Report Date: 2026-11-15 | Industry View: Neutral (Wide-Range Oscillation) | Observation Horizon: 12 months | Base Price Range: USD 44-52/oz (center slowly declining from ~50 to 46)
Abstract: Silver is in the mid-to-late stage of a historic rally—a sixth consecutive year of supply-demand deficit (a shortfall of ~35 Moz in 2026, per the Silver Institute's 2026-11-12 figures) combined with forced physical reallocation (China's export controls, the US tariff episode, London's drained inventories) pushed silver from ~USD 23 in early 2024 to an all-time peak of USD 121.65 on 2026-01-26; a subsequent judicial unwinding of tariffs, rate-hike expectations, and deleveraging sent prices down 60% over ten months to USD 48.63 (Kitco, 2026-11-15). Our neutral view: the squeeze premium has been cleared, but the physical tight balance has not yet flipped—the deficit narrows to ~-14 Moz in 2027 and approaches balance only in 2028 (+7 Moz); combined with the supply rigidity of a 74% by-product structure, China's physical siphoning (Shanghai-London premium +11.7%), and an overhang of 1,250 Moz in ETF holdings, silver will most likely trade in a wide 44-52 USD range over the next 12 months with a slowly declining center, with both fat tails: upside (rate-cut return + squeeze re-ignition) and downside (double rate hikes + weaker gold). Value chain profits are highly bifurcated: low-cost primary miners and streaming companies earn real money (AISC USD 12-16 vs silver at USD 44-50), while smelting and PV silver paste are textbook "revenue growth without profit growth" (Zhangjiagang/Hejing—Juhé Materials: 2026 9M revenue +39%, net profit -95%, operating cash flow -RMB 4.65 billion). Top picks: Wheaton, Pan American, Shengda Resources, Xingye Silver-Tin; avoid Juhé Materials/Dike Co., Ltd., and Hunan Silver.
This study covers the global silver market and the silver mining value chain: upstream mined silver (846.6 Moz in 2025, of which primary silver mines account for 26.1%, lead-zinc by-product 29.4%, copper by-product 28.0%, gold by-product 15.9%) and recycled silver (197.6 Moz); midstream smelting and refining (LBMA good-delivery system, Chinese smelting capacity); downstream processing (PV silver paste/silver powder, electronic pastes, silver brazing strip) and end demand (industrial 58.1%, bars & coins investment 19.3%, jewelry 16.7%, silverware 3.7%, photography 2.1%, in 2025); and financial intermediation (London OTC spot as the physical anchor, COMEX/SHFE futures for price discovery, ETF/ETP holdings of 1,250 Moz). Per the user's request, the research focuses on three topics: industrial and investment demand, the by-product supply mechanism, and PV thrifting and substitution risk. Pricing is globally integrated: London spot lease rates and free float set tightness, and COMEX/SHFE futures and the EFP (futures-spot basis) transmit cross-regional spreads.
Total global silver demand in 2025 (fabrication basis, excluding ETFs) was 1,130.6 Moz (-2%, the lowest since 2021; Silver Institute/Metals Focus, World Silver Survey 2026). By structure: industrial 657.4 Moz (58.1%, first post-pandemic decline of -3%), bars & coins 217.7 Moz (19.3%, +14%), jewelry 189.3 (16.7%, -8%), silverware 42.1 (3.7%, -21%), photography 24.2 (2.1%, multi-decade low). On 12 November, the Institute's mid-year update revised 2026 total demand down to ~1,080 Moz (-5%), with cuts concentrated in PV and jewelry/silverware eroded by high prices.
