1. Price Action & Technical Analysis
Gold (GC=F) ended the week of September 15–20, 2026, on a soft note, closing at 4412.7 on September 20, down 0.28% on the day. Over the past five sessions, the metal has gained 1.4%, but the 20-day change remains negative at -5.72%, indicating a broader corrective phase. The daily closes for the week were: Sep 15: 4291.6 (-1.71%), Sep 16: 4346.3 (+1.27%), Sep 17: 4360.2 (+0.32%), Sep 18: 4424.9 (+0.57%), Sep 20: 4412.7 (-0.28%). This sequence shows a sharp rebound from the Sep 15 low, followed by consolidation near the 4420 area.
On the daily chart, the pivot point (P) for Sep 20 is 4414.6, with R1 at 4418.3 and S1 at 4409.1. The close of 4412.7 is marginally below the pivot, suggesting a neutral-to-bearish intraday bias. The 5-day change of +1.4% shows a short-term bounce. The Average True Range (ATR) for Sep 20 is 106.44, down from 111.47 on Sep 18 and 118.39 on Sep 17, indicating declining volatility. This contraction in ATR often precedes a breakout, but direction is not predetermined.
The 20-day change is negative, so the 20-day simple moving average (SMA) is likely above the current price. The 5-day change is positive, so the 5-day SMA may be turning higher. However, the drawdown from the 52-week high (DD52w) is 25.06%, which is substantial and suggests the metal is in a longer-term correction or bear market phase. The 20-day drawdown is 7.52%, reflecting the recent pullback.
However, we can use the price action and ATR to infer that momentum is mixed. The sharp rebound from 4291.6 to 4424.9 (a 3.1% move in three days) suggests oversold conditions were met with buying, but the failure to hold above 4420 indicates selling pressure. We rely on the provided pivots for Sep 20: P=4414.6, R1=4418.3, S1=4409.1. These are very tight, reflecting the low ATR relative to price.
On the weekly chart, the 5-day change of +1.4% suggests a modest gain for the week, but the 20-day change of -5.72% indicates a lower high and lower low pattern over the past month. The weekly close of 4412.7 is below the 20-day pivot, and the 20-day drawdown of 7.52% shows the recent peak was around 4770 (calculated as 4412.7 / (1 - 0.0752) ≈ 4771). This level is a key resistance zone. The 52-week drawdown of 25.06% implies the 52-week high was approximately 5888 (4412.7 / (1 - 0.2506) ≈ 5888). That is a distant resistance.
On the monthly chart, the picture is less clear without longer-term data, but the 25% drawdown from the 52-week high suggests a significant correction. The Sharpe ratio (30-day) is 1.756, which is positive and indicates that recent returns have been favorable relative to volatility, despite the drawdown. The 20-day volatility is 21.82%, which is elevated. The VaR95 (Value at Risk at 95% confidence) is -2.85%, meaning that over the next day, there is a 5% chance of a loss exceeding 2.85%.
Key technical levels: Immediate support is at the Sep 15 close of 4291.6, which also aligns with the S1 pivot from Sep 15 (4303.5) and the low of that day. Immediate resistance is at the Sep 18 close of 4424.9, followed by the Sep 18 R1 of 4452.4. A break above 4452 would target the 20-day high around 4770. The ATR of 106.44 suggests that daily ranges are about 2.4% of price, so a move to 4452 is within one ATR.
In summary, gold is in a short-term bounce within a medium-term downtrend. The declining ATR and tight pivots suggest a potential breakout, but the crowded positioning and macro headwinds favor a cautious approach. We would need to see a close above 4452 to confirm a bullish reversal, while a close below 4291 would signal a resumption of the downtrend.
2. Fundamental Drivers
Interest rates and the US dollar are primary drivers for gold. The 10-year Treasury yield (^TNX) stands at 5% as of September 19, up 1.03% on the day. However, a 5% nominal yield is historically restrictive for gold, as it increases the opportunity cost of holding a non-yielding asset. The US Dollar Index (DXY) is at 100.25 as of September 20, up 0.03%. A strong dollar makes gold more expensive for foreign buyers, weighing on demand. The combination of high yields and a firm dollar is a classic bearish setup for gold.
Without this, we cannot assess whether real yields are positive or negative. However, the high nominal yield suggests that if inflation is moderating, real yields could be rising, which is bearish for gold. Conversely, if inflation is sticky, real yields may be lower, providing support. The market's focus on the 5% yield level indicates that further increases could trigger risk-off moves that might benefit gold as a safe haven, but the correlation has been inconsistent.
