Copper (HG=F) Deep Dive | 2026-10-04 — Constructive Base, 6.5 Pivot Is the Line | 2026-10-04

Call: LONG copper (HG=F), with the 6.5 S1 pivot as the invalidation level. The prior session settled at 6.55 [2026-10-02], holding the lower third of the 20-day 6.35–6.93 channel (position 35%) after a 5D -3.21% pullback that has retraced the September advance without breaking structure. Three supports underpin the call: (1) the 20-day base at 6.35 is intact and the 52-week high at 6.93 remains the upside reference; (2) visible inventories are not building — COMEX registered stocks were unchanged d/d at 471,400 short tons and SHFE warrants fell 6.7% d/d to 10,011 MT, with only a modest LME rebuild of +575 MT to 248,650 MT; (3) managed-money net length at 78,058 contracts (netPct 25.92%, crowding 68.51) is elevated but not extreme, and the copper/gold ratio at 1.57 sits in the 94th percentile of its one-year range — a pro-growth configuration. The 20-session seasonal window is positive in 9 of the last 15 years (median +1.62%). Invalidation: a settle below 6.5.

1. Bottom Line & Directional Bias

Call: LONG copper (HG=F), reference contract HGZ26.CMX. Invalidation: a daily settle below the 6.5 S1 pivot.

Three reasons. First, price structure: the prior session settled at 6.55 [2026-10-02], still inside the 20-day 6.35–6.93 channel at position 35%. The 5D move of -3.21% is a retracement within that channel, not a break — 6.35 has not been tested. Second, inventories are not confirming the selloff: COMEX registered stocks were unchanged d/d at 471,400 short tons [2026-10-01], SHFE warrants fell 6.7% d/d to 10,011 MT [2026-09-30], and the LME build was trivial at +575 MT to 248,650 MT [2026-10-02]. Third, positioning and cross-asset: managed-money net length of 78,058 contracts (netPct 25.92%, crowding 68.51) is elevated but below the 70th percentile, and the copper/gold ratio at 1.57 sits in the 94th percentile of its one-year range — a pro-growth configuration.

The invalidation is a settle below 6.5. That would put the market back into the lower half of the 20-day range with the 6.35 base exposed, and the constructive read would be wrong. Until then, the bias is to buy the base of the range, not to chase.

2. Price Action & Technical Analysis

Settle 6.55 [2026-10-02], +0.17% on the day. The 5D change is -3.21% and the 20D is -1.73% — the pullback is recent and shallow relative to the 20-day range. ATR14 is 0.13, i.e. 1.99% of price as a full daily range; RV20 is 26.5% annualized. The 20-day channel is 6.35–6.93, with price at position 35%, and the 52-week range is 4.83–6.93 — the market is in the upper third of its annual range but in the lower third of its monthly range. That combination is the definition of a consolidation after a trend, not a reversal.

Pivots from the settle-based snapshot: P 6.57, R1 6.61, S1 6.5, R2 6.68, S2 6.46. Price settled just below the pivot, which is neutral-to-soft on a one-day view; the first meaningful support is S1 6.5, then S2 6.46, with the 20-day low at 6.35 behind them. To the upside, R1 6.61 is the first hurdle, R2 6.68 the second, and the 20-day high at 6.93 the range top. The ordering is consistent: 6.46 < 6.5 < 6.57 < 6.61 < 6.68 < 6.93.

In early Asian trade on the report date, the market is trading around the prior settle; the report-date bar is unfinished and no close should be inferred from it. The last completed weekly bar (2026-09-21 to 2026-09-25) opened 6.72, high 6.93, low 6.7, closed 6.77, +1.11% w/w — a constructive weekly bar that closed near its high. The current week (from 2026-09-28, five sessions) is not closed and shows -3.21%; no weekly-close conclusion can be drawn from it. The weekly read is therefore: last completed week bullish, current week retracing, and the retracement has not yet violated the prior week's low at 6.7 on a weekly-close basis — though on a daily basis price at 6.55 is already below it, which is why the 6.5 pivot, not 6.7, is the operative line.

View: constructive while above 6.5; a settle below 6.5 shifts the tactical bias to neutral and opens 6.35.

3. Supply-Demand Balance & Fundamental Drivers

Visible inventory is the cleanest fundamental signal available and it is not bearish. COMEX registered stocks stood at 471,400 short tons [2026-10-01], unchanged d/d. LME warehouse stock was 248,650 MT [2026-10-02], up 575 MT d/d (+0.2%), with a 20-report change of +14,475 MT — a gradual rebuild, not a flood. SHFE warrants fell 721 MT d/d to 10,011 MT (-6.7%) [2026-09-30], which is a small absolute number and a fast draw; Chinese exchange inventory remains tight.

