📅 2026-10-06 · COMEX銅 · 銘柄別深度レポート

本レポート提供元 MacroTerms.com

データ基準日: 2026-10-05 · 決済基準: 取引所最終日次決済価格

Copper (HG=F) Deep Dive | 2026-10-06 — Constructive Base, 6.93 the Line | 2026-10-06

Call: Bullish copper on a 1–3 week horizon, with the view invalidated on a settle below 6.52 (S2). Three reasons underpin the call. First, positioning has de-risked without breaking: managed-money net fell to 78,058 lots (net 25.9% of OI) in the week to 2026-09-29, a 4,464-lot reduction from 82,522, yet the crowding percentile at 68.5 is mid-range rather than stretched, leaving room for re-addition. Second, the physical tape is not loose — LME warehouse stock at 248,650 MT rose only 575 MT d/d and 14,475 MT over 20 reports, while SHFE warrants fell 721 MT to 10,011 MT and COMEX registered held at 471,400 short tons. Third, the copper/gold ratio at 1.57 sits in the 94th percentile of its one-year range, a pro-growth signal that argues against a demand-led breakdown. The last completed weekly bar (2026-09-28–10-02) closed at 6.55, down 3.21% w/w, so the market is stabilising rather than trending; the 6.64 settle sits mid-channel (50.8% of the 6.35–6.93 20-day range) with ATR14 at 0.129.

1. Bottom Line & Directional Bias

Call: Bullish copper (HG=F), 1–3 week horizon. Invalidation: a daily settle below 6.52 (S2).

Three reasons. First, the positioning reset is constructive rather than damaging: managed-money net length fell 4,464 lots to 78,058 in the week to 2026-09-29, but the crowding percentile of 68.5 on a three-year window is mid-range, not extreme — there is capacity for length to be rebuilt without the trade being crowded. Second, visible inventories are not building aggressively: LME warehouse stock at 248,650 MT (2026-10-02) rose just 575 MT d/d and 14,475 MT over 20 reports, SHFE warrants dropped 721 MT to 10,011 MT, and COMEX registered held flat at 471,400 short tons. Third, the copper/gold ratio at 1.57 sits in the 94th percentile of its one-year range, a pro-growth configuration inconsistent with a demand-led collapse.

The counterweight is momentum: 5D +0.11% and 20D -0.62% on the 6.64 settle, with the last completed weekly bar down 3.21% w/w to 6.55. This is a base-building call, not a breakout call. A settle below 6.52 breaks the 20-day channel floor logic and voids the view.

2. Price Action & Technical Analysis

HG=F settled at 6.64 on 2026-10-05, up +1.4% (settle) on the day. In early Asian trade on 2026-10-06 (07:00), the last print was 6.64 (+0% vs settle), with an Asia range of 6.63–6.65 — a tight, directionless session that neither confirms nor contradicts the prior session's advance.

The medium-term picture is flat-to-soft. The 5-day change is +0.11% and the 20-day change is -0.62%, both computed from settled bars. The 20-day channel runs 6.35–6.93, and at 6.64 the market sits at the 50.8% position of that range — dead centre. The 52-week range is 4.83–6.93, so the 20-day high is also the 52-week high; the market is consolidating just below a major ceiling rather than pressing it.

Volatility is moderate. ATR14 is 0.129, or 1.94% of price — that is the full expected daily range, not a one-sided band. RV20 is 27% annualised. With ATR at roughly 0.13, a stop placed inside 0.1 of entry would sit inside normal daily noise; risk points need to be sized off real levels.

Pivots from the settle-based snapshot: P 6.62, R1 6.68, S1 6.58, R2 6.71, S2 6.52. The settle at 6.64 is above the pivot, which is a mildly constructive intraday posture, but R1 at 6.68 is only four cents away and the 20-day high at 6.93 is the real objective. To the downside, S1 6.58 is the first shelf and S2 6.52 is the invalidation line.

The weekly block matters for context only. The last completed weekly bar (2026-09-28–2026-10-02) opened 6.77, high 6.77, low 6.51, closed 6.55, down 3.21% w/w — a bearish completed bar that failed at the prior week's open. The current week (from 2026-10-05, one session in) is not closed; the +1.4% on that unfinished bar is a rebound attempt, not a weekly reversal. No weekly-close conclusion can be drawn from it.

View: constructive while 6.58 holds on a settle basis; the 6.68–6.71 pivot band is the first test, 6.93 the range objective.

3. Supply-Demand Balance & Fundamental Drivers

Visible copper inventories are the cleanest read available and they are mixed-to-tight, not bearish.

COMEX registered stocks: 471,400 short tons (2026-10-01), d/d +0.0%. Flat on the day. This is a high absolute level and the principal bear argument on the physical side — COMEX has absorbed metal and the registered pool is not drawing down.

LME warehouse stock: 248,650 MT (2026-10-02), d/d +575 MT (+0.2%), 20-report change +14,475 MT. The direction is mildly bearish — LME has been building over the past month — but the pace is modest: roughly 700 MT per report on average. A 14,475 MT build against a 248,650 MT base is a 6.2% increase over the window, not a flood.

