Silver (SI=F) Deep Dive | 2026-10-07 — Base-building above 60, LONG into 63.1 | 2026-10-07

Call: LONG silver, invalidation on a settle below 59.82 (S2). The prior session settled at 61.59 (+0.47%), holding the lower third of a 20-day range (59.99–68.98, position 17.8%) after a -8.08% 20-day drawdown, with the last completed weekly bar (28 Sep–2 Oct) closing at 60.42 (-6.77% w/w) — the current week is unfinished and up 1.94% through two sessions. Three supports for the long: (1) price is coiling directly on the 59.99 range floor with five consecutive settles between 60.42 and 61.59, and the 61.46 pivot now acts as a shelf rather than resistance; (2) managed-money net length has been cut to 7,614 contracts (netPct 7.11%, crowding 38.9, down from 14,386 on 8 Sep), so the froth that drove the September slide has been flushed while CTA trend proxy holds at +62; (3) silver implied vol at 37.19 sits only 3.4 vol points over RV20 of 33.8%, so optionality is not expensive for a directional expression. The bear case is a macro-driven break of 59.82 toward 58.5; the bull case is a rotation through 63.1 toward 65.5.

1. Bottom Line & Directional Bias

Call: LONG silver (SI=F), invalidation on a daily settle below 59.82 (S2). Three reasons underpin the view. First, the tape has stopped going down: after a -8.08% 20-day move, the last five settled bars (30 Sep–6 Oct) printed 60.57, 61.18, 60.42, 61.3 and 61.59, a tight 1.17-point coil sitting on the 59.99 20-day low, with the 61.46 pivot now beneath price rather than above it. Second, positioning has de-risked without breaking: managed-money net length fell to 7,614 contracts on 29 Sep from 14,386 on 8 Sep, a 47% reduction, while the CTA trend proxy held at +62 and crowding eased to the 39th percentile — this is a cleaned-up book, not a capitulation. Third, the options market is not charging a premium for the trade: VXSLV at 37.19 versus RV20 of 33.8% is an IV/RV of 1.1, and GVZ at the 14th percentile of its 1-year range suggests the whole precious complex is priced for calm. The invalidation is clean: a settle below 59.82 breaks the coil floor and reopens the 58.5 area, at which point the base-building thesis is void.

2. Price Action & Technical Analysis

Silver settled at 61.59 on 6 October, +0.47% on the day, +0.71% over five sessions and -8.08% over twenty. The 20-day channel runs 59.99–68.98, leaving price in the bottom 17.8% of that range — a deeply oversold position on a 20-day lookback, but one that has now held for five consecutive sessions without a new low. The 52-week range is 45.85–121.3, so the current level is roughly mid-range on a yearly view and far from the extremes that generated the 51.4% 52-week drawdown.

ATR14 is 2.21, or 3.59% of price as a full daily range — this is a wide tape, and it matters for stop placement. RV20 is 33.8% annualized. The pivot structure from the settle-based snapshot: P 61.46, R1 62.34, R2 63.1, S1 60.7, S2 59.82. Price at 61.59 is 0.13 above the pivot, which is the first constructive tell; the 5-day sequence has repeatedly tested 62.11–62.46 (30 Sep high 62.11, 2 Oct high 62.46, 5 Oct high 62.4) and failed, so 62.34–62.46 is the immediate supply shelf. A settle through R1 opens R2 at 63.1, which is also the top of the recent consolidation.

In early Asian trade on 7 October (08:00), silver last printed 61.67, +0.14% versus the settle, with a session range of 61.61–61.82 — a narrow, quiet Asian book, labelled as such and not to be confused with the settled move. The last completed weekly bar (28 Sep–2 Oct) opened 64.66, high 64.66, low 59.99, closed 60.42, -6.77% w/w — a bearish weekly candle that marked the low of the move. The current week (from 5 Oct, two sessions) is not closed and shows +1.94%; no weekly-close conclusion can be drawn from it.

View: the structure is a base, not a breakdown. Holding 60.7 (S1) keeps the coil intact; reclaiming 62.34 (R1) on a settle is the trigger for the 63.1 test.

