I. Bottom Line & Directional Bias
Direction: bullish. Three core reasons: First, the volume-price structure is healthy. AU2612 [10-09 settle] CNY 899.1/g, +1.55% on the day, main contract open interest 228,458 lots (+1,400 lots), main contract notional open interest CNY 207.59bn, daily change +CNY 1.308bn — an open-interest-build advance led by longs, not a short-covering rebound. Second, international and China markets are resonating. GC=F [10-09 settle] USD 4,216.3/oz, +1.43% on the day, +1.3% over 5 days; SHFE Gold, after tax adjustment, trades at a +0.16% premium to GC=F (+0.03% five days ago) — the onshore-offshore spread is widening modestly but is not extreme, and the import window leans open rather than closed. Third, volatility pricing is not overheated. GVZ at 22.79 sits in the 12th percentile of the past year; GC=F RV20 at 17.9%, IV−RV at +4.9 vol pts, IV/RV at 1.27 — options are paying for event risk but at a low absolute level, so upside is not exhausted.
Direction-invalidation conditions: the bullish thesis fails if AU2612 closes below 895.23 (pivot S1) on the daily chart and main contract open interest turns to a net cut; it is also deemed invalidated if the 10-14 US core CPI y/y beats expectations by more than 2.5% and pushes DXY above 103. Trigger-condition response: main contract open interest sits in the 95th percentile of the past year, so crowding is the largest risk source this issue — position management takes priority over directional judgment.
II. Price Action & Technicals
Settlement basis: AU2612 [10-09 settle] CNY 899.1/g, +1.55% versus the prior settlement; −2.05% over 5 days and −5.19% over 20 days — the medium term remains in a repair channel rather than a trend acceleration. The 20-day range is 889.82–977.26; the current price sits at the 21.5% position of the range, near the lower bound. ATR14 is 15.51 (1.71% of the settlement price, full intraday range), RV20 is 18.9% — realized volatility exceeds the average daily move implied by ATR, indicating unusually large single-day amplitudes recently. Pivots: P 903.23, R1 916.63, S1 895.23. The current price of 899.1 is about CNY 4.1 below P, about CNY 3.9 from S1 and about CNY 17.5 from R1 — all within half an ATR (CNY 7.76), i.e. not a meaningful trading distance — so P and S1 do not constitute standalone entry triggers and serve only as risk references.
The last five daily candles (closing-price basis, not settlement) show a “dip-then-recover” structure: 09-28 closed 904.66, 09-29 closed 898.78, 09-30 closed 910.58, 10-08 closed 893.12, 10-09 closed 908.64. The 10-08 low of 891.1 and the 10-09 low of 889.82 form a double low, and the 10-09 close of 908.64 has essentially recovered all of 10-08's decline — the short-term center of gravity is shifting up.
Report-day session: SHFE Gold night session 10-09 21:05 to 10-12 00:00 opened 904.36, high 909.44, low 903.72, closed 906.66 — −0.22% versus the day session close and +0.84% versus settlement; the night session consolidated in a narrow band and did not damage 10-09's rebound structure. International comparison: GC=F in the Asian session on 2026-10-12 at 07:00 was quoted at 4,209.1, down 0.17% from the [10-09 settle] of 4,216.3, with H 4,222.22 and L 4,209.1 — a modest international give-back but limited in magnitude.
Weekly basis: the last completed weekly candle, 2026-10-08 to 10-09, opened 893.1, high 911.22, low 889.82, closed 908.64, with a weekly change of −0.21% — a narrow consolidation week with no directional breakout; there are no settled daily candles yet this week, so no weekly conclusion is drawn.
Key levels: overhead, R1 at 916.63 is the first resistance; a decisive break and hold above it implies a measured target of 932.14 using one ATR (CNY 15.51). Below, S1 at 895.23 is the first support; a daily close below it points to the 20-day low of 889.82 as the next reference.
III. Supply-Demand Balance & Fundamental Drivers
Inventory: COMEX GOLD registered at 15.07 Moz (468,730 kg) [2026-10-08], daily change 0 (+0%), prior value on 2026-10-07 also 15.07 Moz — unchanged for two consecutive sessions, with no outflow pressure on registered stocks. SHFE GOLD warrants at 116,028 kg [2026-10-09], daily change 0 (+0%), prior value on 2026-10-08 also 116,028 kg — China's deliverable inventory likewise flat. The synchronized flattening of both locations indicates the current price rise is driven by positioning and expectations rather than a physical squeeze, which lowers the probability of a short-squeeze spike but also means that once open interest recedes, price lacks a physical buffer.
