SKHY LONG · 4h
| Entry | 158~161 |
|---|---|
| Stop | 148 |
| Target | 186 / 194.92 |
| R:R | 2.3 |
| Confidence | 46% |


Rationale
- SK Hynix's Q2 2026 earnings are imminent (7/22~23) with consensus at +260% YoY revenue; confirmed HBM4 supply to Nvidia's Vera Rubin platform and 56.4% HBM market share are structural upside catalysts.
- Multi-timeframe (1D/12H/4H) structure retains higher-lows (145.67→150.07→152.5) after the 135→194.92 impulse with EMA20>EMA50 aligned, so the current pullback reads as a retracement within an uptrend rather than a reversal.
- All four technical methodologies (ICT, Wyckoff, Elliott, Divergence) point SHORT off the 175.54 liquidity sweep and rejection (BSL sweep reversal / Upthrust / Elliott wave-C / regular bearish divergence), but their T1-based RR (1.55~1.97) falls short of the 2:1 threshold and is excluded from the confluence tally — this still signals possible further downside into the 152~158 zone, motivating a lower entry.
- A hawkish tilt into the 7/28~29 FOMC and geopolitical risk-off dollar strength (DXY ~100.9) are near-term headwinds for high-multiple semiconductor growth names, and the risk of a repeat BoK-driven chain shock (KOSPI -6.37% on 7/16) further caps confidence.
- News flow is strongly bullish (5 bullish vs 2 bearish items), with Micron/Samsung HBM supply delayed until 2027 supporting SK Hynix's pricing power and margins.
- Only the perspectives that pass the RR>2 filter (Macro LONG at RR 2.3, News LONG at 78 confidence) were tallied for the direction vote, yielding LONG; since a market entry at 162.03 fails the RR>2 test, a 158~161 pullback entry zone was set to secure RR 2.30.
Analysis by methodology
The 1h rally swept above prior swing highs (164.09/172.79) and impulsively rejected right at 175.54, a classic liquidity grab (BSL sweep) followed by reversal. The bearish FVG left behind in the 168.7→162 drop (166~169) plus the 1h EMA20 (167.1) form the short retracement entry zone. Stop sits above the sweep high (175.7), targets at SSL (150.07) and the spring low (145.67). However T1-based RR is only 1.89, below the 2:1 threshold, so this view is excluded from the final confluence tally; holding 159 support without losing 162 would reopen a long re-evaluation.
The 152.62→194.92 surge, reversing on record volume (23.4M), reads as a Buying Climax; the subsequent AR fell to 145.67, forming a 150~175 trading range. 145.67 could be a spring (undercut and recovery, implying re-accumulation), but the more recent retest of 175.54 was rejected on shrinking volume (7.5M), closer to an Upthrust (UT). The base case after a UT is a retest of range lows (152~150), but T1-based RR is only 1.83, below 2:1, so it is excluded from the tally. If 150 support holds firmly on rising volume, this flips to re-accumulation (awaiting SOS), invalidating the short.
Counting 135→194.92 as a 5-wave impulse fails the overlap rule (wave-4 low 145.67 enters wave-1 territory), invalidating the impulse count. Instead this reads as an ABC correction off the 194.92 high: wave A (194.92→145.67), wave B (145.67→175.54, landing precisely on the 0.618 retracement of A at 175.1), with wave C now in progress. The C=A projection targets 126.3, the 0.618×A projection ~145, prioritizing T1 150 / T2 145.67 — but T1 RR is 1.97, just short of 2:1, so it is excluded from the tally. An alternate count (new impulse from 145.67, currently in wave 2) exists, but the precise 0.618 rejection at 175.54 favors the ABC read.