The composition matters. The LME rebuild of roughly 14.5 thousand MT over 20 reports is the only genuine accumulation signal, and at 248,650 MT the LME book is not heavy in absolute terms. COMEX flat at 471,400 short tons removes the most obvious bearish confirmation the market would need to press the 5D -3.21% move further. SHFE at 10,011 MT is the tightest of the three and the most sensitive to Chinese downstream restocking.

The curve is in contango: M1-M2 at -0.0315 (-0.48%), with roll yield of -5.77% and slope 0.0337. Contango of this magnitude is a carry cost for long positions, not a price cap; it tells us prompt physical tightness is not acute, which is consistent with the modest LME rebuild. It does not invalidate a constructive medium-term view, but it does mean longs pay to hold, which argues for defined-horizon positions rather than indefinite carry.

Macro transmission: the dollar index at 101.93 (-0.17% [2026-10-02]) is a mild tailwind, and the 10-year yield at 5.28% (+0.76%) is a headwind for the whole complex. The copper/gold ratio at 1.57, in the 94th percentile of its one-year range, says the market is pricing copper as a growth asset relative to gold — that is the cross-asset confirmation of the constructive fundamental read.

View: inventories are neutral-to-supportive, the curve is a carry cost rather than a signal, and the fundamental balance does not justify the 5D selloff. Constructive above 6.5.

4. Positioning & Fund Flows

Managed-money positioning, as of 2026-09-29: open interest 301,201, longs 92,094, shorts 14,036, net 78,058, a weekly change of -4,464. The prior week (2026-09-22) showed net 82,522, +17,416 w/w; 2026-09-15 showed net 65,106, -17,048; 2026-09-08 showed net 82,154, +9,272. The pattern is a sharp build into 2026-09-22, followed by a modest reduction into 2026-09-29. The latest week's -4,464 is small relative to the +17,416 build, so this is trimming, not liquidation.

Crowding: netPct 25.92% as of 2026-09-29, crowding percentile 68.51 on a three-year window, down from 69.05 the prior week and 69.39 on 2026-09-08. The CTA trend proxy is 62, unchanged across all four weeks, and the hedge ratio is 54.82%. A crowding percentile of 68.51 is elevated but below the 70th percentile and well below the extreme zone; it does not meet the bar for calling the trade crowded. The CTA proxy at a flat 62 across four weeks indicates trend-following flow is stable rather than adding or unwinding.

Implied versus realized: RV20 is 26.5%. The available implied-vol proxies are cross-asset — ^OVX at 51 (49th percentile), ^GVZ at 23.23 (15th percentile), ^VXSLV at 36.9, ^VIX at 15.31 (15th percentile). The broad message is that equity and gold optionality are cheap relative to their own history while energy optionality is mid-range. For copper specifically, RV20 at 26.5% with ATR14 at 1.99% of price describes a market with real daily range but no panic.

View: positioning is supportive but no longer a tailwind of the same magnitude as in mid-September; the reduction is orderly. Net length can rebuild if 6.5 holds.

5. Cross-Asset Relative Value

The key ratio is copper/gold at 1.57 (ratio ×1000), in the 94th percentile of its one-year range and the 52nd percentile of its three-year range. Read carefully: the one-year percentile says copper has been strong versus gold over the past twelve months; the three-year percentile near the middle says this is not an extreme on a longer horizon. A rising copper/gold ratio is a pro-growth signal, and at the 94th one-year percentile the market is already expressing a growth-friendly view. That is a confirmation of the constructive copper call, but it also means the ratio is closer to its one-year ceiling than its floor — further copper outperformance would need either a fresh growth impulse or gold weakness.

The dollar at 101.93 (-0.17%) is a mild positive for dollar-denominated copper. The 10-year at 5.28% (+0.76%) is the offsetting drag: high real rates raise the carrying cost of inventory and weigh on cyclical assets. The contango (roll yield -5.77%) is the third leg — longs pay carry, so the relative-value case for copper rests on price appreciation, not on positive roll.

View: cross-asset configuration is pro-growth and mildly supportive, but the copper/gold ratio is near its one-year top, so incremental upside from relative value is limited. The dollar is a small tailwind; rates are the main cross-asset headwind.

6. Historical & Seasonal Patterns

The seasonality block covers the same calendar start over the next 20 sessions across the last 15 years: mean +1.47%, median +1.62%, up in 9 of 15 years. The best instance was 2011 at +13.06%; the worst was 2012 at -8.53%. The distribution is positively skewed by the 2011 outlier, and the hit rate of 9/15 (60%) is modest. This is context, not a signal — the sample is small and the dispersion is wide (a 21.6-percentage-point spread between best and worst).