PV is the largest industrial end market (186.6 Moz in 2025, 16.5% of total demand) and currently the largest negative variable: PV silver demand is expected to fall to ~147-151 Moz in 2026 (-19%, -35.6 Moz, the largest single-year decline ever). The driver chain is fully quantified: the 2025 average silver price rose 42% to USD 40.03, silver's share of cell costs jumped from 8-10% to over 20%, forcing thrifting—fleet-average loading fell from 8.32 mg/W in 2025 (5,804 tonnes / 698 GW installed) to ~7.34 mg/W in 2026; 2025 technology-route averages were roughly TOPCon 10.7, HJT 11.9, BC 12.4 mg/W; levers include 0BB busbar-free designs (a further 10-20% saving vs SMBB), silver-coated copper paste (silver content already down to 10%), and ultra-fine printing, with an industry target of below 5 mg/W for mainstream cells by 2027 and below 1 mg/W with copper plating at the limit. But the thrifting slope may be overestimated: Fraunhofer ISE (2026-04) did cut TOPCon silver loading to 1.1 mg/W in the lab, but the industry fleet average remains 10-12 mg/W—reaching 5 mg/W within a year requires a further ~55% cut; copper plating faces a yield bottleneck from laser-grooving lattice damage (crystallinity down ~30 percentage points, requiring annealing repair); silver-coated copper in HJT took four years to reduce silver share from 60% to 15%. Accordingly, our three scenarios for 2028 PV silver demand are: 88.4 (fast thrifting) / 130 (neutral, revised up on latest evidence) / 153.1 (slow thrifting) Moz, with the probability of "thrifting slower than planned" no less than 30%.
Non-PV industrial demand is the only strong incremental engine (470.8 Moz in 2025, comprising non-PV electrical & electronics 262.9 + brazing 50.5 + other industrial 157.4): AI data centers (silver paste/interconnects/thermal interface materials, to exceed 10% of electrical & electronics demand in 2026, above 42 Moz), grid and contacts, EVs (European BEV share 15%→22%), and ethylene oxide catalyst change-out cycles—together growing ~4% annually.
Investment demand is the largest marginal buyer and seller of price: bars & coins rose +14% to 217.7 Moz in 2025, briefly surged to 257.6 Moz in 2026 (highest since 2022), then normalized with falling prices; ETF/ETP holdings ended 2025 at a record 1,317.6 Moz (+273), with net redemptions of 55 Moz in 2026 to 1,250 Moz—this overhang both caps rebounds (redemption selling pressure) and acts as a release valve for physical supply on the downside. Regionally: East Asia accounts for 58.5% of industrial demand (China alone is 43% of global industrial demand); India accounts for 38.5% of investment demand and 37.1% of jewelry demand.
Item-by-item: PV thrifting -68.6 (186.6→118 Moz, April baseline); non-PV industrial +58.8 (AI/grid/EV, +4% per year); jewelry & silverware -30.7 (modest recovery from the 2026 trough after high-price destruction); photography -3.9; bars & coins +22.3 (normalization to 240 after the 2026 peak); slow-thrifting adjustment +12 (copper-plating yields and silver-coated copper penetration lagging the roadmap, 2028 PV silver base revised from 118 up to 130); basis revision -33 (the Institute's 2026-11-12 revision of 2026 total demand from 1,112.6 to ~1,080, shifted in equal amount). Net change totals -43.1, with 2028E total demand of ~1,087 Moz (neutral range 1,020-1,170; the largest swing item is PV).
Mined silver in 2025 was 846.6 Moz (+3%), with by-products totaling 73.9% (625.6 Moz): lead-zinc by-product 249.1, copper by-product 237.3, gold by-product 134.7, other 4.5; primary silver mines only 221.1 (26.1%, a record low). The incentive effect of silver prices on mine output is nearly nil: the 2025 average price rose 42% to USD 40.03 and 99% of primary mine output was profitable above cost, yet 2026 mine output actually fell 0.5% to 834 Moz (Institute, November basis)—because 74% of silver is a by-product of copper/gold/lead-zinc mines, with output determined by base-metal prices and capex: copper by-product output has grown at a +2% CAGR since 2016 and +6% in 2025 (production gains at Antamina/Escondida/Las Bambas); gold by-product +5% in 2025 (gold mines at full capacity amid record gold prices); lead-zinc by-product flat as a trend (weakest zinc mine capex). The copper/gold capex cycle points upward but modestly; estimated by-product annual increments are 15-25 Moz—low price elasticity is the most important supply feature of this cycle.
