Gold & silver daily report · 2026-10-16

1. Snapshot, conventions, and conclusion

At 11:07 Beijing time on 2026-10-16, Gold-API showed XAU/USD spot gold at USD4,173.10/oz and XAG/USD spot silver at USD57.854/oz. Trading Economics' same-day cross-check was USD4,155.53 and USD57.246. This report uses the two-source midpoint—gold USD4,164.32 and silver USD57.550—as a non-tradable research reference. The 0.42% and 1.06% dispersion reminds readers that spot/CFD feeds are not one executable pool. FMP's GCUSD and SIUSD readings of 4,190.40 and 57.770 at 02:52 UTC are futures, never spot.

GLD and SLV closed at 382.71 and 51.92 on the latest completed session, 2026-10-15. The current empirical mappings are GLD≈gold spot×0.09190 and SLV≈silver spot×0.90217. These are not legal conversion ratios and will move with fees, cash, time-zone mismatch, and premium/discount.

Bottom line: gold and silver are both in a Bear regime because prices are below falling 200-day averages. Gold sits near the lower end of its technical fair zone and is not proven intrinsically cheap. Silver and SLV are technically cheap versus their own trends, while the gold/silver ratio says silver is relatively expensive versus gold. Technical cheapness, relative value, and intrinsic value are separate questions.

2. Regime and exact bands

Rule: price above a rising 200-day average is Bull; price below a falling 200-day average is Bear; other combinations are Transition. The bands are technical valuation only: cheap at or below 90% of the 200-day average, fair from 90% to 105%, and hot above 105%. Spot averages are futures continuous-close averages translated by the current spot/futures basis and are explicitly proxies.

Calling gold “cheap” would therefore mean only that it is approaching a technical lower band, not that it trades below a verifiable intrinsic value. Gold has no cash flow to discount, so this report does not invent a point intrinsic value. Silver also lacks stable distributable cash flow; industrial demand and deficits are fundamental anchors, not a DCF value.

3. Price stage: metals and ETFs

Gold: bear-market repair, not a trend reversal. The spot reference is 4.24% below its 50-day proxy of 4,348.86 and 7.26% below its 200-day proxy. GCUSD is at the 14.4th percentile of its 52-week range and 21.9% below the 52-week high. RSI(14) is 44.35; five- and 20-session returns are -0.63% and -5.32%. Support is about 4,083 and resistance 4,349–4,357. Holding or small tranches are defensible above support; reclaiming 4,357 would improve the tactical tape, while a break below 4,083 argues for risk reduction. The framework's cheap zone begins only near 4,041.

GLD: 382.71 is 3.68% below its 50-day average and 7.79% below its 200-day average. It sits 15.3% up its 52-week range, 22.8% below the high, with RSI(14) at 46.08. Support is 375.88; resistance is 397.34–401.17. It remains technically fair, not extremely cheap.

Silver: oversold rebound inside a weak structure. Spot is 10.20% below its 50-day proxy and 17.94% below its 200-day proxy. SIUSD is only 3.3% up its 52-week range and 50.0% below the high. RSI(14) is 37.88; five- and 20-session returns are -5.30% and -13.90%. Support is 55.62; resistance is 64.09 and 66.23. It is technically cheap, but only holding 55.62 and reclaiming 64.09 would turn “oversold” into credible basing.

SLV: 51.92 is 10.25% below its 50-day average and 20.72% below its 200-day average. It sits 15.1% up its 52-week range, 50.8% below the high, with RSI(14) at 38.21. Support is 50.22; resistance is 57.85–60.73. Position-controlled tranches fit better than treating one rebound as confirmation.

4. Gold/silver ratio and relative value

The reference gold/silver ratio is 72.36. In FMP's five-year matched-close sample, P10/median/P90 are about 64.02/83.23/90.26; the current ratio is at the 19.3rd percentile. A lower ratio means silver is dearer relative to gold. Silver is therefore relatively expensive, but not extreme, even though it is technically cheap versus its own 200-day trend. A move back toward 83 would restore relative value closer to the five-year median if gold were unchanged.

