Gold’s rebound gains support from lower real yields

Original research record

We lean modestly bullish on gold over 72 hours. Gold has risen while both nominal and inflation-adjusted US yields have eased. That removes some pressure from a non-interest-paying asset, but Waller still expects further tightening if the data follow his outlook and the rebound has not repaired the wider decline.

The US Treasury’s 10-year nominal par yield fell from 5.28% on 7 October to 5.22% on 8 October. Its 10-year real yield fell from 2.92% to 2.87%. These are official daily indicative curve readings, not live trade yields. Lower real yields can reduce the opportunity cost of gold; the shared move does not prove what caused gold to rise. [Nominal curve](https://home.treasury.gov/resource-center/data-chart-center/interest-rates/TextView?field_tdr_date_value=2026&type=daily_treasury_yield_curve) [Real curve](https://home.treasury.gov/resource-center/data-chart-center/interest-rates/TextView?field_tdr_date_value=2026&type=daily_treasury_real_yield_curve)

Dukascopy’s 8 October gold bid rose 0.8851% from its 00:00 open to 23:59 close. The later 02:02:16 GMT 9 October quote was bid 4174.59/ask 4175.20 dollars per ounce. That later snapshot is supporting research evidence only and will not be reused as the final reference.

How we formed this view. Observed change: The official 10-year real Treasury yield fell 5 basis points on 8 October, while the same-day gold bid rose 0.8851%. Transmission: A lower inflation-adjusted bond yield can reduce the opportunity cost of holding gold. This is a plausible supporting channel, not demonstrated causation. What price already reflects: Gold had already rebounded before this proposal. The dated 9 October 02:02:16 GMT bid of US$4,174.59 was 0.7293% above the 8 October 23:59 candle close. We are judging whether the rebound can continue from a later fresh reference, not counting the rise already observed as a forecast success. Opposition: Waller’s further-tightening outlook, still-high real yields, a firmer dollar or profit-taking could reverse the move. GLD’s separate 8 October close remained below its 6 October close, so that cross-check shows an incomplete recovery. Timeboxed judgment: Our proposed bullish 72-calendar-hour call tests limited rebound continuation through the weekend. It excludes the 14 October CPI release and retains Monday holiday-liquidity risk.

Key points The real-yield decline is directly checked, rather than inferred from lower nominal yields. The 8 October rebound and Friday Asian quote move point in the same direction, but they are one related price path, not independent forecasts. The horizon includes Friday’s scheduled Michigan consumer survey and the weekend. The 14 October US CPI report falls outside this 72-hour call. The 72-hour deadline includes the weekend and falls in Monday’s normally open Asian gold session. Monday 12 October is a US federal/bond holiday and Japanese holiday, so liquidity may differ. Provider special notices must be checked before the final freeze.

What could change this view? Waller’s 8 October speech says inflation remains too high and he anticipates additional hikes if data evolve as expected. Lower yields for one day do not establish a new falling-rate regime. GLD rose 0.73% on 8 October but its 378.62 close remained below the 6 October 382.27 close. Higher oil prices could revive inflation and yield pressure; a firmer dollar or profit-taking after the rebound could reverse gold. We have not measured representative flows or positioning. [Waller](https://www.federalreserve.gov/newsevents/speech/files/waller20261008a.pdf) A reversal back below the 8 October late-session bid area around 4144, together with renewed rises in real yields, would weaken the rebound thesis. This is a research review condition, not a trade instruction or a revised deadline.

Bottom line We lean bullish over 72 elapsed hours. We lean modestly bullish on gold over 72 hours. Gold has risen while both nominal and inflation-adjusted US yields have eased. That removes some pressure from a non-interest-paying asset, but Waller still expects further tightening if the data follow his outlook and the rebound has not repaired the wider decline. Research issue: 2026-10-09T03:13:28.717Z. Reference: indicative BID 4176.39 USD per troy ounce; displayed ASK 4177.04. Exact deadline: 2026-10-12T03:13:28.717Z. Website publication time is separate.

How likely is this call? · Three frameworks We assign 52% to a rise greater than 0.5%, 28% to a fall greater than 0.5%, and 20% to a return inside the inclusive ±0.5% band over 72 elapsed hours. Lower real yields and the measured bid rebound support 52% bullish. Still-high yields, Waller’s stated tightening view and the incomplete recovery support 28% bearish. A pause after the bounce supports 20% neutral. This is a weak-to-moderate judgement, not proof of positive after-cost expected return. These are subjective, UNCALIBRATED scenario estimates, not fitted probabilities, confidence intervals or measured success rates. The neutral scenario is separate from the sign-only directional score. The nominal and real yield curves are linked measures. The auction, commentary and gold reaction may share macroeconomic causes. GLD and gold spot are overlapping exposures, not independent confirmations. One gold session is not exchangeable with all earlier yield declines. 52+28+20=100. The three return events are mutually exclusive and exhaustive: above +0.5%, below -0.5%, or inclusive between those bounds. A small move inside the neutral band may still be a directional HIT. A missing endpoint is UNVERIFIABLE, not neutral. Coherence does not demonstrate calibration.