SHFE warrant inventory is 114,831 KG as of September 18, down 12 KG week-over-week. These are relatively small changes and do not indicate a significant shift in physical supply-demand. However, the COT data shows that managed money net length is still large, suggesting that speculative demand remains a factor.
This is a critical missing piece, as ETF flows often reflect institutional sentiment. Without it, we rely on COT and price action.
However, the presence of several central bank speaker events (ECB, BOC, RBA, SNB) in the coming week suggests that monetary policy divergence could be a driver. The SNB policy rate decision on September 24 is high importance; a surprise cut could weaken the Swiss franc and boost gold in CHF terms, but the impact on USD gold is indirect. The RBA Governor speech on September 21 is high importance for AUD, which can influence gold via commodity currency correlations.
Term structure: The gold futures curve is in contango, with the M1-M2 spread at -17.6 (or -0.4%). The roll yield (RY) is -4.79%, and the slope is 18.78. Contango means that longer-dated futures are more expensive than near-dated, which is typical for a storable commodity with carrying costs. For investors rolling long positions, contango creates a negative roll yield, which is a headwind for passive long strategies. The annualized roll yield of -4.79% is significant and could deter some investors. This structure suggests that physical demand is not tight enough to cause backwardation, which is a bearish signal for the spot price.
Overall, the fundamental backdrop is mixed. High nominal yields and a strong dollar are bearish, while central bank buying and geopolitical uncertainty are supportive. The contango structure and crowded positioning add caution. We would need to see a decline in yields or a weaker dollar to shift the fundamental bias to bullish.
3. Positioning & Fund Flows
The CFTC Commitments of Traders (COT) data for gold futures as of September 15, 2026, shows open interest (OI) of 409,899 contracts, with long positions at 142,394 and short positions at 9,278, resulting in a net long of 133,116. This net long has decreased by 1,856 contracts from the previous week. The trend over the past four weeks is as follows: Aug 25: net 144,747 (Δ +3,099), Sep 1: net 136,771 (Δ -7,976), Sep 8: net 134,972 (Δ -1,799), Sep 15: net 133,116 (Δ -1,856). This shows a steady reduction in net length over three consecutive weeks, but the absolute level remains high.
The crowding score is 92.58 as of September 15, up from 92.2 the prior week. This score, likely a percentile, indicates that net positioning is more crowded than 92.58% of historical observations. The net percentage of open interest (netPct) is 32.48%, down slightly from 32.82% the prior week. The CTA (Commodity Trading Advisor) positioning is 62, unchanged over the four weeks, suggesting that trend-following funds have maintained a stable long bias. The hedge percentage is 14.85%, down from 15.6% the prior week, indicating that commercial hedgers are slightly less short.
The high crowding score is a double-edged sword. On one hand, it reflects strong conviction in the bullish narrative. On the other hand, it increases the risk of a sharp unwind if the price breaks key support levels. The gradual reduction in net length suggests that some longs are taking profits, which could be a healthy consolidation. However, if the price falls below 4291.6, we could see a more aggressive liquidation, potentially driving prices toward 4200 or lower.
Options and volatility: The CBOE Gold Volatility Index (^GVZ) is at 23.31 as of September 19, down 6.69% on the day. This is a moderate level, below the highs seen during panic episodes but above the lows of calm markets. The VIX is at 14.81, down 4.08%, indicating low fear in the broader equity market. The decline in GVZ suggests that option premiums are shrinking, which could be a sign of complacency. If volatility increases, gold could see a sharp move, but direction is uncertain.
The COT data is the primary source of positioning information. The open interest in futures has declined from 427,957 on Aug 25 to 409,899 on Sep 15, a drop of 4.2%, indicating that some traders are leaving the market. This could be due to reduced volatility or profit-taking. The volume on Sep 18 was 169,188 contracts, but on other days it was very low (1 or 2 contracts), which is likely a data artifact. The OI on the GC=F contract from the price data is 314,133, which is different from the COT OI of 409,899; this discrepancy may be due to different reporting conventions. We use the COT OI for positioning analysis.
In summary, positioning is crowded long, but the gradual decline in net length suggests that the market is not adding to longs aggressively. This is a neutral-to-bearish signal for the short term, as there is potential for further long liquidation. A break below support could trigger a cascade of selling.
4. Cross-Asset Relative Value
Cross-asset ratios provide insight into gold's relative valuation. The copper-gold ratio (HG_GC_RATIO) is 0.0015, which is in the 73rd percentile of the past year and the 35th percentile of the past three years. This means that copper is relatively expensive compared to gold over the past year, but less so over three years. A high copper-gold ratio often indicates strong industrial demand and a risk-on environment, which can be bearish for gold. The current level suggests that copper has outperformed gold recently, but the three-year percentile is lower, indicating that the ratio is not extreme by longer-term standards.