SHFE warrant: 10,011 MT (2026-09-30), d/d -721 MT (-6.7%). This is the tightest of the three pools in absolute terms and it is drawing. A 10,011 MT warrant book is small; the 6.7% single-day decline signals that Chinese consumers are taking metal rather than leaving it in the warehouse.

The net read: LME building modestly, COMEX flat and heavy, SHFE tight and drawing. That combination is consistent with metal having migrated to the West — a location arbitrage, not a demand collapse. It caps upside velocity but does not break the trend.

On the news flow, the supply-side headlines are the relevant transmission channel. A potential Cobre Panamá restart (SMM, 2026-10-05) is the single largest medium-term bearish supply swing factor in the pipeline; any concrete restart timetable would pressure the back end of the curve. Offsetting that, Freeport reported Q3 copper production of 830 million lb as Grasberg recovery advances (SMM, 2026-10-05) — a recovery, not a surprise expansion — and China is seeking copper concentrate supply commitments as a condition of Anglo–Teck merger approval (SMM, 2026-10-05), which speaks to concentrate tightness at the Chinese smelter level. Greenland's approval of the Tanbreez mine plan and early construction at Marathon are multi-year, not near-term, supply.

Macro transmits through two channels. US 10-year yield at 5.31% (+0.64%) and DXY at 102.1 (+0.17%) are both headwinds — a strong dollar and high real rates raise the discount rate on inventory holding and cap the copper/gold ratio's upside. But the copper/gold ratio at 1.57 (94th percentile, 1Y) says the market is already pricing industrial demand as resilient despite that rate backdrop.

View: physical balance is tight enough at the margin to support a base, not tight enough to force a squeeze. LME builds are the number to watch; a reversal to draws would be the bullish accelerant.

4. Positioning & Fund Flows

CFTC managed-money positioning has de-risked over the past month without capitulating.

As-of	OI	Long	Short	Net	Δ Net
2026-09-29	301,201	92,094	14,036	78,058	-4,464
2026-09-22	301,657	96,421	13,899	82,522	+17,416
2026-09-15	289,463	83,704	18,598	65,106	-17,048
2026-09-08	297,491	98,007	15,853	82,154	+9,272

The pattern is a violent two-way churn rather than a one-directional exit: net length swung +17,416 then -4,464 across the last two reports, with gross longs falling from 96,421 to 92,094 and shorts edging up from 13,899 to 14,036. The reduction in net is almost entirely long liquidation, not new short selling — a constructive distinction.

Crowding: netPct 25.92% of OI, crowding percentile 68.51 on a three-year window. This is elevated but not extreme — it is below the 2026-09-08 reading of 69.39 and the 2026-09-22 reading of 69.05. The CTA trend-following proxy sits at +62, unchanged across all four weeks, indicating trend systems remain net long copper. The hedge ratio at 54.82% is the lowest of the four weeks, meaning commercial hedging pressure has eased.

On volatility pricing, the copper-specific implied measure is not in the feed, but the complex context is informative: ^VXSLV at 36.6 (-0.3 pts) and ^GVZ at 23.18 (14th percentile, 1Y) show metals optionality is cheap relative to history, while ^VIX at 15.52 (19th percentile) shows macro event pricing is subdued. With copper RV20 at 27%, realised volatility is running at a level where long-gamma structures are not obviously overpriced.

View: positioning is a tailwind, not a headwind — the reset has occurred, the crowd is not extreme, and trend systems remain long. A rebuild of net length toward the 82,000 area is the likely path if 6.68 clears.

5. Cross-Asset Relative Value

The single relevant ratio in the feed is HG_GC_RATIO at 1.57 (ratio × 1000), 1Y percentile 93.65%, 3Y percentile 51.98%.

Read this carefully. A high one-year percentile means copper has been strong relative to gold over the trailing year — the ratio is near the top of its 12-month distribution. But the three-year percentile at 51.98% is almost exactly the median, meaning that over a longer window copper/gold is unremarkable. The correct interpretation is that the past year has been a copper-outperformance regime within a neutral multi-year structure, not that copper is historically expensive versus gold. A rising copper/gold ratio is a pro-growth signal, and at the 94th percentile the market is expressing a growth-resilient view.

The curve is in contango: M1-M2 at -0.0315 (-0.48%), roll yield -5.72%, slope 0.037. Contango of this magnitude is a carry cost for long positions — roughly 5.7% annualised — and it is a headwind to holding length passively. It is not a price cap and it is not a signal of oversupply by itself; it reflects the cost of storage and financing against a 5.31% ten-year yield. But it does mean that a long copper position must earn its return from price appreciation, not from roll.

The dollar and rates backdrop is the offsetting factor: DXY 102.1 (+0.17%) and ^TNX 5.31% (+0.64%). Both are headwinds for a dollar-denominated industrial metal. The fact that copper/gold is at the 94th percentile despite this backdrop is the relative-value argument for copper over gold on a growth view.

View: copper/gold at the 94th percentile supports a pro-growth posture; contango at -5.72% roll yield means the trade must be directional and time-bounded, not a carry hold.

