SK Hynix doubles down on $64B factory spending just as demand fears trigger 6% Korea rout

Verdict: The glut thesis is real, but the trigger is not — wait for $180 to break

This is a well-timed thesis about a real mismatch — SK Hynix's $64B plant commitment (per Reuters, July 2, 2026) colliding with Meta's capacity-offloading plans and a 6% Korea rout (per Bloomberg, same date) — but the trade is not live and should not be chased. The strongest point for it: EWY's two-day change is already -3.4% against a -5% trigger and price sits just 1.8% above first support at $180, so a single ugly session brings the entry within reach. The strongest point against: across 1,235 daily bars over 60 months, the five-condition entry never fired once, and EWY's trend-strength reading of 9.4 versus a required 20 suggests the rules may not even catch the exact July 2026 capitulation they were built around. LPL's fundamentals also cut against blanket pessimism — gross margin improved 8.1 points to 9.7% in FY2024 and free cash flow turned positive — though 39.7% share dilution and a -39.2% return on equity keep the strain real. The verdict flips to actionable if EWY's two-day drop reaches -5% while trend strength pushes above 20 and price closes below $180. Until then, treat this as a mechanical watch-list: pre-set levels, no position.

Trade now: the short setup is close on momentum, far on trend strength — here is exactly what flips it live

Nothing is live yet — this is a watch-list setup, and 'wait' here means a concrete set of levels, not a passive stance. The strategy shorts a semiconductor proxy only when five conditions line up on the same day: a two-day price drop worse than -5%, the MACD line crossing below its signal line, a close below first support, volatility (ATR) above 0.5, and a trend-strength reading (ADX) above 20. On the Korea ETF (EWY, last close $183.39), the momentum condition is closest: the two-day change is -3.4%, so another down day or two puts it in range, and price sits only 3.4 points above first support at $180. But the trend-strength condition is far away — ADX reads 9.4 against the 20 threshold — and the volatility check cannot yet be judged because no ATR reading is available. On LPL ($3.26), momentum is actually positive at +0.3% and support sits at $3.20, so the short trigger is much further off. Risk framing is set by the strategy's own rules: a hard stop at -2.5% on the position and a take-profit at +5.1%, an effective reward-to-risk of roughly 2:1 regardless of which symbol triggers first. The thesis itself (per the idea) is that SK Hynix's $64B factory announcement colliding with Meta's plans to sell computing power creates a capacity-glut mismatch — but the idea argues you get paid only by entering after the drop has begun, not before. Chasing the short here, with EWY momentum still above the -5% threshold and ADX that weak, would mean paying full risk for a move that has not yet confirmed. If EWY's two-day change reaches -5% while ADX pushes above 20, the setup flips to actionable; until then the correct action is no position, with the stop and target levels pre-set so execution is mechanical when (and if) the entry fires. One process note: the research author requested a bounded, thesis-preserving parameter search because the compiled thresholds never fired simultaneously across the evaluation history, but no robust alternative setup was established, so the live thresholds stand as published.

A rare setup built for exactly the capitulation Korea just experienced

The macro setup the idea describes is real and well documented in the cited coverage. Per the Reuters piece dated July 2, 2026, SK Hynix committed roughly $64 billion to new chip plants under an AI-driven investment plan; on the same day, per Bloomberg, South Korean stocks tumbled 6% as AI jitters hit chipmakers, and CNBC reported Samsung and SK Hynix shares each fell more than 7% as the rout spread from Wall Street. A supplier layering on massive fixed capacity at the same moment a major customer (Meta, per the idea's account) plans to sell its own computing power is a textbook timing mismatch — the exact state where incremental supply meets softening incremental demand. The timing of the short matters as much as the direction. This idea is explicitly a capitulation-entry structure: it waits for a sharp multi-day momentum drop, a momentum-line cross below its signal, and a break of first support, with trend-strength and volatility filters layered on. That means it does not fight the initial headline bounce; it seeks to enter after the crowd has already broken support — historically where subsequent downside legs in semiconductor selloffs accelerate. The 7% RSI-bounce exit and the RSI-below-25 oversold exit cap the trade's time in the water, which suits a news-driven, high-velocity regime like the July 2026 Korea rout. The capital-allocation argument has teeth when you look at how thin display and memory margins run. In the provided sector data, even the best recent year in this Korean tech cohort produced a gross margin under 20%, and FY2023 collapsed to roughly 1.6% —…

Scores

• Conviction score breakdown: 44
• Thesis support: 65
• Trade readiness: 30
• Risk quality: 45
• Trigger proximity: 40
• Fundamentals trend: 40

Watch items

• EWY — ROC (2) on EWY
• EWY — First support (EWY close)
• EWY — ADX (14) on EWY
• EWY — RSI (14) on EWY
• LPL — ROC (2) on LPL
• LPL — First support (LPL close)
• LPL — RSI (14) on LPL