1. Price Action & Technical Analysis
Copper (HG=F) settled Mon Sep 28 at 6.75, down 0.24% on the session. The five-day change is +2.37% and the twenty-day change is +2.8%, so the pullback is a pause inside a constructive medium-term advance rather than a reversal. The 20-day range is defined by the Sep 22 close of 6.836 at the top and the Sep 23 close of 6.754 at the bottom of the recent swing, with the Sep 24 close of 6.79 and Sep 25 close of 6.766 forming the upper-middle band. Monday's 6.75 close sits just below the Sep 23 low close, which makes the 6.75 area the immediate decision line.
The pivot structure is the cleanest guide. The Sep 25 pivot P was 6.765 with R1 at 6.804 and S1 at 6.733. The Sep 24 pivot was 6.771 with R1 6.817 and S1 6.739. The Sep 23 pivot was 6.807 with R1 6.869 and S1 6.702. The Sep 22 pivot was 6.829 with R1 6.874 and S1 6.787. The pivot band has drifted lower from 6.829 to 6.765 across four sessions, which is a mild loss of upward momentum, but the S1 shelf has held: 6.787, 6.702, 6.739, 6.733. Monday's close at 6.75 is above the most recent S1 of 6.733 and below the most recent P of 6.765, placing copper in the lower half of the immediate pivot range but not at the floor.
ATR14 is 0.1274 as of Sep 25, having eased from 0.1316 on Sep 22 and 0.1315 on Sep 23. Volatility is compressing modestly, which is consistent with a market digesting a fast move rather than trending violently. The 20-day volatility reading is 24.89% and the 30-day Sharpe is 1.053, so risk-adjusted returns remain positive. The 52-week drawdown is 13.49% and the 20-day drawdown is 7.03%, meaning the market is still well off its highs but has recovered a meaningful part of the recent dip.
Channel position is the key crowding tell. The channel position readings were 93.4% on Sep 22, 76.5% on Sep 23, 91.8% on Sep 24 and 89.3% on Sep 25. Copper is holding the top decile of its recent channel on most days, which is bullish momentum but also means there is little room for error. The 20-day high is 6.836 and the 20-day low is 6.754 on a closing basis; Monday's 6.75 close is marginally below that closing low, a small negative that argues for patience rather than chasing.
Volume is thin. The Sep 28 volume was 287 contracts, versus 21,880 on Sep 25, 23,036 on Sep 24, 37,215 on Sep 23 and 37,185 on Sep 22. The Monday print is a holiday-adjacent, low-liquidity session and should not be over-interpreted. Open interest in the futures contract was 174,125 on Sep 25, down from 174,830 on Sep 24 and 175,200 on Sep 23, but up from 171,866 on Sep 22. The net change over the week is a small decline, consistent with profit-taking rather than fresh shorting.
Putting it together: the trend is up on a 5-day and 20-day basis, the pivot band is drifting lower, the channel position is high, and the immediate support shelf is 6.733-6.739 with a deeper shelf at 6.702. Resistance is 6.765 (P), then 6.804 (R1), then 6.817-6.836. The tactical read is a retest buy at 6.75 with a stop below the S1 shelf and a target at the R1 band.
2. Fundamental Drivers
Rates and the dollar are the dominant macro inputs. The US 10-year yield (^TNX) is 5.184, up 0.43%, and the dollar index (DX-Y.NYB) is 100.97, down 0.32%. A high nominal 10-year yield is normally a headwind for copper because it raises the discount rate on industrial capital expenditure and tightens global financial conditions. The offset is that the dollar is softer on the day, which mechanically supports dollar-denominated metals. The combination of a firm yield and a softer dollar is a wash for copper, but the level of the 10-year at 5.18% is a genuine constraint on the bull case: it keeps the cost of carrying inventory and financing projects elevated, and it raises the odds that any demand disappointment is amplified by tighter credit.
The inflation backdrop is mixed. The upcoming Core PCE Price Index m/m for August is forecast at 0.3% versus 0.2% previously, and Personal Spending MoM is forecast at 0.8% versus 0.2% previously. If those prints land at or above forecast, the market will likely price a more hawkish Fed path, which would pressure copper via the rates channel even as it supports the inflation-hedge bid. The Fed commentary in the news flow is notable: Bessent urged the Fed to keep an open mind on rates, citing AI productivity. That is a dovish-leaning argument framed around supply-side productivity, and if it gains traction it would lower real-rate expectations and support copper. The BOJ minutes preview and the September rate hike reference add a layer of global policy uncertainty, but the direct copper transmission is second-order.
