Copper (HG=F) Deep Dive | 2026-10-07 — Constructive Base, 6.61 Pivot Is the Line | 2026-10-07

Call: LONG copper (HG=F / HGZ26.CMX) on a 1–3 week horizon, with the trade invalidated on a settled break below the 6.61 S1 pivot. Three reasons underpin the call. First, the tape has stabilised: the 2026-10-06 settle was 6.65 (+0.13% 1D, +0.7% 5D) and the 2026-10-07 Asia print is 6.66 (+0.07% vs settle), holding the upper half of the 6.35–6.93 twenty-day channel (52.2% position) after the prior completed week's -3.21% washout. Second, visible inventory is being drawn where it matters — LME warehouse stock fell 3,750 MT (-1.51%) to 244,900 MT on 2026-10-05 and SHFE warrants dropped 721 MT (-6.72%) to 10,011 MT on 2026-09-30 — while COMEX registered stock was flat at 471,400 short tons. Third, the copper/gold ratio at 1.6 sits in the 98th one-year percentile, confirming copper's relative strength versus gold. Positioning is supportive but not extreme (net 78,058 lots, crowding 69th percentile), and 20-session seasonality is positive (median +1.76%, up 9 of 15 years).

1. Bottom Line & Directional Bias

Call: LONG copper (HG=F, referencing COMEX December 2026 HGZ26.CMX), invalidated on a daily settle below the 6.61 S1 pivot.

Three reasons carry the call. First, price structure has repaired rather than deteriorated: the 2026-10-06 settle of 6.65 sits above the 6.61 S1 pivot and in the upper-middle of the 6.35–6.93 twenty-day channel (52.2% position), with the 2026-10-07 Asia print at 6.66 (+0.07% vs settle) extending the recovery from the 6.51 low of the 2026-09-28–10-02 week. Second, the physical ledger is tightening at the margin — LME warehouse stock 244,900 MT on 2026-10-05, down 3,750 MT (-1.51%) d/d, and SHFE warrants 10,011 MT on 2026-09-30, down 721 MT (-6.72%) d/d — even as COMEX registered stock held flat at 471,400 short tons. Third, copper is outperforming gold: the copper/gold ratio at 1.6 sits in the 98th one-year percentile, a pro-growth signal that argues against a demand-collapse narrative.

The invalidation is explicit and mechanical: a settled close below 6.61 (S1) breaks the pivot shelf and reopens the 6.51–6.55 zone from the last completed week. A settled break below 6.58 (S2) would confirm trend damage and stand the call down.

2. Price Action & Technical Analysis

The prior session settle (2026-10-06) was 6.65, +0.13% on the day, +0.7% over five sessions and -2.55% over twenty sessions. The twenty-day range is 6.35–6.93, placing the settle at the 52.2% position — dead-centre, which is why the pivot grid rather than the channel extremes governs near-term risk. ATR14 is 0.127, or 1.91% of price as a full daily range; RV20 is 25.8%.

Pivots from the settle: P 6.65, R1 6.69, S1 6.61, R2 6.72, S2 6.58. The settle sits exactly on P, with the 2026-10-07 Asia range of 6.65–6.67 already probing toward R1. The last five settled bars tell a clean story of basing: 09-30 closed 6.62, 10-01 closed 6.54 (the low of the sequence at 6.51), 10-02 closed 6.55, 10-05 closed 6.64 and 10-06 closed 6.65 — three consecutive higher closes off the 6.51 low, with higher lows at 6.51, 6.52, 6.56, 6.61.

The last completed weekly bar (2026-09-28–2026-10-02) opened 6.77, high 6.77, low 6.51, closed 6.55, -3.21% w/w — a bearish weekly candle, but one that terminated at 6.51 and has since been reclaimed. The current week (from 2026-10-05, two sessions) is unfinished and last traded 6.65 (+1.54%); no weekly-close conclusion can be drawn from it. The 52-week range is 4.83–6.93, so the settle is 96% of the way up the annual range — this is a high-level consolidation, not a bottom-fishing exercise.

View: constructive above 6.61, with 6.69 (R1) the first gate and 6.72 (R2) the confirmation level. A settle back under 6.61 shifts the bias to neutral and under 6.58 to defensive.

3. Supply-Demand Balance & Fundamental Drivers

The visible inventory picture is mixed but net-supportive at the margin. LME warehouse stock was 244,900 MT on 2026-10-05, down 3,750 MT (-1.51%) d/d, though the twenty-report change is +8,425 MT — meaning the recent draw is a partial retracement of a larger rebuild. COMEX registered stock was 471,400 short tons on 2026-10-02, unchanged d/d. SHFE warrants were 10,011 MT on 2026-09-30, down 721 MT (-6.72%) d/d, a small absolute base but a sharp percentage draw that speaks to tightness in the Chinese deliverable pool.

