political calendar

1. When a Fed signal is read as very hawkish (market pricing real odds of hikes), ask whether the messenger's incentives point the other way over your horizon.
2. Treat provocative posturing as a negotiating/expectations device likely to be "incrementally walked back" — especially where goosing the economy/market into an election is the incentive.
3. Express it in the asset that benefits from the walk-back plus a seasonal/technical tailwind, and separate a mechanical, temporary source of weakness from a broken thesis.

• A hawkish signal against an election-year growth incentive; a seasonal/technical tailwind lining up; a mechanical (balance-sheet) seller you can identify as temporary.

6. Split a metals trade — speculative store-of-value vs cash-flow commodity

1. Refuse to lump "the mining/metals space" together — test each metal for a fundamental anchor: is there a valuation or cash-flow case, or is the price purely "other people's behavior"?
2. Reject the store-of-value metal if you can't wrap dollars around it (no cash flow to model) — you may still prefer the miner over the metal, but call the metal speculation.
3. Favor the commodity with visible, structural end-demand and a supply deficit, where miners' productivity and cash generation are modelable.

• A metal with no cash-flow anchor (dock it to "speculation"); a commodity with structural demand + supply deficit and modelable miner economics (own that side).

7. Own the bottleneck everyone must pay — priced off contracts, not spot

1. When you can't pick the winner among rivals, step back to the single supplier all of them depend on — the capacity chokepoint the whole value chain has to route through.
2. Prefer the one that prices its output by contracted capacity rather than volatile spot markets — that makes revenue steadier and the valuation more predictable.
3. Use it as the "cleanest" expression of the theme when the direct plays (here, memory) are too ambiguous to model.

• A supplier every competitor must use; contracted (not spot) pricing; a valuation that stays predictable while the direct plays gyrate.

8. Re-rate a "boring" name that quietly changed its business

1. Hunt for a name still classified (and valued) as its old, dull business after it has pivoted into a hotter end-market — the classification lag is the opportunity.
2. Value it against the new comp set: if it trades at a large discount to the sector it's now really in (on EBITDA and forward PE), the gap is the thesis.
3. Prefer under-followed "orphan" names where few analysts have re-rated it yet — and size for the volatility that comes with joining a hot theme.

• A pivot the label/valuation hasn't caught up to; a steep discount to the new peer group; thin analyst coverage as the re-rating fuse.

9. Size on conviction into a mispricing — then trim the adds, not the thesis

1. When a name you already hold gaps to a price that "makes no sense," add aggressively into the dislocation — the mispricing, not a new thesis, is the trigger.
2. Once the adds push the position to an uncomfortable "monster" size, separate the two decisions: keep the thesis, but trim the recent adds back to a sane weight so position risk doesn't swamp the call.
3. Contrast with the hard cap: some managers can't add at all past a max weight, which itself signals "priced to perfection" and to buy only on weakness.

• A held name gapping to an irrational price (add) that then becomes oversized (trim the adds); a hard position cap flagging "priced to perfection."