ETF: GLD holdings at 33,858,916 oz (2026-10-09) versus 33,778,539 oz 20 trading days ago, a change of +0.24% — a modest net inflow. ETF accumulation is directionally aligned with the futures open-interest build, i.e. allocation and trading flows moving together, which is a second layer of support for the bullish conclusion; however, the +0.24% magnitude is limited and insufficient on its own to constitute a trend call.
Term structure: international market in CONTANGO, M1-M2 at −19.2 (−0.45%), RY −5.44%, slope 17.677. The positive structure of near-month discount and deferred-month premium reflects no near-term physical tightness, corroborating the zero change in inventories at both locations; for longs, a positive structure means a positive roll cost, so trend trades must rely on price appreciation to cover roll decay.
Macro transmission: DXY at 102.28 [2026-10-11] and the 10Y UST yield at 5.24% [2026-10-09] — the dollar and rates are both elevated, which in theory pressures gold, yet GC=F [10-09 settle] still recorded +1.43%, indicating that current pricing weight falls more on safe-haven and allocation demand than on real rates. In the next 7-day calendar: 10-14 US CPI y/y expected 3.6% (prior 3.4%), core CPI y/y expected 2.5% (prior 2.4%); 10-15 retail sales m/m expected 0.3% (prior 1.2%); 10-16 Fed Chair Warsh speaks — all high-weight events transmitting through the dollar into SHFE Gold; if inflation data surprises higher without a synchronized dollar strengthening, gold's sensitivity to real rates will decline, reinforcing the bullish logic.
Seasonality: over the past 15 years, the same-date 20-trading-day average return is +0.65%, median +0.91%, with 8 up years out of 15. The sample is limited; this is background reference only and does not constitute a basis for the conclusion.
IV. Positioning Structure & Fund Flows
China seats: SHFE.AU top 20 seats [2026-10-09] longs 59,096 lots (+1,487 lots), shorts 160,373 lots (+2,962 lots), net direction is net short 101,277 lots (net position daily change -1,475 lots, i.e. net short scale expanded). Longs and shorts built in tandem, with shorts adding more in absolute terms (+2,962 lots vs +1,487 lots), indicating rising two-way divergence rather than one-sided consensus; the net short expansion stems from active short positioning rather than long retreat.
Main force capital: AU2612 main contract open interest 228,458 lots (+1,440 lots), main force notional open interest CNY 207.59bn, daily change +CNY 1.308bn, paired with [10-09 settlement] CNY 899.1/g, daily gain 1.55%; the volume-price relationship is open interest build with rising price, characterized as active long positioning rather than short covering. Trigger conditions show main force open interest at the 95th percentile over the past year, with elevated positioning crowding — the variable most warranting vigilance in the current structure: high open interest means that as incremental capital marginal weakens, price sensitivity to open interest cuts rises.
International COT: managed money net position has declined for four consecutive weeks — 2026-09-15 net 133,116 (Δ-1,856), 09-22 net 127,389 (Δ-5,727), 09-29 net 120,318 (Δ-7,071), 10-06 net 110,987 (Δ-9,331) — a cumulative reduction of 22,129 lots over four weeks, with the pace of cuts widening week by week. Over the same window, GC=F [10-09 settlement] 4216.3, +1.3% over 5 days: price rose while managed money net longs kept falling, a persistent divergence between price and positioning — a substantive divergence signaling that trend-following capital's appetite for chasing highs is fading.
Crowding: 2026-10-06 netPct 28.02%, crowd 90.8, easing from 29.6%/92.51 on 09-29 and 32.48%/92.58 on 09-15; CTA reading flat at 62 for four consecutive weeks, hedge 14.45%. Crowding is gradually releasing from elevated levels — directionally supportive — but the absolute percentile remains above 90, still inside the crowded zone.
Volatility: ^GVZ 22.79 [2026-10-09], 1D +0.11 pts, 1-year percentile 12%; GC=F RV20 17.9%, IV−RV +4.9 vol pts, IV/RV 1.27. Implied volatility exceeds realized volatility — options are paying for event risk — yet GVZ sits at a low 1-year percentile, indicating the market has not overpriced upside; this complements the COT crowding decline.
V. Cross-Asset Relative Value & International vs China Markets
Gold/silver ratio: Gold/Silver 69.07 [2026-10-09], 1-year percentile 74.6%, 3-year percentile 24.87%. On the 1-year dimension it sits in the upper range, meaning gold has outperformed silver over the past year; on the 3-year dimension it sits in the lower range, meaning silver has been stronger than gold over the medium term. The two readings point in opposite directions, so the gold/silver ratio is not used as a directional basis for this report — we merely record the current state: short-term gold advantage, medium-term silver advantage.
