I. Bottom Line & Directional Bias
The core rationale rests on three points. First, at the [10-16 settlement], SHFE Gold 2612 was quoted at CNY 898.8/g, with the day session closing at CNY 901.82/g, up +1.36% versus the prior settlement, while main contract open interest stood at 222,955 lots, down 621 lots on the day, and main contract notional open interest at CNY 201.07bn, down CNY 560m on the day; the price rise accompanied by falling open interest is classified as short covering, indicating that near-term selling pressure has been released but longs have not yet begun to add positions actively. Second, among the top 20 seats on [10-16], long positions totaled 230,505 lots (+3,709) and short positions 61,517 lots (-566), for a net long of 168,988 lots, expanding by 9,557 lots versus the prior day; the seat structure remains skewed to the long side. Third, in the international market, GC=F settled at USD 4,181.1/oz on [10-16 settlement], up +0.26%, and the China ex-tax implied premium stood at +0.35% [10-16], showing that international and Chinese markets are directionally aligned. Risk point: the 20-day change is -5.78% and price sits at the 27.1st percentile of the 20-day channel, so the rebound has not yet reversed the medium-term downtrend; the night session closed at CNY 898.62/g, down -0.35% versus the day session, confirming that resistance near R1 901.42 remains effective. Directional invalidation conditions: a daily close below 883.8 (S1) would open the downside toward 869.0; alternatively, a firm hold above 901.42 (R1) would target 914.0.
II. Price Action & Technicals
Technicals are benchmarked to the [10-16 settlement] of CNY 898.8/g. The day session closed at CNY 901.82/g, up +1.36% versus the prior settlement; the 5-day change is -0.75% and the 20-day change is -5.78%, with price at the 27.1st percentile of the 20-day channel of 881.82–955.52; the 20-day high/low (excluding the current day) is 881.82–959.6. ATR14 is CNY 13.56/g (1.5%, full intraday range) and RV20 is 18.4%, indicating mild realized volatility. Classic pivots are derived from [10-15] H/L/C, corresponding to the [10-16] trading day: P 891.62, R1 901.42, S1 883.8. The settlement price of 898.8 sits between P and R1, roughly CNY 2.62 below R1 and less than half an ATR (6.78), so R1 is not yet treated as a valid trading level and serves only as a resistance reference; S1 883.8 lies about CNY 15 below the settlement, more than one ATR, and constitutes valid downside support. The last five daily candles (C = close): 10-12 closed 905.52, 10-13 closed 907.02, 10-14 closed 896.64, 10-15 closed 893.6, 10-16 closed 901.82, forming a “two up — two down — one up” repair pattern; the 10-15 low of 881.82 marks the lower bound of the 20-day channel, and the 10-16 rebound failed to break the 10-13 high of 907.02. 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%; the current week [from 10-12] has not yet closed and is tentatively reported at 901.82 (-0.75%), so no weekly conclusion is drawn. China night session [10-16 21:05 to 10-17 02:30]: open 899.2, high 906.2, low 897.06, close 898.62, down -0.35% versus the day session close and -0.02% versus settlement; the night session high of 906.2 approached the 10-13 high of 907.02, confirming the near-term resistance zone. Trading implication: bias is bullish but requires confirmation; a daily close firmly above 901.42 (R1) would target 914.0, while a break below 883.8 (S1) would open the downside toward 869.0.
III. Supply-Demand Balance & Fundamental Drivers
The supply-demand picture is driven mainly by inventories and open interest changes. SHFE Gold warrants on [10-16] stood at 116,928 kg, up 900 kg (+0.78%) on the day, versus the prior value of 116,028 kg on [10-15] with a daily change of 0; warrants have risen for two consecutive days and turned to an increase on [10-16], exerting mild physical pressure on the market, though the absolute level remains in a low range. COMEX Gold registered inventories on [10-14] were 14.82 Moz (460,954 kg), down 0.06 Moz (-0.4%) on the day; international registered inventories saw a small drawdown, moving in the opposite direction to the rise in Chinese warrants, so onshore and offshore inventory signals are inconsistent and warrant continued monitoring. GLD holdings on [10-15] were 33,884,026 oz, versus 33,767,190 oz 20 trading days earlier, a change of +0.35%; the world's largest gold ETF has seen a modest allocation increase over the past 20 days, providing mild support to gold prices. On macro transmission, the [10-16] US 10-year Treasury yield was 5.1%, up 0.18% on the day, while the US Dollar Index was 101.78, down 0.22% on the day; real rates and the dollar moved in opposite directions, leaving the headwind and tailwind for gold roughly offset. VIX rose to 19.99 on [10-16], up 3.53 points on the day and at the 82nd percentile over one year, with rising risk aversion a positive for gold. The supply-demand balance sheet (WASDE, ICSG, etc.) is not connected, so no balance judgment is made. Fundamental drivers summary: modest ETF allocation increases and the drawdown in international registered inventories are bullish, the rebound in Chinese warrants is bearish, the macro backdrop is neutral, and overall drivers are mildly positive.