The gold-silver ratio (GC_SI_RATIO) is 66.48, in the 51st percentile of the past year and the 17th percentile of the past three years. This means that gold is relatively expensive compared to silver over the past year (above median), but over three years, it is in the lower percentile, meaning silver has been stronger. The ratio is often used as a risk sentiment gauge; a high ratio indicates risk aversion (gold outperforming silver), while a low ratio indicates risk appetite. At 66.48, it is above the historical average of around 60, suggesting some risk aversion. However, the three-year percentile of 17.06% indicates that the ratio has been much higher in the past, so it is not at an extreme.
The oil-gold ratio (CL_GC_RATIO) is 0.0227, in the 93rd percentile of the past year and the 47th percentile of the past three years. This is a very high percentile for the past year, meaning oil is expensive relative to gold. This is unusual because oil and gold are both commodities, but they respond to different drivers. A high oil-gold ratio can indicate inflationary pressures, which could be bullish for gold as a hedge. However, it can also indicate that gold is undervalued relative to oil, which might attract value buyers. The three-year percentile is near the median, so the ratio is not extreme by longer-term standards.
Overall, the cross-asset ratios suggest that gold is relatively expensive versus copper and silver on a one-year basis, but relatively cheap versus oil. However, the high oil-gold ratio could be a warning sign of inflation, which might eventually support gold. The copper-gold ratio at the 73rd percentile suggests that industrial metals are leading, which is typically a risk-on signal that is bearish for gold. The gold-silver ratio at the 51st percentile is neutral.
We also note the term structure contango, which we discussed in section 2. The roll yield of -4.79% is a headwind for long positions. In comparison, other commodities may have different roll yields, but we do not have that data. The contango suggests that the convenience yield for gold is low, meaning there is no shortage of physical metal.
In conclusion, relative value does not offer a compelling case for gold at current levels. The metal is not cheap versus copper or silver, and the oil-gold ratio is elevated but not extreme. We would need to see a decline in the copper-gold ratio or a rise in the gold-silver ratio to signal a shift in favor of gold.
5. Sentiment & News Monitor
The VIX at 14.81 suggests low fear in the broader market. The upcoming central bank speeches (Lagarde, Macklem, Bullock) could generate headlines that affect gold. For example, if Lagarde hints at further ECB tightening, it could strengthen the euro and weaken the dollar, potentially supporting gold. Conversely, if Bullock signals a dovish RBA, it could weaken the AUD and have a mixed impact on gold. The SNB policy rate decision on September 24 is high importance; a cut could weaken the CHF and boost gold in CHF terms, but the impact on USD gold is indirect. Overall, sentiment is cautiously bullish based on positioning, but the lack of fresh news and the crowded trade suggest that the market is vulnerable to a shift in narrative. We would need to see a break of key technical levels to confirm a change in sentiment.
6. Historical & Seasonal Patterns
Typically, September is a mixed month for gold, with some years showing strength due to festival demand in India and China, and others showing weakness due to dollar strength. The 52-week drawdown of 25.06% is significant; historically, such drawdowns have sometimes been followed by rebounds, but past performance is not indicative of future results. We would need to analyze the current macro environment relative to past periods to make a judgment. Given the data gap, we cannot provide a seasonal or historical edge. This section is therefore limited to noting the absence of data.
7. Bull/Bear Scenario Analysis
Bullish scenarios (≥4 bullets):
- If the 10-year Treasury yield (^TNX) falls below 4.8%, then gold could rally as the opportunity cost of holding the metal decreases. The current yield is 5%, so a decline of 20 basis points would be a significant move.
- If the US Dollar Index (DXY) breaks below 99.5, then gold could benefit from a weaker dollar, making it cheaper for foreign buyers. The current DXY is 100.25.
- If central bank buying accelerates, as evidenced by an increase in COMEX or SHFE inventories or news of large purchases, then gold could see a sustained bid. The SHFE warrant inventory fell by 12 KG last week, which is a small decline, but a larger drawdown would be bullish.
- If geopolitical tensions escalate, particularly in the Middle East or Eastern Europe, then gold could attract safe-haven flows.
- If the COT net length continues to decline but price holds above 4291.6, then the market could be forming a base for a rally. The crowding score is high, but a reduction in net length without a price breakdown would be a bullish divergence.
Bearish scenarios (≥4 bullets):
- If the 10-year Treasury yield rises above 5.2%, then gold could sell off as real yields become more attractive. The current yield is 5%, and a break above 5.2% would be a new high for the cycle.
- If the US Dollar Index (DXY) breaks above 101, then gold could face headwinds from a stronger dollar. The current DXY is 100.25.
