1. Price Action & Technical Analysis
Gold (GC=F) closed at 4263 on 2026-09-24, down 0.43% on the day, down 2.56 over the trailing five sessions, and down 7.87 over the trailing twenty sessions. The daily pivot for the session was 4303.3, with R1 at 4321.7 and S1 at 4268.7. Price settled below the pivot and marginally below S1, a bearish short-term configuration. The ATR reading of 98.38 is elevated relative to the 20-day volatility of 20.32%, implying that daily ranges remain wide enough to punish oversized positions. Open interest stood at 315,425 contracts, with volume at 106 (likely a partial or thin print) and change in position at 3.2%.
On the weekly timeframe, the sequence of lower closes is intact: 4398 on 2026-09-20, 4345.8 on 2026-09-21, 4338.9 on 2026-09-22, 4281.3 on 2026-09-23, and 4263 on 2026-09-24. That is a cumulative decline of 135 points, or roughly 3.07%, over five sessions. The 20-day change has deteriorated from -6.04 on 2026-09-20 to -7.87 on 2026-09-24, indicating that the sell-off is not merely a one-day event but a persistent trend.
The daily pivot at 4303.3 acts as a short-term equilibrium; price below it suggests sellers are in control. The 20-day change of -7.87 implies that the 20-day moving average is likely above current price, reinforcing the bearish bias. Without explicit MA values, we treat the pivot and the recent swing highs as the operative reference points. The 2026-09-20 close of 4398 is the most recent swing high, and the 2026-09-24 close of 4263 is the most recent swing low. A break below 4260 would confirm a lower low, while a reclaim of 4303 would neutralise the immediate downside.
However, the persistent lower closes and the negative 20-day change are consistent with a bearish momentum regime. The ATR of 98.38 is high, and the 20-day volatility of 20.32% is above typical historical averages for gold, suggesting that the market is in a high-volatility state. The Sharpe ratio over 30 days is -0.3243, indicating negative risk-adjusted returns over that window. The VaR95 of -2.85% suggests that a 2.85% daily loss is within the 95% confidence interval, which is material for leveraged positions.
Key technical levels to monitor: immediate support at 4260 (psychological and near S1), then 4200 (round number and potential measured move). Immediate resistance at 4303 (pivot), then 4321.7 (R1), then 4345-4350 (recent consolidation). The 2026-09-23 pivot of 4344.9 and R1 of 4376.5 are now overhead resistance. The 2026-09-22 pivot of 4368.4 and R1 of 4391.6 are further resistance. The 2026-09-21 pivot of 4395.2 and R1 of 4411.3 mark the top of the recent range. A close above 4345 would be the first sign of stabilisation; a close above 4395 would suggest a trend reversal.
In summary, the technical picture is bearish in the short term, with price below the daily pivot and below S1, and with momentum and volatility metrics confirming a corrective phase. The market is searching for a base, and the 4260 level is the first line of defence. If that fails, the next support is 4200. If the market can reclaim 4303, the immediate pressure would ease, but the broader trend would still be down until 4345 is reclaimed.
2. Fundamental Drivers
The primary fundamental driver in the current session is the rise in US real yields. The 10-year Treasury yield (^TNX) printed 5.16% on 2026-09-24, up 0.94% on the day. The dollar index (DXY) was 101.25, up 0.15%. Higher nominal yields, combined with the dollar's strength, increase the opportunity cost of holding gold, which pays no coupon. This is the classic macro headwind for the metal.
However, the revised UoM inflation expectations due on 2026-09-25 at 10:00 will be a key input for real-rate calculations. If inflation expectations rise, the real rate could fall even with nominal yields stable, which would be supportive for gold. Conversely, if expectations fall, the real rate would rise, adding pressure.
The absence of this data limits our ability to assess the structural bid. However, the COT data shows that non-commercial net length remains elevated at 133,116 contracts as of 2026-09-15, which suggests that speculative positioning is still crowded. The crowding score of 92.58 and netPct of 32.48% indicate that the long side is crowded, which is a vulnerability if the macro narrative turns more hawkish.
Inventories: SHFE warrant stocks were 116,031 kg on 2026-09-24, up 1,200 kg week-on-week. The increase in SHFE warrants is modest but suggests that physical tightness in China is not acute. The COMEX registered stock level is a key indicator of deliverable supply; without the weekly change, we cannot assess whether it is rising or falling.
