📅 2026-10-01 · COMEX白银 · 单品种早报旧覆盖策略

数据截止: 2026-09-30 · 结算口径: 交易所最终日结算价

Silver (SI=F) Deep Dive | 2026-10-01 — Bearish: 20-Day Channel Floor at Risk | 2026-10-01

Call: Bearish silver (SI=F), with the view invalidated on a settle back above pivot P 61.02. Silver settled at 60.57 on 2026-09-30, down 0.96% on the day, 6.77% over five sessions and 7.35% over twenty, sitting at the 2nd percentile of a 60.38–68.98 twenty-day channel — a channel floor that has not yet been breached on a closing basis. The last completed weekly bar (2026-09-21/25) closed at 64.8, down 3.5% w/w, and the unfinished current week is down 6.54% through three sessions. Registered COMEX silver rose 2.4% d/d to 101.56 Moz and SHFE warrants added 2.1% to 1,490,865 kg, so visible metal is building into a falling tape. Managed-money net length sits at the 55th percentile of its three-year range with CTA proxies pinned at +62, leaving room for further trend-following liquidation. The gold/silver ratio at 69.13 is in the 24th percentile of its three-year range, meaning silver has been structurally strong versus gold, but the one-year percentile of 72.62% shows that strength is now fading. Silver implied vol at 37.87 sits only 2.4 points above RV20 of 35.5%, so downside optionality is not expensive.

1. Bottom Line & Directional Bias

Call: Bearish silver (SI=F). Invalidation: a daily settle back above pivot P at 61.02.

Three reasons underpin the view. First, price structure: silver settled at 60.57 on 2026-09-30, at the 2nd percentile of its 20-day 60.38–68.98 channel, with 5D at -6.77% and 20D at -7.35% — the channel floor is the last line of defence and the momentum is pointed at it. Second, visible supply is loosening: COMEX registered stocks rose 2.4% d/d to 101.56 Moz and SHFE warrants added 2.1% to 1,490,865 kg, both building into a declining price, which is the opposite of a squeeze setup. Third, positioning is not washed out: managed-money net length is at the 55th percentile of its three-year range and the CTA proxy is pinned at +62, so trend-following flow still has room to sell.

The invalidation is clean and arithmetic: a settle above 61.02 (pivot P) would put price back inside the prior balance area and neutralise the channel-floor breakdown thesis. Until then, rallies into 61.02–61.66 are supply to sell, not reversals.

2. Price Action & Technical Analysis

Silver settled at 60.57 on 2026-09-30, down 0.96% (settle) on the day, -6.77% (settle) over five sessions and -7.35% (settle) over twenty. The 20-day channel runs 60.38–68.98, placing the settle at the 2nd percentile — effectively on the floor. The 52-week range is 45.39–121.3, so the market is in the lower third of its annual envelope but far from the extremes.

In early Asian trade on 2026-10-01 (06:55), silver last printed 60.71, +0.24% versus the prior settle (Asia), with a session range of just 60.7–60.8 (Asia). That is a very narrow Asian box and should be read as consolidation, not a reversal signal; the reference level remains the 60.57 settle.

Volatility is elevated and two-sided. ATR14 is 2.32, equal to 3.84% of price — that is the full expected daily range, not a one-sided band. RV20 is 35.5%. The practical implication is that a stop placed less than roughly one ATR from entry sits inside normal daily noise; risk must be sized accordingly.

Pivots from the settle-based snapshot: P 61.02, R1 61.66, R2 62.75, S1 59.93, S2 59.29. The ordering matters. Price is below P, so the immediate bias is to sell rallies into 61.02. A settle above P would flip the tactical bias; a settle below S1 59.93 opens S2 59.29, and a break of the 20-day channel floor at 60.38 on a closing basis would confirm the next leg lower.

On the weekly timeframe, the last completed bar (2026-09-21 to 2026-09-25) opened 66.69, high 68.11, low 63.51 and closed 64.8, -3.5% w/w. That completed bar was already a lower close. The current week (from 2026-09-28, three sessions) is not closed and shows -6.54%; no weekly-close conclusion can be drawn from it. The takeaway from the completed weekly bar is that the prior week's high at 68.11 is now distant resistance, and the market has lost the 64–65 area that had been support.

View: bearish while below 61.02; the 60.38 channel floor is the trigger for the next leg, and 59.29 (S2) is the first objective.

3. Supply-Demand Balance & Fundamental Drivers

Visible inventory is the cleanest fundamental signal in the current tape, and it is bearish. COMEX registered silver rose to 101.56 Moz (3,158,871 kg) on 2026-09-29, up 73,404 kg or +2.4% d/d. Two sessions earlier, on 2026-09-28, registered stocks were 99.2 Moz (3,085,467 kg), so the build is not a one-day blip — it is a two-session accumulation of roughly 2.36 Moz. SHFE warrants rose to 1,490,865 kg on 2026-09-30, up 30,504 kg or +2.1% d/d, following a 21,070 kg build the prior session. Both major deliverable venues are adding metal while price falls. That combination — rising registered stocks into a declining price — removes the scarcity narrative that typically drives silver squeezes and instead points to metal moving from private hands into warehouses.