Copper/gold ratio: Copper/Gold 1.59 [2026-10-09], 1-year percentile 98.02%, 3-year percentile 58.33%. The 1-year percentile is near an extreme high, reflecting that growth expectations have been priced strongly relative to gold over the past year; the 3-year percentile is neutral, so the ratio's medium-term position is not extreme. A rising copper/gold ratio corresponds to improving growth expectations and exerts marginal pressure on gold's safe-haven premium — a counter-variable that the bullish conclusion must keep tracking.
Oil/gold ratio: Oil/Gold 0.0218 [2026-10-09], 1-year percentile 89.68%, 3-year percentile 47.09%. The 1-year percentile is elevated, the 3-year percentile neutral: crude has been relatively strong versus gold over the past year and flat over the medium term. The Crack 3:2:1 spread stands at USD 66.75/bbl [2026-10-09], with refinery margins sustained — energy-side inflation pressure has not abated, providing indirect support to gold's inflation-hedge properties.
Macro relative-value backdrop: DXY 102.21, 10Y 5.24%, VIX 14.84, GVZ 22.79, OVX 48.9 [2026-10-09]. VIX sits at the 8th 1-year percentile and GVZ at the 12th 1-year percentile — implied volatility for risk assets and gold is low in tandem, the market is in a low-volatility regime overall, and gold's safe-haven premium is not fully priced.
International vs China markets: SHFE.AU (SHFE Gold) trades at a +0.16% premium to GC=F on a tax-adjusted basis [2026-10-09], versus +0.03% five days earlier — the premium has widened modestly. This ratio is comparable in nature; a positive value means China is relatively expensive and the import window tends to open. The current +0.16% premium is mild, with no extreme deviation, so it does not support a directional conclusion of “post-holiday catch-up decline” or “China expensive” — it serves only as a flag for onshore-offshore spread convergence risk.
Risk metrics: DD52w 23.87%, DD20d 6.42%, Sharpe30 -4.1603, VaR95 -2.94%; the 20-day drawdown of 6.42% and VaR95 of -2.94% both point to lingering downside volatility — position management should be weighted alongside the bullish bias.
VI. Historical Statistics & Seasonality
Seasonal statistics (same date, next 20 trading days, past 15 years): average +0.65%, median +0.91%, up in 8/15 years, win rate 53.3%. In sample structure, the best of the 8 up years is 2016 at +6.71%, and the worst of the 7 down years is 2015 at -5.71% — an asymmetric extreme distribution: the upside tail is fatter than the downside tail, and the median (+0.91%) exceeds the mean (+0.65%), meaning a few large down years dragged the mean lower, with typical-year gains concentrated below 1%.
On win rate, 53.3% is only marginally above random and does not constitute a strong seasonal signal; but the median is positive and above the mean, and with AU2612 [10-09 settlement] at CNY 899.1/g sitting at the low 21.5th percentile of the 20-day range 889.82–977.26, the historical “low position + positive skew” combination offers weak background support for the bullish bias. Note the sample is only 15 years and does not strip out macro regime differences (e.g. the 2015 strong-dollar cycle), so this section's conclusions are background only and do not enter the directional basis.
Relation to Section I's bullish conclusion: seasonality provides no independent entry reason — it only shows that the historical distribution over the next 20 trading days is not a headwind for longs. The real directional weight still rests on Section II's volume-price structure and Section IV's positioning crowding.
VII. Scenario Analysis
Base case (probability 55%): bullish bias, range-bound grind higher. Trigger: AU2612 daily close holds above pivot P 903.23, with main contract open interest showing no consecutive net reduction; internationally, GC=F holds near [10-09 settlement] 421.3. Target levels: first target R1 916.63; on a valid break and hold, a 1-ATR (CNY 15.51) measured move points to 932.14. Corresponding action: execute per the strategy already defined in Section VIII — entry 908.64, stop-loss 893.13, target 931.91, horizon 1-5d; if price first retraces to near 903.23, activate the second-tier entry 903.23, stop-loss 887.72, target 926.5, horizon 1-2w. The base-case direction is consistent with the bullish conclusion in Section I; the core rationale is open-interest-build gains and resonance between international and China markets, as set out in Sections I and II.
Bullish case (probability 25%): accelerated upside. Trigger: 10-14 US core CPI y/y actual prints below the 2.5% expectation (i.e., no upside surprise), DXY fails to hold above 103, while AU2612 daily close breaks above R1 916.63 with main contract open interest continuing to increase. Target levels: starting from R1 916.63, a 1-ATR (CNY 15.51) measured move points to 932.14; if 932.14 is validly broken, the next reference is the 20-day range top at 977.26. Corresponding action: do not chase above 916.63; wait for a retest of 916.63 to confirm support before entering, with stop-loss placed below 903.23 (i.e., the 887.72 tier), target 932.14, position size at the reference upper limit per Section VIII.