The headlines from the past 48 hours include a Sina Futures article titled “每日论金 | 金价下跌，主要原因是这个” (Daily Gold: The main reason for the price drop) and “金价还能考虑逢低配置吗？ ” (Can gold still be considered for buying on dips?). These headlines suggest that the market is focused on the recent price decline and the question of whether to buy the dip. Mining news includes OceanaGold's 97-gram hit backing Waihi expansion and Kinross Gold cutting output and boosting shareholder payout. These are company-specific and do not directly drive the gold price, but they reflect the operating environment for producers.
The term structure is in contango, with M1-M2 at -30.5 (-0.71%) and the roll yield at -8.51%. A contango market implies that futures are more expensive than spot, which is typical when carrying costs (storage, insurance, financing) exceed convenience yield. The negative roll yield is a headwind for long-only futures positions, as rolling from one contract to the next incurs a cost. The slope of 20.61 indicates a relatively steep contango, which may encourage physical holders to sell forward and could weigh on spot prices.
Overall, the fundamental backdrop is challenging for gold in the near term: high nominal yields, a firm dollar, and a crowded long positioning. The offsetting factors would be a rise in inflation expectations, a dovish Fed pivot, or a geopolitical shock. The market is likely to remain under pressure until there is a clear change in the rate narrative.
3. Positioning & Fund Flows
The CFTC Commitments of Traders (COT) data for the four weeks ending 2026-09-15 shows a persistent reduction in net non-commercial length. On 2026-08-25, net length was 144,747 contracts (long 159,819, short 15,072), with a weekly change of +3,099. On 2026-09-01, net length fell to 136,771 (long 149,721, short 12,950), a change of -7,976. On 2026-09-08, net length was 134,972 (long 145,804, short 10,832), a change of -1,799. On 2026-09-15, net length was 133,116 (long 142,394, short 9,278), a change of -1,856. The trend is clear: longs are liquidating and shorts are covering, but the net reduction is driven more by long liquidation than by new shorts. The open interest in the COT report fell from 427,957 on 2026-08-25 to 409,899 on 2026-09-15, a decline of 18,058 contracts, confirming that positions are being reduced.
The crowding metrics show that the net position as a percentage of open interest (netPct) was 32.48% on 2026-09-15, down from 33.82% on 2026-08-25. The crowding score was 92.58, down slightly from 92.02, but still at a high level. The CTA positioning was 62, unchanged over the four weeks, and the hedge ratio was 14.85%, down from 15.95%. The high crowding score suggests that the long side is still crowded, which is a risk if the price continues to fall. However, the gradual decline in net length and the reduction in open interest indicate that the market is deleveraging, which could eventually create a cleaner base.
Options and volatility: The CBOE Gold Volatility Index (^GVZ) was 22.58 on 2026-09-24, down 0.83%. The VIX was 15.67, up 3.23%. The GVZ at 22.58 is elevated relative to the VIX, which suggests that gold-specific volatility is higher than broad equity volatility. This is consistent with the high ATR and 20-day volatility readings. The options market is likely pricing a wider range of outcomes for gold, which may be reflected in higher implied volatilities for both calls and puts. Without specific option open interest or skew data, we cannot assess the directional bias in options, but the elevated GVZ suggests that hedging demand is present.
The absence of this data means we cannot confirm whether the recent price decline has been accompanied by ETF outflows. Historically, ETF flows tend to follow price momentum, so a sustained decline could trigger further redemptions. Conversely, if ETFs are stable or rising, it would suggest a structural bid that could cushion the downside.
In summary, positioning is still crowded long, but the deleveraging process is underway. The reduction in open interest and net length is a necessary condition for a sustainable bottom, but it is not sufficient. The market needs to see a catalyst to reverse the trend. Until then, the risk of further long liquidation remains.
4. Cross-Asset Relative Value
The gold-silver ratio (GC_SI) was 66.81 on 2026-09-24, with a one-year percentile of 54.76% and a three-year percentile of 18.25%. This means that gold is relatively expensive compared to silver on a three-year basis, but closer to the middle of the one-year range. A high gold-silver ratio typically indicates risk aversion or industrial weakness, as silver has more industrial demand. The current ratio is not extreme, but the three-year percentile of 18.25% suggests that silver has been outperforming gold over the past three years, and the ratio is now near the lower end of its three-year range. If the ratio mean-reverts, silver could continue to outperform, or gold could catch up.