The macro backdrop transmits to silver through two channels. First, the US 10-year yield at 5.29% (^TNX, +0.72%) and DXY at 101.46 (+0.09%) represent a restrictive real-rate and firm-dollar combination that is historically a headwind for non-yielding precious metals. Second, the SMM commentary noting a “holiday atmosphere strong in silver spot market” and a pullback in the probability of rate hikes suggests spot demand is seasonally quiet into the Chinese holiday, which reduces the physical bid that normally cushions dips.

The term structure is in contango: M1-M2 at -0.23 (-0.38%), roll yield -4.54%, slope 0.2311. Contango here reflects carry — the cost of financing and storing metal against a 5.29% policy-sensitive yield curve — and is not itself a price cap. It is, however, a roll cost for long holders, which modestly penalises buy-and-hold positioning and reinforces the case that speculative length has an incentive to trim.

There is no offsetting supply-side tightness in the data: no mine disruption, no refinery outage, no drawdown in registered stocks. The fundamental picture is one of adequate and building deliverable supply against a softening spot bid.

View: bearish. Rising COMEX and SHFE inventories plus contango carry costs argue against a near-term squeeze; the burden of proof is on the bulls to show a drawdown.

4. Positioning & Fund Flows

CFTC managed-money positioning is the key swing factor and it is not yet washed out. As of 2026-09-22, open interest was 106,474, longs 19,303, shorts 5,994, for a net of 13,309, up 185 w/w. The net-to-OI ratio was 12.5%, with a crowding percentile of 55.01 on the three-year window and a CTA trend-following proxy of +62. The hedge ratio was 24.99%.

The four-week sequence shows net length oscillating in a narrow band: 12,598 (2026-09-01), 14,386 (2026-09-08), 13,124 (2026-09-15), 13,309 (2026-09-22). The crowding percentile has drifted from 53.27 to 55.01, and the CTA proxy has been pinned at +62 throughout. Two observations follow.

First, this is not a crowded long — the 55th percentile is mid-range, not extreme, so there is no mechanical short-squeeze fuel from positioning alone. Second, and more importantly for the bear case, the CTA proxy at +62 means trend-following accounts are still net long a market that has fallen 6.77% over five sessions and 7.35% over twenty. Trend systems typically flip or reduce when price breaks a channel floor. With the 20-day floor at 60.38 and the settle at 60.57, the trigger for systematic selling is close at hand. That is the asymmetry: positioning is not stretched enough to force a squeeze, but it is long enough to feed a liquidation cascade if the floor gives way.

On the options side, ^VXSLV (silver implied vol) is 37.87, down 1.26 points d/d, against RV20 of 35.5%, giving an IV minus RV of +2.4 vol points (IV/RV 1.07). Implied vol is only marginally above realised, so downside optionality is not expensive — a put or put-spread structure to express the bearish view is reasonably priced relative to the realised move the market has already delivered.

View: bearish. Mid-range crowding means no squeeze fuel, while a +62 CTA proxy means systematic flow is a seller if 60.38 breaks.

5. Cross-Asset Relative Value

The primary cross-asset lens for silver is the gold/silver ratio at 69.13. Its one-year percentile is 72.62% and its three-year percentile is 24.21%. The interpretation is important and frequently misread. A low or falling ratio means silver is outperforming gold; a high or rising ratio means silver is lagging. The three-year percentile of 24.21% tells us that, over a three-year horizon, the ratio has spent most of its time higher — i.e. silver has been structurally strong versus gold. But the one-year percentile of 72.62% tells us the ratio is now in the upper portion of its one-year range, meaning silver's relative strength versus gold is fading and gold is currently the better-performing leg.

That is consistent with the volatility picture: ^GVZ (gold implied vol) at 23.74, 1Y percentile 20%, versus ^VXSLV at 37.87. Gold's implied vol is in the bottom quintile of its one-year range while silver's is materially higher, reflecting silver's higher beta and the recent disorderly move. In a risk-off or rate-driven precious-metals bid, gold typically leads and silver lags — which is exactly what a rising gold/silver ratio from a low three-year base would look like.

Elsewhere, ^VIX at 16.34 (1Y percentile 33%) and ^OVX at 52.24 (1Y percentile 52%) indicate no broad risk-off panic. This matters for silver because its industrial demand component means it is sensitive to growth expectations; a calm equity-vol regime with a firm dollar and a 5.29% 10-year yield is a mix that favours gold over silver on the precious side and offers no growth impulse on the industrial side.

View: bearish silver versus gold. The ratio's one-year percentile at 72.62% says silver is the lagging leg; relative-value longs should be expressed in gold, not silver.

6. Historical & Seasonal Patterns

Seasonality for the same calendar window — the next 20 sessions from 1 October — over the last 15 years shows a mean return of +2.37%, a median of +1.31%, and an up-rate of 10 out of 15 years. The best instance was 2011 at +11.67% and the worst was 2012 at -7.43%.