Inventories are the physical anchor. LME warehouse stock was 251,500 MT on Sep 25 and 251,175 MT on Sep 24, essentially flat. COMEX registered stock was 432,817.96 MT on Sep 24, a large buffer. SHFE warrant stock was 16,620 MT on Sep 24, down 2,468 MT week-on-week. The SHFE draw is the tightest part of the complex and is the strongest physical argument for the bulls, but the LME and COMEX buffers mean there is no global scarcity. The curve confirms this: the M1-M2 spread is -0.034, or -0.5%, which is contango, and the roll yield is -6.06% with a slope of 0.0339. Contango means nearby metal is cheaper than deferred, so there is no incentive to hoard and no squeeze pressure on shorts. A contango structure with high net length is a classic crowded-long setup: the position is large, but the physical market is not forcing anyone to cover.
Geopolitics is running through the oil channel. Oil prices jumped higher at the Globex open for the new week, Iran's Hormuz proposal was called a cynical bid for upfront concessions, and a Hormuz missile strike raised risk before the open. At the same time, Saudi crude exports via the US-guaranteed Hormuz route surged 80% to 6 million barrels per day, which pulled some supply-channel risk premium out of the market. Trump said talks with Iran would continue this week but that Tehran's demands were too high and its proposal had been rejected, with Qatar mediating. The net effect for copper is indirect: higher oil raises mining and freight costs and can lift the broad commodity complex, but a sustained oil spike that forces tighter monetary policy is a negative for industrial metals. The copper-specific headlines are minor: a Zimbabwe lithium mining variable, an NGO warning on a Donlin mine spill, and a trustless-asset comment. None of these change the copper supply-demand balance.
3. Positioning & Fund Flows
The COT data is the most important risk marker. As of Sep 22, open interest was 301,657, longs were 96,421, shorts were 13,899, and net was 82,522, a change of +17,416 from the prior week. The prior week, Sep 15, had net 65,106 with a change of -17,048. The week before that, Sep 8, had net 82,154 with a change of +9,272, and Sep 1 had net 72,882 with a change of -3,389. The pattern is a sharp rebuild of net length after a mid-September washout: net went 72,882 to 82,154 to 65,106 to 82,522. The Sep 22 print is the highest of the four weeks and the largest weekly build.
Crowding metrics confirm the stretch. Net as a percentage of open interest was 27.36% on Sep 22, up from 22.49% on Sep 15, 27.62% on Sep 8 and 25.79% on Sep 1. The crowding score was 69.05 on Sep 22, versus 63.36 on Sep 15, 69.39 on Sep 8 and 68.96 on Sep 1. CTA positioning is 98 in all four weeks, which is a persistent maximum-length signal from trend-following models. Hedge positioning was 56.61% on Sep 22, down from 58.27% on Sep 15, 56.46% on Sep 8 and 53.66% on Sep 1. The combination of a 27.36% net share, a 69.05 crowding score and CTA at 98 means the marginal buyer is largely already positioned. That does not make a top, but it caps the upside from here and raises the cost of being wrong.
The futures open interest in the contract data tells a slightly different story: 174,125 on Sep 25, 174,830 on Sep 24, 175,200 on Sep 23, 171,866 on Sep 22. The COT open interest of 301,657 is the broader reporting universe, while the contract-level OI is the front-month proxy. The contract-level OI rose into Sep 23 and then drifted lower, which is consistent with some long liquidation into the Sep 24-25 consolidation. That is a healthy sign: the market is not adding leverage into the highs.
Options and volatility are not showing stress. The VIX is 14.87, down 5.11%, which is a calm equity-vol backdrop. Copper's own 20-day volatility is 24.89%, and the VaR95 is -2.97%, so a one-day 95% loss scenario is roughly 3%. There is no options data in the set, so the vol read must come from realized measures and the VIX proxy. The message is that there is no panic bid for protection, which is typical of a grinding uptrend but also means a positioning unwind would be unhedged.