The copper-gold ratio (HG_GC) was 0.0016, with a one-year percentile of 91.27% and a three-year percentile of 47.62%. This ratio is a proxy for global growth expectations relative to safe-haven demand. A high copper-gold ratio suggests that copper is expensive relative to gold, which could indicate strong growth expectations or a supply deficit in copper. The one-year percentile of 91.27% is very high, meaning that copper has been strong relative to gold over the past year. This is a headwind for gold, as it suggests that the market is pricing a pro-growth environment rather than a risk-off one. However, the three-year percentile of 47.62% is more neutral, indicating that the current level is not extreme relative to the longer history.
The oil-gold ratio (CL_GC) was 0.0217, with a one-year percentile of 87.7% and a three-year percentile of 44.84%. A high oil-gold ratio suggests that oil is expensive relative to gold, which could be due to supply constraints or strong demand. The one-year percentile of 87.7% is high, meaning that oil has outperformed gold over the past year. This is another sign that the market is focused on growth and inflation rather than safe-haven demand. The three-year percentile of 44.84% is near the middle, suggesting that the current level is not extreme relative to the longer history.
Overall, the cross-asset ratios suggest that gold is relatively cheap compared to copper and oil on a one-year basis, but not on a three-year basis. The gold-silver ratio is in the middle of its one-year range but near the low end of its three-year range. These relative value signals are mixed. If the growth narrative fades, gold could outperform copper and oil, and the ratios could mean-revert. If the growth narrative strengthens, gold could continue to underperform. The key driver will be the trajectory of real yields and the dollar.
5. Sentiment & News Monitor
However, the headlines from the past 48 hours provide a qualitative read. The Sina Futures headlines are focused on the recent price decline and whether to buy the dip. The first headline, “每日论金 | 金价下跌，主要原因是这个” (Daily Gold: The main reason for the price drop), suggests that the market is trying to explain the decline, likely pointing to the rise in yields and the dollar. The second headline, “金价还能考虑逢低配置吗？ ” (Can gold still be considered for buying on dips?), indicates that investors are questioning whether the sell-off is a buying opportunity. This is a classic sentiment pattern during a correction: initial denial, then questions about buying the dip, then capitulation. The current tone is cautious but not panicked.
The mining headlines are company-specific: OceanaGold's 97-gram hit backing Waihi expansion is a positive exploration result, and Kinross Gold cutting output and boosting shareholder payout is a mixed signal (lower output but higher returns). These do not directly affect the gold price but reflect the operating environment.
Overall, the news bias over the past 48 hours is neutral-to-bearish, with a focus on the reasons for the decline and the question of whether to buy. There is no evidence of panic selling or a major bullish catalyst. The sentiment is likely cautious, which is consistent with the technical and fundamental picture.
6. Historical & Seasonal Patterns
However, we can make some general observations based on the time of year. Late September is typically a period of transition for gold. Historically, gold has often found a seasonal low in late September or early October, followed by a rally into the end of the year, driven by festive demand in India and China, as well as year-end portfolio adjustments. However, this is a tendency, not a guarantee. The current year is different because of the high nominal yield environment, which is a powerful headwind.
Without specific 10-year analogue data, we cannot quantify the probability of a seasonal rally. We note that the 52-week drawdown of 25.06% is significant, and the 20-day drawdown of 4.68% is moderate. The Sharpe ratio over 30 days is negative, indicating that the risk-adjusted returns have been poor. The market is in a corrective phase, and seasonal patterns may be less reliable when the macro backdrop is unfavourable.
We advise caution in relying on seasonal patterns without confirmation from price action and macro data.
7. Bull/Bear Scenario Analysis
Bull case (≥4 bullets):
- If the 10-year Treasury yield (^TNX) reverses lower from 5.16%, the opportunity cost of holding gold would fall, potentially triggering a short-covering rally. The first target would be the daily pivot at 4303.3, then R1 at 4321.7.
- If the dollar index (DXY) weakens from 101.25, gold would become cheaper for non-US buyers, supporting demand. A drop below 100 would be a significant bullish signal.
- If the revised UoM inflation expectations (due 2026-09-25) come in higher than expected, real yields could fall even if nominal yields remain stable, which would be supportive for gold. This could lead to a test of 4345-4350.