5. Macro: restrictive, but easing at the margin

FRED source-of-record observations are 10-year TIPS real yield 2.84% on 2026-10-14, nominal 10-year yield 5.19% on 2026-10-14, and NFCI -0.499 on 2026-10-09. FMP's fresher market cross-check put the 10-year yield at 5.10% on 2026-10-15. Yahoo returned 429 and FMP denied the DXY history field, so DXY falls back to Investing.com: about 102.06 on 2026-10-16, down 0.48% the prior day but still near its 52-week high area.

High real yields and a high dollar remain restrictive for non-yielding metals. The one-day decline in nominal yields and DXY is marginally supportive, while negative NFCI indicates loose broad financial conditions. Net assessment: restrictive, but slightly less so than yesterday, not universally supportive.

6. ETF holdings and flows

SPDR's official archive shows GLD gold holdings falling from 1,054.20 tonnes on 2026-10-14 to 1,053.91 tonnes on 2026-10-15, a -0.29 tonne daily change. This is a dated holdings change, not a dollar cash-flow estimate.

The iShares SLV request timed out at 18 seconds with an incomplete page and did not expose a verifiable 2026-10-15 tonnage. The last verified official observation is 15,371.95 tonnes and 547.25 million shares on 2026-10-14, labeled HOLDINGS_PROXY. No 2026-10-15 inflow or outflow is claimed. NAV of 51.21 was visible for 2026-10-15, but NAV is not holdings flow.

7. Supply-demand fundamentals

World Gold Council Q2 2026 data show central-bank net purchases of 289 tonnes and 345 tonnes for H1. Reserve diversification remains a strategic support, but H1 was the weakest first half since 2022; the pace matters. Global gold ETFs lost about 45 tonnes in Q2, showing that high real yields and a stronger dollar can still suppress investment demand.

The Silver Institute's World Silver Survey 2026 forecasts industrial demand of 639.6Moz, down 3%, and photovoltaic demand of 151.0Moz, down 19%, largely because of thrifting and substitution. Rising solar installations do not guarantee rising silver loadings. The 2026 market deficit is forecast at 46.3Moz, sustaining a tight long-run backdrop but not preventing price declines during deleveraging.

8. Scenario mapping

Assuming today's ETF/spot mapping remains unchanged and ignoring fees, basis, and premium/discount changes:

9. Three horizons

Tactical, weeks: Gold is a weak repair inside a Bear regime; 4,083–4,357 is the decision range. Silver needs to reclaim 64.09 before the rebound becomes structurally meaningful. Keep the core, use small tranches, and do not chase one strong day.

Cyclical, months: Upgrade to Bull only after price holds above the 200-day average and that average turns upward. Sustained declines in real yields and DXY would validate a reversal; otherwise rallies remain risk-reduction opportunities.

Strategic, one year plus: Reserve diversification supports gold; electrification and multi-year deficits support silver. PV thrifting, a strong dollar, and high real yields are genuine disconfirming evidence. Strategic exposure can coexist with disciplined rebalancing.

10. Action map and data health

• Gold ≤4,041 or GLD≤373.56: technical cheap zone; tranche only within a defined risk budget, with no intrinsic-value guarantee.
• Gold>4,357 and GLD>401.17: tactical repair confirmation; gold>4,490/GLD>415 plus a rising 200-day average is needed for Bull.
• Gold<4,083 or GLD<375.88: support failure; reduce tactical risk.
• Silver≤63.12 or SLV≤58.94 is already technically cheap; require a reclaim of 64.09/SLV 57.85 before increasing pace.
• Silver<55.62 or SLV<50.22: stop using “cheap” as a standalone buy reason and control drawdown.
• Ratio<64: silver reaches a more expensive relative extreme; ratio>83: closer to the five-year median.

Data health: Gold-API and Trading Economics provided same-day spot readings; no STALE_CRITICAL flag. Yahoo query2 and query1 both returned HTTP 429 for GC=F, SI=F, GLD, SLV, DX-Y.NYB, and ^TNX. Technicals therefore use FMP continuous daily-close fallback and all returns were calculated from consecutive closes. FMP explicitly returned ACCESS DENIED for DX-Y.NYB history, so DXY falls back to Investing.com. FRED CSV succeeded; the GLD official archive succeeded on one retry; the fresh SLV holdings field was unavailable and is marked HOLDINGS_PROXY.