DRAM LONG · multi (15m~1D)

The 1D/12H higher timeframes still show a bearish lower-highs sequence (81.34→74.16→66.1), but a high-volume (174M) sellside liquidity (SSL) sweep at 48.63 was followed by a sharp reversal candle, confirming a short-term 4H CHoCH (change of character). Current price 58.85 sits in a discount relative to the ~65 equilibrium (EQ) of the 81.34~48.63 range, with the 55~56.5 zone marking a 4H FVG/OTE pullback-buy area. Entry at 55.5 (band mid), stop at 51.5 (below the swept low — a break here reads as SSL-sweep failure), T1 at 62.0 (prior structural BSL), T2 at 66.1 (swing high), T3 at 70.15 gives RR=(62.0-55.5)/(55.5-51.5)=2.09, clearing the 2:1 bar. A closing acceptance above 66.1 would fully break the higher-timeframe bearish structure and open room for further upside liquidity expansion.

The sharp high-volume (136M/174M) decline into 48.63 followed by an immediate bounce is a candidate Selling Climax (SC), with the subsequent rally to 58.85 read as the Automatic Rally (AR). If this structure holds, price is now in Accumulation Phase A, with the next step being confirmation of a Secondary Test (ST) in the 53~54.5 zone on declining volume that holds above the SC low. Because this follows a large decline from the 81.34 top, however, a redistribution reading (rather than accumulation) remains live, so confidence is capped at a modest 50. Entry at 53.75 (band mid, ST retest zone), stop at 48.0 (SC-low breakdown), with creek-resistance targets at 63.0 (T1), 66.5 (T2), 70.0 (T3); RR against T1 alone is 1.61 (short of 2.0), but adopting the achievable T2 (66.5) as the qualifying target gives RR=(66.5-53.75)/(53.75-48.0)=2.22, clearing the 2:1 bar. A high-volume breakdown through the 48.0 SC low fully invalidates the accumulation thesis.

Counting 28→81.34 as a 5-wave advance, the decline from 81.34 to 48.63 becomes the (A)-wave of a larger correction, with the current bounce to 58.85 read as a (B)-wave retracement. The 0.382 retracement of the 32.71-point decline sits near 61.2 and the 0.5 level near 65 — this zone marks the (B)-wave exhaustion resistance. Short entry at 61~64 (mid 62.5), stop at 67.5 (above the 0.618 retracement, the count-invalidation level), targeting the (C)-wave decline path at 50.0 (T1), 46.0 (T2), 42.0 (T3), giving RR=(62.5-50.0)/(67.5-62.5)=2.50~2.78 — the highest RR of all six perspectives. However, an alternative count treating 48.63 as a completed correction (e.g., the end of a smaller-degree pullback within a still-live impulse) remains viable, capping confidence at 57. A high-volume closing break above 66.1 (the prior swing high) would invalidate this bearish count and require reinterpretation as the start of a new advancing impulse.

On the 4H chart, the 48.63 low (RSI 39.8, MACD histogram -0.82) prints a lower price than the prior 58.2 low (RSI 41.0, MACD histogram -1.58) while the MACD histogram is less negative — a textbook bullish regular divergence, already partially realized in the 48.63→58.85 bounce. The 12H MACD histogram is also converging higher (-3.27→-2.85), supporting improving momentum on a higher timeframe. On the 15m chart, however, the double top at 59.24/59.35 shows RSI holding (74.2→74.5) while the MACD histogram falls (0.13→-0.03) — a bearish divergence signaling a likely short-term pullback. That pullback is viewed as a buy opportunity in the 55~56 zone: entry at 55.0 (band mid), stop at 51.8 (a break here negates the bullish divergence structure), targets at 62.0 (T1), 66.1 (T2), 70.0 (T3), giving RR=(62.0-55.0)/(55.0-51.8)=2.19, above the 2:1 bar.

DRAM is a memory-semiconductor theme ETF holding SK Hynix, Micron, Samsung Electronics and peers; the AI/HBM demand surge (BofA estimates 2026 DRAM revenue +51%, ASP +33%, Micron's HBM capacity sold out through 2027) is a strong sector tailwind. Against this, the Fed's new chair regime turning hawkish (the share of committee members expecting a hike jumped from 0% in March to roughly 50% recently) has driven a 5-week Nasdaq/tech losing streak, with the imminent 7/28~29 FOMC meeting a live macro headwind. Technically, the daily EMA20 (60.76) is forming a golden cross above EMA50 (57.61) and the swing-low structure has improved (46.43→55.38→58.20), favoring trend-following longs, but this remains an unproven bounce within a large downtrend still -28% off the 81.34 June high. A market entry at the current 58.85 fails to clear RR>2, so the 55.5~57.0 pullback zone (where daily EMA50 and 4H EMA20 converge) is proposed as the primary buy zone: entry at 56.25 (band mid), stop at 52.8 (below the recent swing-low cluster), targets at 63.5 (T1, daily BB midline) and 70.15 (T2, prior swing high), giving RR=(63.5-56.25)/(56.25-52.8)=2.10. Confidence is capped at 50 given the tension between the sector tailwind and macro headwind.

Samsung's Q1 DRAM sales rose +292% YoY with operating profit up ~48x, Micron's Q3 guidance points to roughly $31B in DRAM sales (a record) at ~81% gross margin, and SK Hynix's conventional DRAM margin (77-78%) and Samsung's (72-74%) confirm the most profitable point in the cycle's history — the AI/HBM-driven memory supercycle is a clear tailwind. The global DRAM market is forecast at $633B in 2026 (+192.7% YoY), and news team's overall read is BULLISH at 68% confidence, fundamentally supporting the 54~56 pullback-buy technical thesis. However, SK Hynix's US IPO (closed 7/10, 17.8M ADRs, $29.4B raised) creates a valuation re-rating risk since the DRAM ETF's 49% Korean-asset weighting now overlaps a directly listed competitor, and the 2022-2023 DRAM bust precedent (a 70%+ downside is possible if HBM ASP declines or hyperscaler capex cools) remains a live cyclical reversal risk. Micron's Q3 earnings (7/30), the NVIDIA HBM4 supplier decision (Samsung vs. SK Hynix — a win would reprice roughly a quarter of the ETF's holdings), and the monthly TrendForce DRAM contract price are the near-term catalysts, clustering right around the 54~56 entry window; confidence is set at 68, while the RR and trade levels follow the confluence final plan (entry 55 / stop 51.5 / target 63).