The term structure is in contango: M1-M2 at -0.026 (-0.39%), with roll yield of -4.68% and slope 0.0316. This is a carry cost for longs, not a price cap — it tells us prompt physical is not acutely scarce, which is consistent with LME stock still being above its recent low. The practical implication is that a long position pays roughly 4.7% annualised to hold, so the trade needs a directional move, not a carry argument.

On the news flow, the 48-hour headlines are supply-side and mostly constructive for the medium term: the Puquios copper project secured authorisation to begin key water infrastructure in Chile, Canterra grew its Newfoundland copper resource 55% (albeit with falling grades), and Eastport identified additional nickel-copper anomalies in Botswana. None of these change near-term balances; they reinforce the multi-year mine-supply pipeline. The more relevant macro transmission is the energy shock headline — $264 billion wiped off top mining stocks — which, if it reflects an energy cost shock, raises smelter and mine operating costs and is ultimately cost-push supportive for copper, while simultaneously threatening demand if it feeds into a broader growth scare.

Macro transmission runs through the dollar and rates: DXY at 101.85 (-0.32%) and US 10-year yield at 5.269 (-0.79%) on 2026-10-06. A softer dollar and falling long-end yields are the two most direct macro tailwinds for a dollar-denominated industrial metal, and both moved in copper's favour into the settle.

View: the physical ledger is not tight enough to force a squeeze, but the direction of travel (LME and SHFE draws, flat COMEX) plus a softer dollar supports the long bias. Contango is a cost to manage, not a reason to avoid the trade.

4. Positioning & Fund Flows

CFTC managed-money positioning as of 2026-09-29: open interest 301,201 lots, longs 92,094, shorts 14,036, net 78,058, a week-on-week change of -4,464 lots. The prior week (2026-09-22) showed net 82,522 with a +17,416 weekly build, and 2026-09-15 showed net 65,106 with a -17,048 reduction. The pattern is a sharp two-week build into 2026-09-22 followed by a modest trim into 2026-09-29 — profit-taking into the 6.77 weekly high rather than a directional reversal.

Crowding is moderate: netPct of 25.92% on 2026-09-29 sits at the 69th percentile of the three-year distribution, down from 69.05 on 2026-09-22 and 69.39 on 2026-09-08. The CTA trend-following proxy is steady at 62 across all four weeks, and the hedge ratio has drifted from 58.27% (09-15) to 54.82% (09-29), meaning commercial hedging pressure has eased slightly as price fell.

The key divergence to note: net length fell 4,464 lots in the week to 2026-09-29 while price fell from the 6.77 weekly high to a 6.55 weekly close — positioning and price moved in the same direction, so there is no bullish divergence to exploit. What matters is that the subsequent three-session recovery to 6.65 has occurred without a reported rebuild in net length, which leaves room for fresh fund buying if 6.69–6.72 is reclaimed.

View: positioning is a tailwind, not a headwind. The 69th percentile crowding leaves capacity for incremental length; a push through 6.72 with rising open interest would confirm genuine new money rather than short-covering.

5. Cross-Asset Relative Value

The copper/gold ratio stands at 1.6 (ratio ×1000 basis), in the 98th percentile on a one-year window and the 58th percentile on a three-year window. The one-year reading is the striking one: copper has been structurally strong against gold over the past twelve months, and the ratio is near the top of its annual distribution. Read correctly, this is a pro-growth signal — the industrial metal is outperforming the monetary metal, which is not the configuration you see heading into a demand collapse.

The three-year percentile at 57.94% tempers the enthusiasm: relative to the longer window, copper/gold is only mid-range, so the one-year strength is a recovery within a broader range rather than a structural breakout. For a copper long, the ratio is confirmation rather than a fresh catalyst — it argues the demand side of the equation is intact, but it does not by itself generate the next leg higher.

Cross-referencing the volatility complex: VIX at 15.01 (12th percentile) and gold implied vol at 22.97 (14th percentile) both indicate a market pricing very little macro stress. Copper's own RV20 at 25.8% is the highest of the three readings.

View: the copper/gold ratio supports the long bias on a one-year view but is not stretched enough on a three-year view to be a contrarian warning. Use it as confirmation, not as the trade trigger.

6. Historical & Seasonal Patterns

Seasonality for the same calendar start (early October) over the next twenty sessions, last fifteen years: mean +1.52%, median +1.76%, positive in 9 of 15 years. The best instance was 2011 at +9.08%; the worst was 2012 at -7.62%. The distribution is roughly symmetric around a modest positive median, with the tails wide — a 9-of-15 hit rate is a 60% base rate, which is a mild edge, not a strong one.