Cost basis (by-product credit method, Metals Focus): 2025 primary silver mine weighted cash cost USD 5.21/oz, AISC USD 12.21/oz (rising to 16.27 in H1 2026, +33% cost inflation); regional AISC: North America 10.28 / Asia 10.74 / Central & South America 13.98 / Oceania 14.15 / CIS 16.23 USD. At the 2025 average of USD 40.03, 99% of primary output was profitable; marginal capacity (90-99th percentile) AISC ~USD 25-40; new-project DFS use base prices of USD 40-50 (AbraSilver Diablillos used 50; at a floor of 40, IRR is still 31%), with reserve economic cutoffs cumulative up 2.5-4.0 USD/yr over two years. By-product mines' net silver cost can be deeply negative—for them, silver is a zero-marginal-cost "free option."
Project pipeline (first-production timing is key): Endeavour's Terronera achieved commercial production in 2025Q4 (blocked by community protests for two weeks in 2026-08 before resuming); Coeur consolidated New Gold for USD 7 billion (2026 silver guidance ~20.3 Moz); PAAS completed its acquisition of MAG for a 44% stake in Juanicipio (2026 attributable guidance 6.0-6.5); Vizsla's Panuco first production in H2 2027; AbraSilver Diablillos awaiting FID in 2027Q2 with first production late 2029; Hecla's Keno Hill declining rather than growing—2026 guidance cut to 2.2-2.6 Moz (3.0 in 2025), Q4 throughput down to ~300 t/d with further reductions from January 2027; after the Grasberg mudflow, Freeport restored only 50% of capacity by May 2026, with full output not until end-2027. Construction timelines mean the effective 2026-2028 increments are essentially locked to the named projects: 8-12 years for greenfield, 2-4 years from FS to production—a price spike cannot be converted into new mines within two years. Recycled silver was 197.6 Moz in 2025 (13-year high) and ~211 in 2026F (+7%, first time above 200 since 2012), but capped by North American/European refining bottlenecks. Miner hedge books surged from 5.3 Moz at end-2024 to 50.0 at end-2025 (5.9% of mined output, highest since 2013)—locking in prices at highs is itself miners' vote on durability; 2026 net hedging supply goes to zero. Reserves: primary silver mine reserves 3,762 Moz (net -50 in 2025), ~17 years static life; USGS global reserve basis 610,000 tonnes, ~23.5 years. Silver was added to the US critical minerals list in November 2025.
Mined silver supply path (our estimates): 834 (2026E, Institute) → 845 (2027E, incorporating Keno Hill cuts, half-year Grasberg contribution, Fresnillo grade decline) → 870 (2028E, Panuco ramp-up + restarts) Moz; total supply 1,045 → 1,062 → 1,094 (including recycling 210/216/223 and net official sales of 1, hedging at zero).
| Year | Demand | Supply | Deficit | Notes |
|---|---|---|---|---|
| 2024A | 1,157.4 | 1,019.6 | -137.8 | WSS 2026; incl. net hedging 0 |
| 2025A | 1,130.6 | 1,090.4 | -40.2 | Net hedging supply 44.7 |
| 2026E | 1,080 | 1,045 | -35 | Institute 2026-11-12 basis (April old basis -46.3) |
| 2027E | 1,076 | 1,062.3 | -13.7 | Our estimate: mined output revised down to 845 |
| 2028E | 1,087 | 1,093.8 | +6.8 | Our estimate: PV silver base 130 |
The meaning of the balance matters more than the numbers: the cumulative 2021-2026E deficit is ~746-751 Moz, but visible inventories did not deplete commensurately—~248 Moz was absorbed by net ETF inflows (cumulative +7,960 tonnes 2021-2025) and ~250-300 Moz sank into retail bars & coins, ultimately clearing as a squeeze in January 2026 via the exhaustion of London free float (~136 Moz in September 2025, less than 0.3 days of spot market volume). Current inventories (divergent, not loose): COMEX total 332.8 Moz (registered 101.7, 2026-11-12; total down ~37% from the September 2025 peak); LBMA London vaults 928.8 Moz (end-October 2026, +4.3% YTD but turning down month-over-month in October); SHFE warrants hovering low at 44.1 Moz. Paper re-stocking coexists with physical tightness: from 2026-01-01 China imposed license + state-trading management on silver exports (44 companies, case-by-case approval), reversing from net exporter to net importer (528 tonnes imported in March alone, the largest in nearly two decades), with the Shanghai silver benchmark at a +11.7% premium to London in October 2026 (63.69 vs 57.00 USD)—physical released by Western ETF redemptions is being siphoned East; "global synchronized easing" does not hold. We set the Shanghai-London premium as the veto indicator for this view: if it stays above 8% for two consecutive months, the price center must be revised up.