4. Cross-Asset Relative Value
The copper-gold ratio is 0.0016, with a one-year percentile of 90.87% and a three-year percentile of 47.49%. This is the single most important relative-value signal in the report. A one-year percentile of 90.87% means copper has outperformed gold over the past year by a wide margin; the ratio is near the top of its one-year range. The three-year percentile of 47.49% means the ratio is roughly mid-range over a longer horizon, so the one-year strength is not a three-year extreme. The mean-reversion implication is that the ratio is stretched to the upside on a one-year view, which favors gold over copper on a relative basis. For a copper-only mandate, this is a headwind: it argues against adding fresh length at the current ratio and favors buying dips rather than breakouts.
The ratio context also matters for the crowding read. Copper's high net length and the high copper-gold ratio are two expressions of the same macro trade: reflation, industrial demand optimism and a softer dollar. If that trade unwinds, both the ratio and the net length would compress together, which is why the crowded positioning is a bigger risk than the contango alone. The saving grace is the three-year percentile at 47.49%, which says the ratio is not in a bubble relative to its multi-year history.
There is no gold-silver or oil-gold ratio in the data, so the cross-asset read is limited to copper-gold. The oil headlines are the other cross-asset input: a higher oil price is a cost-push input for copper miners and a broad commodity tailwind, but it is also an inflation signal that can tighten policy. The net cross-asset message is that copper is the crowded leg of the reflation trade, and the relative-value screen says the easy money in that trade has been made.
5. Sentiment & News Monitor
Sentiment is cautiously constructive but not euphoric. The VIX at 14.87, down 5.11%, signals a calm macro backdrop. Copper's own risk metrics are moderate: 20-day volatility 24.89%, Sharpe30 1.053, VaR95 -2.97%. The 48-hour headline bias is mixed-to-supportive. The copper-specific headlines are low-impact: a Zimbabwe lithium mining variable, an NGO warning on a Donlin mine spill, and a trustless-asset comment. None of these are balance-sheet events for copper.
The macro headlines are more active. Oil jumped at the Globex open, Iran's Hormuz proposal was rejected as too expensive, a Hormuz missile strike raised risk, and Saudi exports via the guaranteed route surged 80% to 6 million barrels per day. The net oil signal is two-sided: geopolitical risk premium up, supply-channel risk premium down. The Fed commentary from Bessent is dovish-leaning and productivity-focused. The BOJ minutes preview and the September rate hike reference add policy uncertainty. On balance, the news flow does not justify a directional break; it justifies trading the range.
6. Historical & Seasonal Patterns
Using the historical percentiles in the data, copper's copper-gold ratio at the 91st one-year percentile is the standout: the metal has been a one-year outperformer against gold, and that is a mean-reversion warning rather than a momentum signal. The three-year percentile of 47.49% says the ratio is mid-range over a longer window, so the stretch is a one-year phenomenon. The 52-week drawdown of 13.49% says copper is still below its one-year high, so the one-year outperformance versus gold is relative, not absolute.
The one-year range context: the 20-day high is 6.836 and the 20-day low is 6.754 on a closing basis, and the 52-week drawdown of 13.49% implies the 52-week high is well above the current 6.75. The 20-day drawdown of 7.03% says the recent peak-to-trough move is contained. A negative roll yield of -6.06% is a carry cost for long holders, which is a structural headwind for passive length.
Seasonally, late September into early October is a period when Chinese demand signals matter most, and the calendar delivers exactly that: China Manufacturing PMI forecast 50.1 versus 49.8 previous, China Non-Manufacturing PMI forecast 49.3 versus 49.0 previous, and RatingDog Manufacturing PMI forecast 51.7 versus 51.5 previous. If the official manufacturing PMI prints above 50, it would be the first expansion signal in the forecast set and would support the retest-buy thesis. The historical percentile framework says copper is a relative-value short versus gold but an absolute-range long above 6.73.
7. Bull/Bear Scenario Analysis
Bull case:
- The 5-day change of +2.37% and 20-day change of +2.8% confirm an uptrend, and Monday's -0.24% close is a shallow pause, not a reversal.
- The S1 shelf at 6.733-6.739 has held across four sessions, and the deeper 6.702 pivot has not been tested.