Evidence behind this view We started with OTR Database Data, then checked the original public sources below. [US Treasury nominal yield curve](https://home.treasury.gov/resource-center/data-chart-center/interest-rates/TextView?field_tdr_date_value=2026&type=daily_treasury_yield_curve) [US Treasury real yield curve](https://home.treasury.gov/resource-center/data-chart-center/interest-rates/TextView?field_tdr_date_value=2026&type=daily_treasury_real_yield_curve) [Waller 8 October speech](https://www.federalreserve.gov/newsevents/speech/files/waller20261008a.pdf) [Dukascopy 8 October gold BID archive](https://datafeed.dukascopy.com/datafeed/XAUUSD/2026/09/08/BID_candles_min_1.bi5) [Dukascopy indicative bid/ask board](https://www.dukascopy.com/plugins/expandGlass/indexIntl.php) [GLD daily history cross-check](https://stockanalysis.com/etf/gld/history/) [Michigan consumer survey release calendar](https://data.sca.isr.umich.edu/fetchdoc.php?docid=79628) [BLS CPI release calendar](https://www.bls.gov/schedule/news_release/cpi.htm?lv=true) [Dukascopy gold session details](https://www.dukascopy.jp/cfd-services/trading-hours/) Monday’s holiday note is based on the [JPX calendar](https://www.jpx.co.jp/english/corporate/about-jpx/calendar/?mod=article_inline) and [SIFMA fixed-income schedule](https://www.sifma.org/wp-content/uploads/2026/03/SIFMA-Capital-Markets-Outlook-2026.pdf).

It does not mean that all global gold or Japanese derivatives trading is closed.

What past price moves show The retained 8 October Dukascopy BID archive has 1440 minute records. We used 1380 in-session records and excluded the 60-minute 21:00 UTC break. Bid open 4108.005, high 4145.805, low 4105.135 and last 23:59 candle close 4144.365 give+0.8851%. The last close was near the day high. The 01:56 and 02:02 GMT 9 October live board showed a further rise, but exact intraday attribution is unproven. The separate GLD series rose 0.73% on 8 October after falling 1.67% on 7 October. Its daily US-session close is a cross-check, not a matching spot return. We do not have a matched-event backtest or a completed week of same-provider gold bars.

Evidence against this view Waller’s 8 October speech says inflation remains too high and he anticipates additional hikes if data evolve as expected. Lower yields for one day do not establish a new falling-rate regime. GLD rose 0.73% on 8 October but its 378.62 close remained below the 6 October 382.27 close. Higher oil prices could revive inflation and yield pressure; a firmer dollar or profit-taking after the rebound could reverse gold. We have not measured representative flows or positioning. [Waller](https://www.federalreserve.gov/newsevents/speech/files/waller20261008a.pdf)

Why the direction changed Earlier OTR gold/GLD bearish calls remain unchanged. This candidate adds the 8 October fall in real yields and the subsequent observed rebound. A new bullish judgement does not erase an older bearish result or claim that the long-term inflation debate is settled. The gold dealer quote and GLD fund shares have different prices, sessions and costs.

What people say vs the facts The claim that a Treasury auction supported gold is a plausible market explanation, not a proven causal finding. We checked the actual yield curves rather than treating commentary as an independent vote. No social sentiment percentage, crowd consensus or verified trader positioning is claimed.

How past forecasts performed The retained OTR ledger contains a 7 October GLD bearish 24-hour call with its original deadline on 8 October 11:30:23 UTC, but that snapshot does not contain a verified current endpoint for it. We do not award a HIT or MISS from today’s gold quote or from a later ETF close. No resolved same-instrument 72-hour rebound sample is available to calibrate this judgement.

Frozen research reference Dukascopy Bank SA Gold rolling-spot indicative quote / XAU/USD, USD per troy ounce. Directly displayed indicative BID 4176.39, ASK 4177.04. Scoring reference is BID. Provider quote-panel snapshot clock 2026-10-09T03:13:08Z; this is a genuine displayed snapshot clock, not an individual trade or central-exchange timestamp. Retrieved 2026-10-09T03:13:26.639Z. Frozen research issue 2026-10-09T03:13:28.717Z; snapshot age at issue 20.717 seconds. No executed-trade or firm-fill claim. Optional midpoint 4176.715 is calculated from displayed BID/ASK and is not the scoring reference.

Resolution contract Frozen P0 is the Dukascopy directly displayed indicative BID 4176.39 USD per troy ounce for Gold rolling-spot indicative quote / XAU/USD. Exact deadline 2026-10-12T03:13:28.717Z, 72 elapsed calendar hours after research issue 2026-10-09T03:13:28.717Z. Use the nearest legitimate in-session one-minute BID OPEN from the same product, comparing UTC bar-start timestamps with exact expiry; earlier observation wins an equal-distance tie. Retain source prices, source interval, actual observation timestamp and signed offset. A bar-start quote proxy is not an exact executed trade. Exclude carried records in confirmed closed sessions; do not automatically discard genuine open-session zero-volume quote bars. Do not replace missing historical data with a current quote. Return=(P1/P0-1)*100. BULLISH HIT if P1>P0; flat or lower is MISS. The inclusive ±0.5% neutral scenario band is separate and never changes this sign-only score. If no legitimate dated endpoint exists, mark UNVERIFIABLE and do not silently extend expiry. No trading profitability is implied. Rolling-spot gold normally trades Sunday 22:00UTC through Friday 21:00UTC with daily 21:00–22:00UTC breaks. Exact expiry falls Monday 12 October in the normally open session. US federal/bond and Japanese holidays may change liquidity; any actual session gap must remain visible.