The full chain of this rally: the US reciprocal tariff episode (2025-08, the 39% rate on Switzerland at one point touched 1 kg silver bars) → record EFP, dealers over-stocking New York and draining London → the October 2025 London squeeze, silver breaking USD 50 for the first time in 40 years → China's export controls layered on → first break of USD 100 on 2026-01-23, peak of USD 121.65 on the 26th (COMEX registered coverage ratio at one point just 14%) → the Supreme Court's 2026-02-20 ruling that IEEPA tariffs exceeded authority, premium clearing, a single-day -38% → back to USD 48.63 in November 2026 (-60.0% from peak, still +21.5% vs the 2025 average). The gold/silver ratio recovered from January's extreme low of 43 to 79.9 (2026-11-14, gold USD 3,883)—already above the upper bound of JPMorgan's "physical-balance-determined normalization range" of 70-75; silver is no longer expensive relative to gold. Historical cycle template comparison: -78% within weeks after the 1980 Hunt squeeze; -72.5%/4.7 years peak-to-trough in 2011-2015; this round -60%/10 months from peak, already in the mid-to-late stage of the historical end-game drawdown—but the bottom is likely significantly above the historical template: recycling supply (+7%) and jewelry demand destruction (-16%) were strongly activated above USD 40, forming a composite soft floor rather than relying on an accounting-style "incentive price."
The term structure confirms "tight but not squeezed": COMEX shallow contango (2026-11-13: Nov 48.58 → next Feb 49.02 USD, annualized +3.6%, below the ~5% cost of carry, implying near-zero lease rates—the deep backwardation and double-digit lease rates of the squeeze period have fully normalized, though still above the pre-crisis norm of ~0.5% YTD); SHFE calendar spread flat at +0.33%. None of the three preconditions of a squeeze (thin free float, falling exchange inventories, spiking lease rates) has been truly eliminated—they are merely suppressed by rate hikes—even with London free float recovering to ~267 Moz, that is only 0.6 days of spot volume.
Pricing framework: ① composite floor ~USD 40-42 (recycling supply curve + jewelry/silverware demand elasticity + by-product structure costs + gold linkage β—three legs, not a single "incentive price"); ② balance direction: deficit narrows to -14 in 2027 and approaches balance in 2028, with the scarcity premium fading quarterly but not reversing; ③ positioning: the 1,250 Moz ETF overhang caps rebounds, London free float is fragile, and China's net-import siphoning provides a floor; ④ macro: real rates at 2.91%, the market pricing ~77% odds of a December hike (a crowded hawkish consensus, with large reverse-repricing risk); JPMorgan/HSBC's 2026-2027 forecasts (63-75 USD) were made at higher price levels, and our divergence must be stated explicitly: their path implies no further macro tightening, while our base case implies one of the double hikes lands. Base-case center declines slowly from 50 → 46 USD (-8%), with a probability-weighted expected price of ~52 USD.
Silver is a "fragmented industry with oligopolistic sub-industries," and two sets of measures must be distinguished: all-in concentration (denominator = global mine silver production of 846.6 million oz) CR3 ~14.4% (Fresnillo 48.7 + KGHM 43.3 + Newmont ~30), CR5 ~19.8%; primary silver miner concentration (denominator = primary silver pool of 221.1 million oz, the basis in the chart below) CR3 ~40.5%, CR5 ~55.1%, and consolidation is accelerating — record $10.8 billion in M&A announcements in 2025 (Coeur–New Gold $7 billion, PAAS–MAG $2.1 billion).