6. Historical & Seasonal Patterns

The seasonality block covers the same calendar start, next 20 sessions, over the last 15 years: mean +1.46%, median +1.38%, up 10 of 15 years. The best instance was 2011 at +10.67%; the worst was 2012 at -7.62%.

A 10-of-15 hit rate is 66.7% — a genuine but modest edge, and the sample is small enough that it should be context, not conviction. The mean and median are close (+1.46% vs +1.38%), which indicates the distribution is not dominated by a single outlier in the central tendency, though the +10.67% best case shows the right tail is fat. The -7.62% worst case shows the left tail is equally real.

Applied to the 6.64 settle, the median seasonal path implies roughly 6.73 over the next 20 sessions, with the mean implying about 6.74. That sits just above R2 at 6.71 and below the 20-day/52-week high at 6.93 — a seasonal path that is consistent with a grind toward the range top rather than a breakout.

View: seasonality is a mild tailwind that reinforces the base case of a push toward 6.71–6.93; it does not by itself justify a position, and the 2012 precedent warns that October can break the other way.

7. Scenario Analysis

Base case — 55%: grind higher toward 6.71–6.93. Trigger: a daily settle above R1 at 6.68, confirming the pivot band has been reclaimed. Target: 6.71 (R2) first, then the 20-day/52-week high at 6.93. Action: hold existing length, add on the 6.68 settle confirmation, trail stops beneath 6.58 (S1). This path requires LME stock builds to stay modest (sub-1,000 MT per report) and the dollar to remain contained near 102. The seasonal median of +1.38% over 20 sessions is the supporting prior.

Bull case — 25%: breakout through 6.93. Trigger: a settle above 6.93 on rising open interest, most plausibly catalysed by a shift in LME warehouse stock from builds to draws or by a concentrate-supply headline out of the Anglo–Teck approval process. Target: 7.1–7.2, an extension measured off the 0.58-wide 20-day channel. Action: add to length on the breakout settle, move the stop to 6.68 (R1) to protect the position. This scenario is where the CTA proxy at +62 would force incremental trend buying.

Bear case — 20%: settle below 6.52 (S2). Trigger: a daily settle under 6.52, which would break the lower pivot shelf and put the 20-day channel floor at 6.35 in play. Catalysts: a credible Cobre Panamá restart timetable, an acceleration in LME builds beyond the recent 575 MT/day pace, or a dollar move through 103 on the 5.31% ten-year yield. Target: 6.35, then 6.2. Action: exit length on the 6.52 settle, stand aside; do not initiate shorts against a 94th-percentile copper/gold ratio without confirmation of a demand break.

Probabilities sum to 100%. The base case agrees with the section 1 call.

8. Trading Strategies & Risk Management

Strategy 1 — Long on pivot reclaim (primary). Entry: 6.68 on a daily settle above R1. Stop: 6.52, one cent below S2 and roughly 1.2 ATR from entry. Target: 6.93, the 20-day and 52-week high. Horizon: 1–3 weeks. Size: half of normal risk budget at initiation, scaling to full on a settle above 6.71 (R2). Conviction: 7/10. The rationale is that the 6.68 reclaim confirms the pivot band and opens the seasonal path toward the range top; the 6.52 stop sits beyond a real level rather than inside the 0.129 ATR daily noise.

Strategy 2 — Long on pullback (secondary). Entry: 6.58 on a daily settle holding S1. Stop: 6.45, below the S2 shelf and approximately one ATR from entry. Target: 6.71 (R2). Horizon: 5–10 sessions. Size: one-third of normal risk budget. Conviction: 6/10. This is the lower-risk entry if the market retests the pivot before advancing, but it requires the 6.58 shelf to hold on a closing basis; a settle below it invalidates the setup and hands the initiative to the bear case.

Risk management: total copper exposure should not exceed the normal single-asset risk budget across both strategies. The -5.72% roll yield means positions held beyond three weeks bleed carry, so both strategies are time-bounded. The FOMC minutes on 2026-10-08 (BJT 02:00) is the key event risk inside the horizon; reduce size into it if the position is already profitable.

9. This Week's Data Calendar

- **BJT 10-07 04:30	ET 10-06 16:30** — API Crude Oil Stock Change (OCT/02), USD, medium impact.
- **BJT 10-07 22:30	ET 10-07 10:30** — EIA Crude Oil and Gasoline Stocks Change (OCT/02), USD, medium impact.
- **BJT 10-08 02:00	ET 10-07 14:00 — FOMC Meeting Minutes, USD, high impact** (GC, SI, DXY) — the key event for copper's macro channel this week.
- **BJT 10-08 16:30	ET 10-08 04:30** — FOMC Member Waller speaks, USD, medium impact.
- **BJT 10-14 09:30	ET 10-13 21:30 — China CPI and PPI y/y, CNY, high impact** (HG, CL, ZS) — the demand-side read for copper.

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 2026-10-06

Gold (GC=F) Deep Dive | 2026-10-06 — Oversold Bounce Setup, LONG | 2026-10-06

Copper (HG=F) Deep Dive | 2026-10-06 — Constructive Base, 6.93 the Line | 2026-10-06

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