IV. Positioning & Fund Flows
Positioning is bullish but requires distinguishing between seats and the main contract. Top 20 seats on [10-16]: long positions 230,505 lots (+3,709), short positions 61,517 lots (-566), net long 168,988 lots (expanding by 9,557 lots versus the prior day), with a long/short ratio of about 3.75:1, indicating a clearly bullish seat structure. Main contract open interest on [10-16] was 222,955 lots, down 621 lots on the day, and main contract notional open interest was CNY 201.07bn, down CNY 560m on the day; the +1.36% price rise accompanied by falling open interest is classified as a rally on open interest cut, i.e., short covering, rather than active long accumulation; this coexists with the expansion in seat net longs, suggesting that longs are still entering at the seat level while the main contract side is dominated by short liquidation, so the sustainability of the rebound depends on whether it shifts to a rally on open interest build. International comparison: COT [10-13] OI = 396,345, longs 124,747, shorts 17,658, net long 107,089, weekly change -3,898; net longs over the past four weeks have fallen from 127,389 on [09-22] to 107,089 on [10-13], with managed money net positions continuing to decline, consistent with the -5.51% 20-day price decline and showing no clear divergence. COT crowding (as-of 2026-10-06, 11 days ago, data not updated): netPct 28.02%, crowding 90.8 (3-year percentile), CTA 62, hedge 14.45%; crowding remains elevated but has retreated from 92.58 on [09-15]. Implied versus realized volatility: ^GVZ on [10-15] was 22.34, down 0.44 points on the day and at the 10th percentile over one year; GC=F RV20 is 18.3%, IV−RV is +4.0 vol points and IV/RV is 1.22, so options are paying some premium for event risk, but with GVZ at a one-year low, the market's pricing of gold volatility is not aggressive.
V. Cross-Asset Relative Value & International vs China Markets
Cross-asset relative value is based on the [2026-10-16] snapshot. International vs China markets: SHFE.AU vs GC=F, with the China ex-tax implied premium at +0.35% [10-16] versus +0.01% five days earlier; the Chinese market is relatively expensive versus the international market but the premium is mild, and onshore/offshore directions are aligned ([10-16] China +1.36%, international +0.26%), with no basis inversion issue. Ratios and percentiles: the gold/silver ratio is 72.96 [10-16], at the 87.7th percentile over 1Y and the 29.37th percentile over 3Y, i.e., high over one year but low over three years, indicating silver is relatively weak versus gold and gold is relatively expensive; the copper/gold ratio is 1.57 [10-16], at the 92.46th percentile over 1Y and the 54.23rd percentile over 3Y, a one-year high, with the rising copper/gold ratio reflecting strong growth expectations and diverting some of gold's safe-haven appeal; the oil/gold ratio is 0.0243 [10-16], at the 100th percentile over 1Y and the 55.29th percentile over 3Y, an extreme one-year high, with crude oil relatively strong versus gold. Additional cross-asset snapshot: DXY 101.78, 10Y 5.1%, VIX 19.99 [10-16]; the 3:2:1 crack spread is USD 57.77/bbl [10-16] (no direct transmission to gold, background only). Related spreads (international comparison): HG_GC_RATIO 1.57, P1Y 88.49%, P3Y 51.46%; GC_SI_RATIO 72.96, P1Y 84.92%, P3Y 28.44%; CL_GC_RATIO 0.0235, P1Y 98.41%, P3Y 53.44%. Term structure (international comparison): CONTANGO, with M1-M2 at -18.9 (-0.45%), RY -5.41%, slope 17.601; the international market shows near-month discount and a contango curve, which is neutral for cost of carry. Summary: both the gold/silver and copper/gold ratios are at one-year highs, with gold relatively expensive and growth expectations strong, constraining upside for gold prices; however, the international-China premium is mild and directions are aligned, so the bullish bias is maintained.
VI. Historical Statistics & Seasonality
On seasonality, this section cites only the in-house seasonal statistics block (same date over the past 15 years, 20 trading days) as background; the sample is limited and does not constitute a directional basis. The statistics show: average change of -0.8%, median of -0.68%, with up years at 6/15, i.e. a historical same-period win rate of about 40%, slightly below parity; the best year was 2011 at +5.76%, the worst was 2014 at -6.98%, with large year-to-year dispersion and extreme years exceeding ±5%. Combined with this report's modestly bullish bias, seasonality does not provide resonant support: over the past 15 years, both the same-period median and mean are negative, and the win rate is below half, indicating that from a calendar-effect perspective the current window is closer to the historically weak zone than the strong zone. Therefore, the support for the bullish-leaning view in this report comes mainly from positioning and price structure (as set out in Section I and Section II), not from seasonality; seasonality merely flags that if the rebound extends, watch for mean-reversion pressure typical of this period. Operationally, seasonality does not alter the base-case direction, but argues for restraint on targets: the worst historical year's -6.98% far exceeds the intraday move that the current ATR14 (CNY 13.56/g) can cover, so stop-loss discipline takes priority over target space. Note that the sample is only 15 years, and the best/worst years were likely dominated by macro shocks of those years and should not be extrapolated to the current year; this section does not issue an independent trading signal on this basis, but serves only as a background constraint for scenario probability setting.