- If the COT net length continues to decline and price breaks below 4291.6, then a long liquidation could accelerate, driving prices toward 4200. The crowding score of 92.58 indicates that there are many longs who could be forced to sell.
- If the term structure moves further into contango, with the M1-M2 spread widening beyond -20, then the negative roll yield would become more punitive for long holders, potentially deterring investment.
- If the gold-silver ratio falls below 60, it could indicate a shift to risk-on sentiment, which is typically bearish for gold. The current ratio is 66.48.
Near-term balance (1-2 weeks): The market is at a crossroads. The rebound from 4291.6 has stalled near 4420, and the declining ATR suggests a breakout is imminent. However, the crowded positioning and high yields favor the bears. We would give a slight edge to the bearish scenario in the near term, with a target of 4300 if support breaks. A close above 4452 would negate this view.
Medium-term balance (1-3 months): The fundamental drivers are mixed. Central bank buying and geopolitical risks provide support, but high yields and a strong dollar are headwinds. The 25% drawdown from the 52-week high suggests that the metal is in a corrective phase. We would need to see a shift in monetary policy or a weakening of the dollar to turn bullish. Until then, we expect range-bound trading between 4200 and 4600.
8. Trading Strategies & Risk Management
Strategy 1: Short-term bearish (LONG volatility via options or SHORT futures). Given the crowded long positioning and the failure to hold above 4420, we favor a short position on a break below 4400. Entry: 4395 (stop limit), Stop: 4455 (above the Sep 18 high), Target: 4300 (near the Sep 15 low), Timeframe: 1-5 days, Size: 1% risk per trade. Conviction: 7/10. This trade aligns with the bearish near-term bias. If the price instead breaks above 4455, the stop would be triggered, limiting losses.
Strategy 2: Medium-term bullish (LONG futures or call options). For investors with a longer horizon, we see value in buying gold on dips toward 4300, as central bank demand and geopolitical uncertainty provide a floor. Entry: 4310 (limit), Stop: 4250 (below the Sep 15 low), Target: 4550 (near the 20-day high), Timeframe: 1-3 months, Size: 2% risk per trade. Conviction: 6/10. This trade requires patience and a tolerance for drawdown. If the price breaks below 4250, the stop would be triggered.
Risk management: Use ATR-based stops. The current ATR is 106.44. Position sizing should be adjusted to account for the 20-day volatility of 21.82%. The VaR95 of -2.85% suggests that daily losses can be significant. We recommend risking no more than 1-2% of capital per trade. Diversification across other assets is also advised. Do not use excessive leverage. Monitor the COT data for signs of further long liquidation. The next COT release is on September 22 (for the week ending September 15), which could provide fresh positioning clues.
9. This Week's Data Calendar
| Date | Time (UTC) | Event | Importance |
|---|
| 2026-09-21 | 11:00 | ECB President Lagarde Speaks | MEDIUM |
| 2026-09-21 | 11:05 | BOC Gov Macklem Speaks | MEDIUM |
| 2026-09-21 | 23:10 | RBA Gov Bullock Speaks | HIGH |
| 2026-09-22 | 07:00 | ECB President Lagarde Speaks | MEDIUM |
| 2026-09-23 | 03:15 | French Flash Manufacturing PMI | MEDIUM |
| 2026-09-23 | 03:15 | French Flash Services PMI | MEDIUM |
| 2026-09-23 | 03:30 | German Flash Manufacturing PMI | MEDIUM |
| 2026-09-23 | 03:30 | German Flash Services PMI | MEDIUM |
| 2026-09-23 | 04:30 | GBP Flash Manufacturing PMI | MEDIUM |
| 2026-09-23 | 04:30 | GBP Flash Services PMI | MEDIUM |
| 2026-09-23 | 21:30 | AUD Employment Change | HIGH |
| 2026-09-23 | 21:30 | AUD Unemployment Rate | HIGH |
| 2026-09-24 | 03:30 | CHF SNB Monetary Policy Assessment | HIGH |
| 2026-09-24 | 03:30 | CHF SNB Policy Rate | HIGH |
| 2026-09-24 | 04:00 | CHF SNB Press Conference | HIGH |
| 2026-09-24 | 08:30 | CAD Core Retail Sales m/m | MEDIUM |
| 2026-09-24 | 08:30 | CAD Retail Sales m/m | MEDIUM |
| 2026-09-24 | 08:30 | USD Unemployment Claims | MEDIUM |
| 2026-09-24 | 09:00 | CNY CB Leading Index m/m | LOW |
| 2026-09-24 | 19:01 | CNY Bank Holiday | LOW |
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.