- SHFE warrant stock fell 2,468 MT to 16,620 MT, the tightest part of the inventory complex.
- The dollar index at 100.97 is down 0.32%, a mechanical tailwind for dollar-denominated copper.
- China Manufacturing PMI is forecast at 50.1 versus 49.8 previous; a print above 50 would validate the demand-recovery leg.
- CTA positioning at 98 means trend models remain long, and the Sep 22 net build of +17,416 shows fresh institutional demand.
Bear case:
- Net length at 82,522 with a 27.36% net share and a 69.05 crowding score means the marginal buyer is already positioned.
- The curve is in contango (M1-M2 -0.034, -0.5%) with a -6.06% roll yield, so there is no squeeze forcing shorts to cover.
- LME stock at 251,500 MT and COMEX registered at 432,817.96 MT are ample global buffers.
- The copper-gold ratio at the 91st one-year percentile is stretched and favors mean reversion toward gold.
- The 10-year yield at 5.184 is a high discount-rate headwind for industrial capex and inventory financing.
- Monday's close at 6.75 is marginally below the Sep 23 closing low of 6.754, a small technical negative.
Near-term balance: the trend and the physical tightness in SHFE favor the bulls, but the crowding, contango and stretched copper-gold ratio cap the upside. The base case is a range between 6.7 and 6.84, with a bias to buy the 6.75 retest and sell into 6.8-6.84. Medium-term, the market needs either a China PMI beat or a dovish Fed surprise to break 6.84; absent that, the crowded long is vulnerable to a positioning flush toward 6.7.
8. Trading Strategies & Risk Management
Primary plan: buy the 6.75 retest. Entry 6.75, stop 6.725, target 6.803, horizon 1-5 days, conviction 6. The entry sits at Monday's close and just above the 6.733 S1 shelf; the stop at 6.725 is below that shelf and below the Sep 23 S1 of 6.702 on a closing basis, so a close below 6.725 invalidates the retest thesis and opens the 6.7 pivot zone. The target at 6.803 is the Sep 25 R1 of 6.804, a level that capped the Sep 24-25 bounce. Size at half of normal risk because the crowded positioning and contango argue against a full-size trend entry; the reward-to-risk on the level structure is approximately 2.1 to 1.
Invalidation: a daily close below 6.725. If that occurs, stand aside and wait for either a reclaim of 6.75 or a deeper flush into 6.7 with a stabilization signal. Do not add to the long below 6.725, and do not initiate a short, because the 5-day and 20-day trends remain positive and the SHFE warrant draw is a genuine tightness signal.
Risk management: the ATR14 of 0.1274 means the 6.725 stop is a tight, tactical stop; a wider stop at 6.7 would be more consistent with the daily range but would reduce the reward-to-risk below the target. Given the crowded long and the contango, the preferred approach is the tight stop with reduced size. Watch the China PMI prints on Wed Sep 30 Beijing time and the US Core PCE on Wed Sep 30 08:30 ET as the two events most likely to force a range break.
9. This Week's Data Calendar
| Time (New York / Beijing) | Event | Forecast | Previous |
|---|
| Tue Sep 29, 00:30 / Tue Sep 29, 12:30 | AUD Cash Rate | 4.6% | 4.35% |
| Tue Sep 29, 21:30 / Wed Sep 30, 09:30 | CNY Manufacturing PMI | 50.1 | 49.8 |
| Tue Sep 29, 21:30 / Wed Sep 30, 09:30 | CNY Non-Manufacturing PMI | 49.3 | 49.0 |
| Tue Sep 29, 21:45 / Wed Sep 30, 09:45 | CNY RatingDog Manufacturing PMI | 51.7 | 51.5 |
| Wed Sep 30, 08:30 / Wed Sep 30, 20:30 | USD Core PCE Price Index m/m | 0.3% | 0.2% |
| Wed Sep 30, 08:30 / Wed Sep 30, 20:30 | USD Final GDP q/q | 1.5% | 1.5% |
| Fri Oct 2, 08:30 / Fri Oct 2, 20:30 | USD Non-Farm Employment Change | 98K | 162K |
| Mon Oct 5, 10:00 / Mon Oct 5, 22:00 | USD ISM Services PMI | 54 | 55.4 |
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.