Profit pool distribution (current silver price $44–50 vs primary AISC $12–16): ① Low-cost primary miners — cash profit on the order of $30/oz; Shengda Resources' gross margin in the first three quarters of 2026 was 69.3%, net margin 39.6%, operating cash flow of RMB 980 million exceeding net profit; ② By-product producers — silver is a free option; KGHM's 43.3 million oz converts almost fully into profit; ③ Streaming/royalty — Wheaton's silver streaming cost is $8.95 with cash operating gross margin ~85%; revenue, net profit and cash flow all hit records in 2025, the best return quality in the entire chain; ④ Midstream smelting — Hunan Silver's revenue in the first three quarters of 2026 was RMB 13.4 billion but net profit attributable to parent only RMB 320 million (net margin 2.4%), operating cash flow of –RMB 2.42 billion (rising silver prices drain working capital); ⑤ PV silver paste — Paste Technology (Solayer), global No.1 in shipments, saw revenue +39% over the same period while net profit attributable to parent fell –95% (only RMB 12 million), gross margin 4.2%, operating cash flow –RMB 4.65 billion: a triple squeeze from volume-price double hit (PV silver –19%) + price pressure from concentrated cell-manufacturer customers + working capital tied up by high silver prices; all foreign brands (DuPont, Heraeus, Samsung SDI paste businesses) have exited. Bargaining power conclusion: the mining end is strong toward upstream suppliers, has no pricing power over the market but is bolstered episodically by supply rigidity; midstream/downstream is the weakest link of the chain — concentrated downstream customers capture the prosperity dividend. Direction of profit migration: shrinking PV silver demand → demand profit pool migrates toward non-PV industrial (+4%); midstream/downstream continues to consolidate and shake out; mining super-profits revert with mean reversion in silver prices, but supply rigidity makes a step-up in the earnings center not unlikely (see I7, Controversy 2).
Valuation positioning: SW Precious Metals Index (801053.SI) PE 19.91x (~1st percentile since July 2014), PB 3.30x (~57th percentile), as of 2026-11-13. The combination of low PE + mid-range PB is classic cyclical pricing: the divide between "discounting cycle-peak earnings" and "rising earnings center with reasonable valuation" is precisely the core controversy in I7 — if the 2027 deficit persists (–14 million oz), the 1st-percentile PE leans toward the latter. Stock anchors: Shengda Resources PE(TTM) 20.3 (5th percentile over 1 year) / PB 4.48 (72nd percentile over 10 years).
Key player profiles: Fresnillo (global No.1; 48.7 million oz in 2025, –13.5%; 2026 guidance cut to 42–46.5; single-country concentration in Mexico) | Pan American (22.8; consolidating 44% of Juanicipio; 2026Q3 standalone free cash flow $377 million, cash $1.9 billion; the consolidation winner) | KGHM (43.3; the copper by-product archetype; silver = free option but stock tracks copper) | Hecla (17.0; 2026 guidance 15.1–16.5, the Keno Hill production-down year is the main drag) | First Majestic (silver 15.4 / silver-equivalent 31.1; Gatos consolidation lowers costs; Mexico high-beta) | Coeur (~20.3 post-consolidation; volume step-up) | Wheaton (streaming; costs locked) | Shengda Resources (purest silver mine play in A-shares) | Xingye Silver-Tin (dual leverage from Yinman + tin; FY2025 net profit attributable RMB 1.7 billion) | Hunan Silver (smelting, a "pass-through middleman") | Paste Technology (Solayer) (a specimen of illusory prosperity in silver paste).