3. Supply-Demand Balance & Fundamental Drivers

COMEX registered silver stood at 101.68 Moz (3,162,604 kg) as of 2 October, up 0.58 Moz (+0.57%) from 101.1 Moz on 1 October, and up from 101.38 Moz on 30 September. This is a modest but persistent rebuild of registered stocks over the three observations available — the kind of flow that caps aggressive upside but is far too small to signal a glut. SHFE warrant stocks were 1,490,865 kg as of 30 September, up 30,504 kg (+2.09%) day-on-day. Both exchanges are adding metal at the margin, which argues against a squeeze narrative and supports the base-case grind rather than a vertical move.

The term structure is in contango: M1–M2 at -0.26 (-0.42%), roll yield -5.09%, slope 0.1239. Contango here reflects carry — the cost of financing and storing metal against a 5.27% ten-year yield — and is not a price cap; it is a roll cost for anyone holding long futures. The practical read: longs pay roughly 5% annualized to roll, which is a real drag and argues for expressing the view over days-to-weeks rather than quarters, or via options where the carry is embedded in the premium.

Macro transmission is through the dollar and rates: DXY at 101.85 (-0.32%) and ^TNX at 5.27% (-0.79%) on 6 October. A softer dollar and lower long-end yield are the two channels that matter most for silver, and both moved in the metal's favour into the settle. The single headline in the window — a Discovery silver project clearing a key Mexico hurdle — is a multi-year supply signal, not a near-term flow, and does not change the 20-session balance.

View: inventories are a mild headwind, carry is a real cost, but the dollar/rates channel is turning supportive. Net, fundamentals are neutral-to-constructive and do not block the long.

4. Positioning & Fund Flows

Managed-money positioning has been cut hard. Net length fell to 7,614 contracts on 29 September from 13,309 on 22 September — a weekly change of -5,695 contracts, the largest reduction in the four-week window. Open interest was 107,047, up slightly from 106,474, so the reduction was long liquidation rather than a short build. Net length as a share of open interest is 7.11%, with crowding at the 39th percentile of the three-year distribution — down from 58.9 on 8 September. This is not a crowded long; it is a market that has already absorbed selling.

The CTA trend-following proxy reads +62, unchanged across all four weeks, and hedge pressure is 24.83%, also broadly stable. The stability of the trend proxy while net length collapsed is the key divergence-free observation: price fell and length fell together, which is de-risking, not a positioning/price divergence. There is no evidence of a crowded short either — shorts are only 9,272 contracts against 16,886 longs.

On volatility, VXSLV at 37.19 rose 0.59 points on 6 October while RV20 sits at 33.8%, leaving IV minus RV at +3.4 vol points (IV/RV 1.1). Options are paying a modest premium to realized, which is normal around an FOMC minutes release but not stretched. For context, GVZ at 22.97 is in the 14th percentile of its one-year range and VIX at 15.01 is in the 12th percentile — the broader complex is priced for calm, which makes long optionality relatively cheap.

View: the positioning reset is the single most constructive input. With crowding at 38.9 and net length halved, the path of least resistance is higher.

5. Cross-Asset Relative Value

The gold/silver ratio is 67.81, with a 1-year percentile of 64.29% and a 3-year percentile of 21.43%. The two percentiles tell different stories and both matter. On a one-year view the ratio is in the upper two-thirds — silver has lagged gold over the past twelve months, consistent with the -8.08% 20-day silver drawdown against a more resilient gold tape. On a three-year view the ratio sits in the bottom quartile — silver has been structurally strong against gold over the longer horizon. The current state is therefore a ratio that has backed up within a longer-term silver-favourable trend; it is not an extreme, and a low three-year percentile does not by itself forecast a reversal.

There is no copper/gold or crack spread data in the window, so no pro-growth read is available from those channels. The relevant cross-asset signal is the dollar: DXY at 101.85, down 0.32%, with the ten-year at 5.27%, down 0.79%. A weaker dollar mechanically lifts the ratio-adjusted value of silver, and the two moved together into the 6 October settle.

View: the ratio at 67.81 is mid-range on a one-year basis and low on a three-year basis — a neutral-to-supportive backdrop for silver versus gold, with the dollar the swing input.

6. Historical & Seasonal Patterns

The seasonality block covers the same calendar start, next 20 sessions, over the last 15 years: mean +1.05%, median +2.32%, up in 10 of 15 years. The best instance was 2011 at +10.05%; the worst was 2014 at -7.35%. The distribution is positively skewed — the median exceeds the mean, and the best year is far larger than the worst — which is consistent with silver's tendency to produce sharp upside tails in this window.