VII. Scenario Analysis
Scenario analysis follows the key-level framework from the first half: pivot P 891.62, R1 901.42, S1 883.8, ATR14 at CNY 13.56/g, with invalidation conditions consistent with Section I. Base case (probability 55%, modestly bullish): after the day session closed at CNY 901.82/g, +1.36% versus the prior settlement, price consolidates between P 891.62 and R1 901.42, with top-20 seat net longs of 168,988 lots (up 9,557 lots versus the prior day) providing structural support; the trigger is a daily close firmly above 901.42 (R1), with a target of 914.0 (901.42 + 0.9×ATR14 ≈ 913.6, rounded to 914.0), and operations follow the established strategy (see Section VIII), using the night session high of 906.2 as a near-term resistance reference. Bull case (probability 20%): if the daily close holds firmly above 901.42 (R1) and subsequent sessions do not retest S1 883.8, while international GC=F continues higher from the [10-16 settlement] level of USD 4,181.1/oz, look to 914.0, with further upside referenced against the post-breakout space after the 10-13 high of 907.02 across the last five daily bars was cleared; operationally, scale out near the established strategy target of 921.82 and do not add new positions. Bear case (probability 25%): the trigger is a daily close below 883.8 (S1), looking down to 869.0 (883.8 − 1×ATR14 ≈ 870.2, rounded to 869.0); if the night session low of 897.06 is also lost, near-term weakness is confirmed, and the operation is to close the established long and stand aside, reassessing only after price re-establishes firmly above 891.62 (P). The three scenario probabilities sum to 100%. Emphasize: both the base and bull cases require a firm hold above R1 901.42 to be valid; if price only touches it intraday without a firm daily close above, treat it as the base case. Once the bear case is triggered, 869.0 is the downside target, with no lower level set.
VIII. Trading Strategies & Risk Management
Execute the trading strategy per the established plan: go long, entry 901.82, stop-loss 888.06, target 921.82, horizon 1–5 days, risk/reward 1:1.5. The stop at 888.06 is derived from the risk conversion: a stop distance of CNY 13.76/g, i.e. 1.01×ATR (ATR14 is 13.56), slightly more than one ATR, covering normal intraday volatility without being randomly stopped out; at the same time, this level sits between pivot P 891.62 and S1 883.8, so if price breaks below 888.06 it signals that P 891.62 support has failed and the short-covering logic is broken, hence the stop is placed here rather than below S1 883.8. On sizing, the contract multiplier is 1,000, with risk per lot of CNY 13,810 (including fees and slippage); a single-trade risk budget of CNY 66,853 corresponds to a reference cap of 4 lots, and actual opening must not exceed that cap. Add/trim rules: after entry, if the daily close holds firmly above 901.42 (R1), hold to the target of 921.82 but do not add; if price hits the stop at 888.06 first, exit immediately, with no averaging down; if neither target nor stop is reached within the horizon, close the position at expiry (1–5 days). This section sets no additional entry, stop, or target numbers; all stop distances, ATR multiples, and lot counts are taken from the risk conversion, not independently calculated. Note: the established strategy entry of 901.82 is the [10-16] day session close; if the night session or subsequent open has moved far from that price, reassess execution per the risk rules and do not chase.
IX. This Week's Data Calendar
This week's data calendar (Beijing time): 10-19 09:30 China new home prices m/m; 10-19 10:00 China energy production, fixed asset investment ytd/y, household income, industrial capacity utilization, industrial value added y/y, real estate development investment ytd/y, total retail sales of consumer goods y/y, services production index y/y; 10-20 09:15 China 1-year Loan Prime Rate (prior 3%), 5-year Loan Prime Rate (prior 3.5%); 10-20 09:30 China GDP y/y. Most events are tagged CNY/HIGH or MEDIUM, mainly affecting ferrous metals, non-ferrous metals, and chemicals; the transmission to gold is primarily via risk appetite and rate expectations, and should be assessed in conjunction with changes in the [10-16] US 10-year Treasury yield of 5.1% and the dollar index at 101.78.
This report is generated automatically from public quantitative and macro data for research and market tracking only. It does not constitute any investment advice.