Trade and tariffs (net effect: most distortions have cleared, tail risks remain): the US IEEPA reciprocal tariff episode (August 2025 Swiss tariff hit silver bars → squeeze → Supreme Court ruled 6–3 on 2026-02-20 it exceeded authority); Section 122 bridge tariffs of 15% (from 2026-02-24) struck down by the Court of International Trade in May 2026; under the US–Swiss framework, Section 301 still imposes ~12.5% on Swiss silver bars (from 2026-07-24; PAMP/Argor shipments to the US taxed) — supplementary NYC delivery brand inflows remain suppressed. China: from 2026-01-01 silver exports are folded into state-trading + licensing management (MOFCOM Announcement No. 68 of 2025; 44 firms, review of each shipment, continued in 2027), combined with record imports — a major driver of the London squeeze; this is the largest structural positive on the policy front this cycle. India: effective import duty on gold/silver bars raised from 6% to 15% (effective 2026-05-13), raising landed costs, suppressing jewelry/investment demand and stimulating smuggling — the main policy negative on the demand side. Resource countries: Mexico's 2023 mining law reform upheld by the Supreme Court (30-year concessions + public tenders); the new General Water Law (effective end-2025) places mining water use at the bottom of priority and reverses the burden of proof — systematic tightening of new-supply permitting in the world's largest silver producer (172.9 million oz, 20.4% of global output); Peru's Fujimori government (sworn in June 2026) is pro-mining, but community blockades in the southern corridor have become cyclical; Bolivia is pivoting to attract investment but constrained by ~500,000 cooperative miners; Russian silver has not been directly embargoed, but after LBMA halted new good-delivery status, flows shifted east (–80%… no, +80% to China in 2025H1). ESG/permitting: water rights and community consent are the biggest soft constraint on new supply; the US listing silver on the 2025 critical minerals list (2025-11-06) eased domestic permitting. Net direction: structurally bullish for prices (Chinese controls + Latin American constraints + critical mineral status), while realized negatives (tariff premium clearing, India tariff) suppress demand elasticity — policy raises the floor and amplifies volatility, rather than pushing prices up one-sidedly.
| Scenario | Probability | Price Range | Narrative |
|---|---|---|---|
| Bear | 25% | $32–40 | December and following-March hikes both delivered, real rates above 3%, weaker gold, accelerating ETF outflows (–100 million oz/yr), smooth PV thrifting (2028 PV silver 88–118) — a repeat of the 2011 template, but with a floor above the historical mapping (27–34) |
| Base | 45% | $44–52 (center 50→46) | Deficit of –14 in 2027, near-balance by 2028; macro "higher for longer"; slow ETF outflows (–45/yr); Shanghai-London premium maintained at 5–10% |
| Bull | 30% | $58–85 | No December hike or dovish dot plot, rate-cut trade returns; ETFs flip to net inflows; free float tightens again, lease rates spike re-igniting a squeeze; or PV thrifting significantly behind plan (2028 PV silver 150+) — $63–80 (JPMorgan/UBS targets) becomes the magnet zone |
| Date | Event | Directional Impact | Controversy |
|---|---|---|---|
| 2026-12 | Fed rate decision (hike priced at ~77%) | Delivered → shift toward bear case; not delivered → bull case triggered | D3 |
| 2027-01 | Silver Institute annual update, miner annual reports and 2027 guidance (Keno Hill throughput) | Window for validating/falsifying balance sheet and pipeline | D2 |
| 2027Q1 | Execution of China's 2027 annual silver export license list | Tightened → bullish; relaxed → bearish | D2/D3 |
| 2027-03 | Fed rate decision (second hike window) | Same as December | D3 |
| 2027-04 | WSS 2027 release (initial 2027 balance) | Official check on the –14 deficit assumption | D2 |
| 2027Q2 | AbraSilver Diablillos final investment decision | Leading signal of supply additions | D2 |
| 2027H1 | LBMA monthly vaults / weekly ETF holdings (ongoing) | Free float and overhang selling pressure | D2/D3 |
| 2027H2 | Vizsla Panuco first production | Key delivery on the 870 mine-supply path | D2 |
| 2027 Full Year | PV "mainstream cell <5 mg/W" target milestone | Achieved → D1 bears win; missed → bulls win | D1 |
| 2027-06/09 | Mexico water law enforcement, Peru political anniversary window | Supply disruption risk | D2 |
Chart reading: the mining end and streaming fall in the upper-left quadrant — under the base case, silver price prosperity is neutral-to-weak (center drifting lower), but low costs and supply rigidity keep return quality very high; these are assets to be "allocated with timing based on valuation and macro inflection, not avoided"; Fresnillo/Hecla sit lower on the x-axis as 2026–2027 are production-down years. The lower-right quadrant contains only recycling & smelting: recycling volumes have rising prosperity (+7%/yr), but net margins are 2–3% and rising silver prices devour working capital (operating cash flow –RMB 2.4 billion) — volume-driven prosperity ≠ returns; the chain's most typical value trap. PV silver paste sits lower-left (volume-price double hit with poor returns) and belongs to "avoid" rather than "value trap."