The sample is small (15 observations) and should be treated as context, not edge. The hit rate of 10/15 (67%) is the more robust statistic than the magnitude, and it aligns directionally with the long call. The 2014 worst case of -7.35% is a useful reminder of the downside tail: a repeat would take silver from 61.59 to roughly 57.06, which is below the 59.82 invalidation — so the seasonal tail risk and the technical invalidation are broadly consistent.

View: seasonality is a mild tailwind with a fat right tail; it supports the long but is not the reason for it.

7. Scenario Analysis (Base / Bull / Bear)

Base case — 50% — grind higher within the coil. Trigger: price holds above S1 at 60.7 and settles through the 61.46 pivot, with the Asian session (61.67, +0.14%) confirming no overnight supply. Target: 62.34 (R1), then 63.1 (R2). Action: hold the long, add on a settle above 62.34, trail stops beneath 60.7. This path agrees with the section 1 call and requires no new catalyst — just the absence of a dollar spike.

Bull case — 30% — breakout extension. Trigger: a settle above 63.1 (R2) on rising open interest, most plausibly catalysed by a dovish read of the FOMC minutes (BJT 10-08 02:00) that pushes DXY below 101.5 and the ten-year below 5.2%. Target: 65.5, the midpoint of the September breakdown. Action: add to the long, move the stop to breakeven-plus, and consider taking partial profit into 65.5 given the contango roll cost of -5.09% annualized.

Bear case — 20% — coil failure. Trigger: a daily settle below 59.82 (S2), which would break the five-session floor and confirm the 20-day downtrend is resuming. Target: 58.5, with the 52-week low at 45.85 far below but not in play on this horizon. Action: exit the long on the settle, stand aside, and re-engage only on a reclaim of 60.7. The bear case is the invalidation scenario and is deliberately the lowest-probability branch given the positioning reset and the 17.8% range position.

Probabilities sum to 100%. The base case is the call; the bull case is the upside path; the bear case is the stop-out.

8. Trading Strategies & Risk Management

Strategy 1 — Core long (1–5 day horizon, conviction 7/10). Entry at market, 61.59–61.7, referencing the 6 October settle of 61.59 and the Asian print of 61.67. Stop at 59.75, just beyond S2 at 59.82 and roughly 0.8 ATR below entry, so it sits outside the normal daily range rather than inside it. Target 63.1 (R2), a 1.51-point move against 1.84 points of risk — a modest reward-to-risk that improves materially on a settle above 62.34. Size at half normal, given ATR14 of 2.21 (3.59% of price) means a full-size position carries a wide daily P&L swing.

Strategy 2 — Breakout add (1–2 week horizon, conviction 6/10). Add on a daily settle above 62.34 (R1), stop at 60.65 beneath S1 at 60.7, target 65.5. This leg only activates if the base resolves upward; it is not a standalone entry. Horizon is capped at two weeks because the contango roll yield of -5.09% annualized erodes a held long futures position.

Risk management: total exposure across both legs should not exceed 1.5x normal unit size. The FOMC minutes at BJT 10-08 02:00 is the event most likely to gap the position; consider reducing size into that print or expressing part of the position via options, where VXSLV at 37.19 versus RV20 of 33.8% offers IV/RV of 1.1. A settle below 59.82 voids both legs.

9. This Week's Data Calendar

BJT 10-07 22:30 | ET 10-07 10:30 — EIA Crude Oil Stocks Change (OCT/02), medium impact, affects CL/BZ. BJT 10-08 02:00 | ET 10-07 14:00 — FOMC Meeting Minutes, high impact, affects GC, SI, DXY; the key event for silver this week. BJT 10-08 16:30 | ET 10-08 04:30 — FOMC Member Waller speaks, medium impact, GC/SI/DXY. BJT 10-14 09:30 | ET 10-13 21:30 — China CPI and PPI y/y, high impact, affects HG, CL, ZS.

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.

單品種深度報告深度分析

COMEX黃金GC=F 2026-10-07

Gold (GC=F) Deep Dive | 2026-10-07 — Fading Crowded Longs Into Support, Tactical LONG | 2026-10-07

Silver (SI=F) Deep Dive | 2026-10-07 — Base-building above 60, LONG into 63.1 | 2026-10-07

今日風險地圖

單品種深度報告 深度分析

往期研報封存

⟠ 本月策略回測表現

單品種深度報告深度分析