Bearish case (probability 20%): invalidation downside. Trigger: AU2612 daily close breaks below 895.23 (pivot S1) with main contract open interest turning to net reduction, or 10-14 US core CPI y/y surprises above 2.5% and pushes DXY above 103. Target levels: after a break of 895.23, the next reference is the 20-day low at 889.82; if 889.82 fails, a 1-ATR (CNY 15.51) measured move from S1 895.23 points to 879.72. Corresponding action: once the invalidation condition triggers, do not maintain long exposure; exit entered positions per the Section VIII stop-loss terms (893.13 or 887.72), do not reverse into shorts, and wait for price to re-establish above 903.23 before reassessing re-entry.
The three scenario probabilities sum to 100%. The base case carries the highest weight, as the volume-price structure in Section II and the flat inventories in Section III jointly support a “non-squeeze grind higher”; the 20% bearish probability mainly corresponds to the crowding risk noted in Section IV, with main contract open interest at the 95th percentile of the past year.
VIII. Trading Strategies & Risk Management
The finalized strategies are quoted verbatim below:
- Long entry 908.64 / stop-loss 893.13 / target 931.91 / horizon 1-5d / risk-reward 1:1.5
- Long entry 903.23 / stop-loss 887.72 / target 926.5 / horizon 1-2w / risk-reward 1:1.5
Stop-loss placement: the first-tier stop at 893.13 sits about CNY 2.1 below pivot S1 895.23, leaving a buffer after a daily close below S1 to avoid being shaken out by intraday false breaks; the second-tier stop at 887.72 sits about CNY 2.1 below the 20-day low of 889.82, corresponding to a loss of the range floor. Both stops are directionally consistent with the Section I invalidation condition (daily close below 895.23 with main contract open interest turning to net reduction) — the stop prices sit below the invalidation levels, ensuring positions are not stopped out by noise before invalidation is confirmed.
Position sizing and add/trim rules: on a risk-conversion basis, for the bullish 908.64 entry the stop distance is 15.51, equal to 1×ATR (ATR14 15.506); contract multiplier 1,000; risk per lot CNY 15,560 (including fees and slippage); a single-trade risk budget of CNY 66,924 corresponds to a reference upper limit of 4 lots. For the bullish 903.23 entry the stop distance is 15.51, equal to 1×ATR (ATR14 15.506); contract multiplier 1,000; risk per lot CNY 15,560 (including fees and slippage); a single-trade risk budget of CNY 66,924 corresponds to a reference upper limit of 4 lots. Combined exposure across both tiers must not exceed twice the single-trade risk budget. Add rule: the second-tier quota may be merged in only if price holds above R1 916.63 and main contract open interest continues to increase; trim rule: halve the position on an AU2612 daily close below 895.23, and liquidate entirely if main contract open interest turns to net reduction.
Horizon management: the first tier's 1-5d corresponds to the event window (10-14 CPI, 10-15 retail sales); the second tier's 1-2w corresponds to the trend window. If neither the target nor the stop is hit within the horizon, exit on a time stop and do not roll.
IX. This Week's Data Calendar
Beijing time; only high-weight events relevant to SHFE Gold via USD transmission are listed:
- 10-14 09:30 China CPI y/y, expectation 1.0%, prior 0.8%, threshold ±0.2%; PPI y/y, expectation 4.5%, prior 3.8%, threshold ±0.7%.
- 10-14 20:30 US core CPI m/m, expectation 0.2%, prior 0.3%, threshold ±0.1%; core CPI y/y, expectation 2.5%, prior 2.4%, threshold ±0.1%; CPI m/m, expectation 0.6%, prior 0.4%, threshold ±0.2%; CPI y/y, expectation 3.6%, prior 3.4%, threshold ±0.2%.
- 10-15 20:30 US core PPI m/m, expectation 0.3%, prior 0.2%, threshold ±0.1%; core retail sales m/m, expectation 0.5%, prior 1.4%, threshold ±0.9%; PPI m/m, expectation 0.5%, prior 0.4%, threshold ±0.1%; retail sales m/m, expectation 0.3%, prior 1.2%, threshold ±0.9%.
- 10-16 11:30 Fed Chair Warsh speaks.
Of these, the 10-14 core CPI y/y print is the direct trigger for the Section I invalidation condition and must be tracked closely.
This report is generated automatically from public quantitative and macro data for research and market tracking only. It does not constitute any investment advice.