At the 15m swing high, price made a higher-high (174.56→175.54) while RSI made a lower-high (75.7→63.0) — a clean regular bearish divergence that already preceded the current drop. On 1h, 175.54 confirmed a higher-high with RSI at 73.9, so no higher-timeframe divergence exists; the signal is confined to the 15m. The current leg is making lower-lows in both price and RSI (167.04→161.28, 28.2→23.9), so no bullish divergence has formed yet at the low; with 15m RSI at 23.9 (oversold), a bounce into 166~169 followed by renewed selling is favored. Because the signal has already played out, T1 RR is only 1.55 — too weak, and excluded from the tally.
The macro backdrop is mixed — a hawkish tilt into the 7/28~29 FOMC and geopolitical risk-off dollar strength (DXY ~100.9) are near-term headwinds, but SK Hynix's own HBM4 supercycle (confirmed Nvidia Vera Rubin supply, 56.4% HBM share) and the imminent Q2 2026 earnings (7/22~23, consensus +260% YoY revenue) offer a stronger offsetting tailwind. From a trend-following view, 1D/12H/4H all retain a higher-low structure (145.67→150.07→152.5) after the 135→194.92 impulse, with EMA20>EMA50 intact, so the current pullback reads as a retest of the 158~162 EMA confluence rather than a trend reversal. A market entry (162) fails the RR>2 test, so a pullback entry zone of 158~161 (primary, RR 2.3) was set, invalidated on a close below 148 (HL structure break). Earnings gap risk and the possibility of a repeat BoK-driven semiconductor chain shock (as on 7/16) cap confidence at 52.
SK Hynix is a core beneficiary of the AI memory supercycle: record Q1 2026 results (net profit ~$29B, 77% net margin), a successful Nasdaq listing ($26.5B raised, ~7x oversubscribed), 56.4% HBM market share, and the start of HBM4 mass production. HBM demand is sold out through end-2026, with Micron and Samsung supply delayed until 2027, supporting pricing power and margins. China's antidumping probe and the US export-license shift (VEU wind-down) are risks, but low China dependence limits near-term impact. Bullish items (5) far outnumber bearish (2), but post-listing volatility and gap risk into the 8/10 Q2 earnings print mean the direction is strongly LONG while near-term timing risk must be respected.
Invalidation
A 4H close below 148 breaks the 4H/12H higher-low structure (150.07/152.5) and immediately invalidates this long thesis. A close below 145.67 (the crash low / candidate spring) would validate the Elliott wave-C extension (retest of 135) or a failed-Wyckoff-UT renewed-decline scenario, flipping to the opposite (SHORT) case. Conversely, a strong volume-confirmed close above 175.54 (the 1h BSL sweep high) would confirm an early breakout with no further pullback, warranting a re-raised entry zone and higher confidence. Because the Q2 earnings print (8/10) can produce a gap, price may move sharply through the stop around the release even after a limit fill, so extra caution around that volatility window is warranted.
Context
Across timeframes, the 1D chart sits right at the range midpoint/equilibrium (164.96) of 135.0~194.92, landing exactly on EMA20/BB-mid (161.77), while 12H and 4H both retain EMA20>EMA50 alignment atop a gentle higher-low (double-bottom-like) sequence of 145.67→150.07→152.5. The 1H/15M show a sharp pullback right after the 175.54 sweep (breaking below 1H EMA20 167.1/EMA50 163.4, RSI at 23.9 oversold), so downside pressure dominates near-term, but this reads as a correction within the higher 4H/12H/1D trend. The key pivot cluster is 158~162, where 4H EMA20 (161.49)/EMA50 (160.92), 12H EMA20 (160.78), and 1H EMA200 (160.81) all converge, near the 1D equilibrium (161.77). On the macro/event side, SK Hynix's Q2 2026 earnings (7/22~23) are imminent with record-beating consensus, so buying into this pullback plausibly coincides with an earnings catalyst, while the Fed FOMC (7/28~29) and geopolitical risk overlap to raise volatility around the same window. This entry zone was chosen because a market entry (162.03) fails the RR>2 test, and because the near-term downside targets flagged in common by all four technical methodologies (152~158) overlap with the macro-derived optimal pullback zone — i.e., where the short-term technical correction is expected to complete is the same level identified as the structural long re-entry point.