| Scenario | Biggest Beneficiaries | Biggest Losers |
|---|---|---|
| Bear (32–40) | Wheaton (cost locked at $8.95), KGHM (free by-product option, main business tracks copper) | Shengda Resources / Xingye Silver-Tin (high-beta pure silver in A-shares + high PB percentile), First Majestic (higher own-mine costs + single-country concentration) |
| Base (44–52) | Pan American (volume step-up + cash flow), Wheaton, Shengda Resources / Xingye Silver-Tin (cash profit band still ~$30/oz) | Paste Technology (Solayer) / Dion (Dike) (PV silver –19%, negative cash flow from funded working capital) |
| Bull (58–85) | Shengda Resources / Xingye Silver-Tin (highest purity, A-share elasticity), First Majestic (high beta), Wheaton | Paste Technology (the higher the silver price → the more capital tied up and the faster thrifting, hurt inversely) |
Beneficiaries (preferred): ① Wheaton Precious Metals — asset-light silver price exposure with locked costs, best return quality in the chain; ② Pan American Silver — Juanicipio consolidation ramping, the M&A consolidation winner, 2026Q3 standalone free cash flow of $377 million; ③ Shengda Resources (000603.SZ) — the purest A-share pure-silver mine elasticity (net margin 39.6%, cash flow exceeding net profit); main risks are silver price downside and the PB 4.48 percentile; ④ Xingye Silver-Tin (000426.SZ) — dual-metal elasticity from Yinman silver + tin, FY2025 net profit attributable RMB 1.7 billion (note its silver-tin by-product structure). Avoid: Paste Technology (Solayer) / Dion (Dike) (PV silver paste; volume-price double hit + customer price pressure + capital tied up; classic growing-revenue-not-profit); Hunan Silver (smelting, a "pass-through middleman"); high-cost single-country second-tier miners absent from the bull case with the greatest bear-case elasticity. Fresnillo/Hecla kept on watch (production-down years; could be upgraded if 2027 guidance stabilizes). Individual profiles (e.g., Shengda Resources, Wheaton) can be created on request for price targets and single-stock deep dives.
① Macro: December and March hikes both delivered with real rates above 3%, pressuring precious metals broadly (bull case falsified); ② Demand: PV thrifting slope faster than expected (2028 PV silver falling to the 88 band), balance flipping to surplus early; ③ Supply: smooth commissioning of Panuco/Diablillos et al. combined with the above demand scenario, 2028 surplus exceeding expectations; ④ Policy: relaxation of Chinese export controls or a cut in India's tariff altering trade flows and demand elasticity; ⑤ Events: squeeze re-ignition (upside risk) or accelerated ETF redemption (downside risk) amplifying volatility; ⑥ The 2027–2028 balance sheet in this report is our own estimate; pipeline delivery and revisions to institute data may push actual outcomes ±30 million oz from the neutral path.
Key data sources: World Silver Institute/Metals Focus World Silver Survey 2026 and interim update of 2026-11-12; LBMA London vault data; USGS 2025 critical minerals list; MOFCOM Announcement No. 68 of 2025; Kitco spot quotes; CME/SHFE exchange data; Fraunhofer ISE; company disclosures from AbraSilver/Pan American/Hecla et al.; JPMorgan/HSBC research (forecasts made at higher price levels; divergences noted). Price percentiles are estimates from daily distributions.