This is a mildly positive seasonal tilt, and intellectual honesty requires flagging it as the main counterweight to the bearish call. However, three caveats apply. First, the sample is small — 15 observations — and the median of +1.31% is modest relative to the current ATR14 of 2.32, meaning the seasonal edge is smaller than a single day's expected range. Second, the distribution is wide: the spread between the best and worst years is over 19 percentage points, so the mean is not a reliable central estimate. Third, seasonality is a context input, not a trigger; it does not override a price sitting at the 2nd percentile of its 20-day channel with inventories building.

The correct use of this block is as a risk flag: it argues against pressing short exposure to maximum size, and it argues for taking profits at the S2 59.29 objective rather than holding for an extended trend. It does not change the directional call.

View: seasonality is a mild headwind to the bear case and a reason to manage size and take profit at targets, not a reason to be long.

7. Scenario Analysis (Base / Bull / Bear)

Base case — 55% probability: grind lower toward S2 59.29. Trigger: price remains below pivot P 61.02 and settles below the 20-day channel floor at 60.38. Target: 59.29 (S2), with 59.93 (S1) as the first waypoint. Action: hold short exposure established below 61.02, trail stops to the 61.02–61.66 zone, and take partial profit at 59.93. This scenario is consistent with the section 1 call: falling price, building COMEX and SHFE inventories, and a CTA proxy at +62 that has room to reduce.

Bull case — 25% probability: reclaim 61.02 and squeeze toward R1 61.66 / R2 62.75. Trigger: a daily settle back above 61.02 (pivot P), ideally accompanied by a drawdown in COMEX registered stocks or a dovish surprise from the week's labour data. Target: 61.66 (R1) initially, then 62.75 (R2). Action: this is the invalidation scenario — cover shorts on a settle above 61.02 and stand aside; do not initiate longs against the primary trend unless 62.75 is reclaimed on a closing basis. The seasonal tailwind (mean +2.37%, up 10 of 15 years) and the mid-range 55th percentile crowding are the supports for this path.

Bear case — 20% probability: accelerated breakdown below 59.29. Trigger: a decisive settle below S2 59.29, likely on a hot inflation print or a hawkish FOMC minutes read that lifts the 10-year yield further above 5.29% and extends dollar strength beyond DXY 101.46. Target: the 52-week low at 45.39 becomes the medium-term reference, with the 20-day drawdown already at 11.77%. Action: add to shorts on a confirmed close below 59.29, with stops back above 60.38; this is the tail scenario where systematic CTA selling compounds the move.

Probabilities sum to 100%. The base case agrees with the section 1 call. The bull case is the invalidation path, not a second conclusion.

8. Trading Strategies & Risk Management

Strategy 1 — Short silver on rallies into pivot resistance (primary). Entry: 61.02 (pivot P), scaled between 60.9 and 61.2. Stop: 62.75 (R2). Target: 59.29 (S2), with a partial at 59.93 (S1). Horizon: 1–5 days. Conviction: 7/10. Size: half of normal risk budget, given the mildly positive seasonal tilt and the mid-range crowding percentile.

Strategy 2 — Momentum short on a confirmed channel-floor break (secondary). Entry: on a daily settle below 60.38 (20-day channel floor). Stop: 61.66 (R1). Target: 59.29 (S2), then trail. Horizon: 1–5 days. Conviction: 6/10. Size: quarter of normal risk budget, added only if Strategy 1 is already working.

Risk management notes: ATR14 of 2.32 (3.84% of price) means position size must be calibrated so that a one-ATR adverse move is survivable; VaR95 of -6.28% and 20-day drawdown of 11.77% confirm this is a high-volatility instrument. Do not add to shorts into the 59.29 target; the seasonal window argues for realising profit there. If price settles above 61.02, both strategies are invalidated and should be closed.

9. This Week's Data Calendar

BJT 10-01 22:00 | ET 10-01 10:00 — FOMC Member Waller Speaks; ISM Manufacturing Employment SEP (F 51.5, P 51.2, surprise outside F±0.3); ISM Manufacturing PMI SEP (F 54.8, P 54.6, surprise outside F±0.2). BJT 10-02 20:30 | ET 10-02 08:30 — Average Hourly Earnings m/m (F 0.3%, P 0.3%, surprise outside F±0.1%); Non-Farm Employment Change (F 89K, P 162K, surprise outside F±73K); Unemployment Rate (F 4.1%, P 4.1%, surprise outside F±0.1%). BJT 10-05 22:00 | ET 10-05 10:00 — ISM Services PMI SEP (F 54, P 55.4, surprise outside F±1.4). BJT 10-08 02:00 | ET 10-07 14:00 — FOMC Minutes. All events map to GC, SI and DXY.

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.

单品种早报深度分析

COMEX黄金GC=F 2026-10-01

Gold (GC=F) Deep Dive | 2026-10-01 — Long into 4,143 Support, Invalidation 4,132 | 2026-10-01

Silver (SI=F) Deep Dive | 2026-10-01 — Bearish: 20-Day Channel Floor at Risk | 2026-10-01

今日风险地图

单品种早报 深度分析

往期早报

⟠ 本月策略回测表现

单品种早报深度分析