The practical read: the seasonal window leans mildly positive, with a median gain of +1.62% over 20 sessions. Applied to a 6.55 settle, the median path would put copper near 6.66 by early November — roughly the R2 6.68 pivot, still below the 20-day high at 6.93. That is consistent with the base case in section 7 and does not by itself justify a large position. Seasonality supports the constructive bias but is not the reason for it; the inventory and positioning evidence carries the call.

View: mildly positive seasonal tailwind over the next 20 sessions, median +1.62%, but the wide dispersion means it should size positions, not set them.

7. Scenario Analysis (Base / Bull / Bear)

Base case — 50%: range grind higher. Trigger: 6.5 holds on a settle basis and the market reclaims the 6.57 pivot. Path: 6.55 → 6.61 (R1) → 6.68 (R2), with the 20-day high at 6.93 as the outer bound. Action: hold long exposure established near 6.5–6.55, add only on a settle above 6.61, take partial profit into 6.68. This agrees with the section 1 call.

Bull case — 30%: breakout retest. Trigger: a settle above 6.68 (R2) on rising open interest, with LME stock flat-to-lower and SHFE warrants continuing to draw. Path: 6.68 → 6.93 (20-day and 52-week high). A settle above 6.93 would open the range extension. Action: add to longs on the 6.68 settle, trail stops to 6.55, target 6.93. The copper/gold ratio at the 94th one-year percentile is the confirming cross-asset signal; a further rise in that ratio would validate the path.

Bear case — 20%: range failure. Trigger: a daily settle below 6.5 (S1), confirmed by a build in LME stock beyond the recent +14,475 MT 20-report pace or a further reduction in managed-money net length beyond the latest -4,464. Path: 6.5 → 6.46 (S2) → 6.35 (20-day low). Action: exit longs on the 6.5 settle, stand aside; do not initiate shorts until 6.35 breaks, because the 52-week range low at 4.83 is far away and the intermediate structure is still a range. A settle below 6.35 would shift the tactical bias to bearish and target the lower half of the annual range.

Probabilities sum to 100%. The base case is the section 1 call; the bull and bear cases are the weighted alternatives, not competing conclusions.

8. Trading Strategies & Risk Management

Strategy 1 — Long the base of the range (primary). Entry 6.52–6.56, i.e. at or near the 6.55 settle and above the 6.5 S1 pivot. Stop 6.37, below the 20-day low at 6.35 and beyond one ATR14 (0.13) from entry. Target 6.68 (R2) for the first tranche, 6.9 for the second, just below the 20-day/52-week high at 6.93. Horizon 5–15 sessions. Conviction 6/10. Size: half of normal risk budget, because the market is mid-range and the 5D trend is negative; add the second half only on a settle above 6.61 (R1).

Strategy 2 — Breakout continuation (conditional). Entry on a daily settle above 6.68 (R2). Stop 6.52, below the 6.5 pivot and roughly one ATR14 from entry. Target 6.93, the 20-day and 52-week high. Horizon 5–10 sessions. Conviction 7/10 if the breakout comes with rising open interest and flat-to-lower LME stock; 4/10 otherwise. Size: full risk budget only on the confirming inventory print.

Risk management: the invalidation for the entire constructive view is a settle below 6.5; both strategies are void below that level. The contango (roll yield -5.77%) means carry is a cost, so keep horizons defined and avoid indefinite holds. The 10-year at 5.28% is the macro risk to both positions; a sharp further rise in yields would pressure the whole complex.

9. This Week's Data Calendar

BJT 10-05 22:00 | ET 10-05 10:00 — ISM Services PMI (SEP), forecast 54, previous 55.4; surprise if outside 54 ± 1.4. Affects DXY and the metals complex, including copper via the growth channel.
BJT 10-07 04:30 | ET 10-06 16:30 — API Crude Oil Stock Change (OCT/02); energy-linked, indirect read-through to industrial sentiment.
BJT 10-07 22:30 | ET 10-07 10:30 — EIA Crude Oil and Gasoline Stocks (OCT/02); same channel.
BJT 10-08 02:00 | ET 10-07 14:00 — FOMC Minutes; the highest-impact event for DXY and metals this week.

This report is generated automatically from public quantitative and macro data for research and market tracking only. It does not constitute investment advice or a recommendation to trade.

銘柄別深度レポート深度分析

COMEX金GC=F 2026-10-04

Gold (GC=F) Deep Dive | 2026-10-04 — Bearish, 4143.1 20-day floor in focus | 2026-10-04

Copper (HG=F) Deep Dive | 2026-10-04 — Constructive Base, 6.5 Pivot Is the Line | 2026-10-04

本日のリスクマップ

銘柄別深度レポート 深度分析

過去のレポート

⟠ 今月の戦略バックテスト

銘柄別深度レポート深度分析