- If the crowded long positioning (netPct 32.48%, crowding 92.58) unwinds in an orderly way and the market finds a base at 4260, a contrarian rally could ensue as shorts cover. The 2026-09-20 high of 4398 would be the medium-term target.
- This is a low-probability, high-impact scenario.
Bear case (≥4 bullets):
- If the 10-year yield continues to rise above 5.16%, the opportunity cost of holding gold would increase further, pressuring the metal towards 4200. The next support after 4260 is 4200.
- If the dollar index strengthens above 101.25, gold would face additional headwinds. A break above 102 would likely accelerate the decline.
- If the revised UoM inflation expectations (due 2026-09-25) come in lower than expected, real yields would rise, reinforcing the bearish case. This could trigger a break below 4260.
- If the crowded long positioning continues to unwind, as it has over the past four weeks (net length down from 144,747 to 133,116), further long liquidation could push price towards 4200. The reduction in open interest from 427,957 to 409,899 confirms that positions are being closed.
- If the contango in the futures curve (M1-M2 -30.5, roll yield -8.51%) persists, long-only futures holders will continue to face negative roll costs, which could discourage new longs and encourage physical holders to sell forward.
Near-term balance (1-2 weeks): Bearish-to-neutral. The price is below the daily pivot and S1, momentum is negative, and the macro backdrop is unfavourable. The market is likely to test 4260 and possibly 4200. A reclaim of 4303 would neutralise the immediate downside, but the broader trend would remain down until 4345 is reclaimed.
Medium-term balance (1-3 months): Neutral-to-bearish. The high nominal yield environment is a structural headwind. However, if the market prices in a peak in yields, gold could find a base. The key will be the trajectory of real yields and the dollar. The crowded long positioning is a risk, but the deleveraging process is underway. A sustained break below 4200 would open the door to 4000. A sustained break above 4400 would signal a trend reversal.
8. Trading Strategies & Risk Management
Strategy 1: Tactical short on rallies. Direction: SHORT. Entry: 4300-4310 (near daily pivot 4303.3). Stop: 4345 (above recent consolidation and 2026-09-23 pivot 4344.9). Target: 4260 (S1 and psychological support), then 4200. Timeframe: 1-5 days. Conviction: 7/10. Size: 0.5-1.0% risk per trade. Rationale: Price is below the pivot, momentum is negative, and the macro backdrop is bearish. The stop is placed above the recent pivot to avoid noise. If price reclaims 4345, the short thesis is invalidated.
Strategy 2: Scale-in long on a confirmed base. Direction: LONG. Entry: 4260-4265 (if price stabilises and prints a bullish reversal candle). Stop: 4230 (below the 4260 support). Target: 4303 (pivot), then 4345. Timeframe: 3-10 days. Conviction: 5/10. Size: 0.3-0.5% risk per trade. Rationale: The 4260 level is a key support. If the market can hold it and reclaim the pivot, a short-covering rally could ensue. The stop is tight to limit losses if the support fails. This is a counter-trend trade, so conviction is lower.
Risk management: Given the ATR of 98.38 and 20-day volatility of 20.32%, position sizes should be conservative. The VaR95 of -2.85% suggests that a 2.85% daily loss is possible. Use stop-loss orders and avoid over-leveraging. The Sharpe ratio over 30 days is negative, so risk-adjusted returns are poor. Monitor the 10-year yield and the dollar index closely, as they are the primary drivers. The revised UoM inflation expectations on 2026-09-25 could be a volatility event. The BOE Gov Bailey speech on 2026-09-25 could also affect the dollar and gold. The China Manufacturing PMI on 2026-09-29 could impact industrial demand expectations and the copper-gold ratio.
9. This Week's Data Calendar
| Date | Time | Event | Importance |
|---|
| 2026-09-25 | 05:15 | BOE Gov Bailey Speaks | HIGH |
| 2026-09-25 | 10:00 | Revised UoM Consumer Sentiment | MEDIUM |
| 2026-09-25 | 10:00 | Revised UoM Inflation Expectations | MEDIUM |
| 2026-09-27 | 21:30 | China Industrial Profits ytd/y | MEDIUM |
| 2026-09-29 | 21:30 | China Manufacturing PMI | HIGH |
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.