The honest read is that this is context, not a signal. A 60% hit rate with a +1.76% median over a fifteen-observation sample carries wide error bars, and the -7.62% worst case is more than four times the median in magnitude. What seasonality does provide is a reason not to fight the long side on a purely calendar basis, and a reminder that October has historically been capable of sharp two-way moves in copper.

Combined with the technical setup — price basing above the 6.61 pivot after a washout to 6.51 — the seasonal window is a mild tailwind that reinforces the constructive stance without independently justifying it.

View: seasonality is a supporting factor for the long bias, with a 60% historical hit rate and a +1.76% median move over the next twenty sessions. It does not change the invalidation level at 6.61.

7. Scenario Analysis (Base / Bull / Bear)

Base case — 50% probability: grind higher toward 6.72–6.77. Trigger: the 2026-10-07 Asia print at 6.66 holds above the 6.65 pivot and the market takes out R1 at 6.69 on a settled basis. Target: 6.72 (R2) initially, then the 6.77 high of the last completed week. Action: hold existing length, add on a settled close above 6.69, keep stops below 6.61. This scenario is consistent with the section 1 call and requires no change in the physical or positioning backdrop — merely continuation of the three-session higher-close sequence.

Bull case — 25% probability: breakout through 6.77 toward 6.93. Trigger: a settled close above 6.72 (R2) accompanied by a rebuild in CFTC net length above the 78,058 reading and a further LME draw below 244,900 MT. Target: 6.93, the twenty-day and 52-week high. Action: add to length on the breakout, trail stops to 6.65, and treat 6.93 as the first scale-out zone. The bull case is helped by the softer dollar (DXY 101.85, -0.32%) and falling 10-year yield (5.269, -0.79%), and by the 98th percentile copper/gold ratio confirming pro-growth conditions.

Bear case — 25% probability: loss of 6.61 opens 6.51–6.55. Trigger: a settled close below 6.61 (S1), with confirmation below 6.58 (S2). Target: 6.55, then 6.51 — the low of the last completed week. Action: exit length on the 6.61 settle, stand aside, and re-engage only on a reclaim of 6.65. The bear case is supported by the -4.68% roll yield (a persistent cost to holding), the +8,425 MT twenty-report build in LME stock, and the fact that the last completed weekly bar was -3.21% w/w. A hawkish FOMC minutes read on 2026-10-08 would be the most likely catalyst.

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

8. Trading Strategies & Risk Management

Strategy 1 — Core long (conviction 7/10). Entry 6.65 (at the pivot, working the current Asia range of 6.65–6.67), stop 6.58 (below S2, roughly 0.07 below entry and beyond the 6.61 pivot shelf), target 6.72 (R2) for the first scale, 6.77 for the second. Horizon 1–3 weeks. Size: 60% of intended copper risk budget, with the remaining 40% reserved for the breakout add above 6.69. The stop sits beyond a real level and outside the noise band implied by ATR14 of 0.127.

Strategy 2 — Breakout add (conviction 6/10). Entry on a settled close above 6.69 (R1), stop 6.61 (back below the pivot), target 6.93 (the twenty-day and 52-week high). Horizon 1–2 weeks. Size: the remaining 40% of the risk budget. This leg is only triggered if the base case resolves upward; if 6.61 settles first, both legs are cancelled and the book is flat.

Risk management: total copper risk is capped at the two-leg budget, and the invalidation at 6.61 is a settled-close condition, not an intraday touch — the 2026-10-07 Asia low of 6.65 shows the market is not currently testing it. The -4.68% roll yield means the position carries a cost of roughly 0.09% per week at current price, which is immaterial over a 1–3 week horizon but argues against holding indefinitely without progress. VaR95 of -2.97% and 20-day max drawdown of 7.03% frame the tail: a single bad session can take out roughly 3% and the stop must be respected mechanically.

9. This Week's Data Calendar

BJT 10-07 22:30 | ET 10-07 10:30 — EIA Crude Oil and Gasoline Stocks Change (OCT/02), medium impact, relevant to the energy complex rather than copper directly. BJT 10-08 02:00 | ET 10-07 14:00 — FOMC Meeting Minutes, high impact, transmits to copper via DXY and the 10-year yield. BJT 10-08 16:30 | ET 10-08 04:30 — FOMC Member Waller speaks, medium impact. BJT 10-14 09:30 | ET 10-13 21:30 — China CPI and PPI y/y, high impact, the key demand-side print for HG.

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.

單品種深度報告深度分析

COMEX黃金GC=F 2026-10-07

Gold (GC=F) Deep Dive | 2026-10-07 — Long-side mean reversion, crowded but not broken | 2026-10-07

Copper (HG=F) Deep Dive | 2026-10-07 — Constructive Base, 6.61 Pivot Is the Line | 2026-10-07

今日風險地圖

單品種深度報告 深度分析

往期研報封存

⟠ 本月策略回測表現

單品種深度報告深度分析