Gold Forecasts Were Cut Again. The 2027 Number Went Up

Gold settled Tuesday at $4,060.54 on Investing.com's XAU/USD series, up 1.09%, after an Asian session that dipped to $3,989.33. That is the tape. The more interesting document is a survey published the same morning: a Reuters poll of 30 analysts and traders that cut the 2026 gold forecast for a second consecutive quarter.

The headline reads like another downgrade. The numbers say something else.

The cut is $14. The revision is up

The median 2026 forecast came in at $4,495 an ounce, down from $4,509 in July. On a $4,500 number that is 0.3%. Line up the last three polls:

• April 2026: median 2026 forecast $4,916, the highest annual call in Reuters polls going back to 2012.
• July 2026: $4,509, a $407 cut and the first downgrade in 11 quarters.
• October 2026: $4,495, a $14 cut.

The 2026 line has stopped falling: the downgrade collapsed by 96.6% in one quarter, which matters more than the level. Read the other line: the 2027 median went from $4,610 to $4,650 — $5,100 in April, $4,610 in July, and now raised by $40.

The same 30 people trimmed this year's number and lifted next year's. That is not a bearish document; it has given up on timing and kept the thesis. The stated reasons are the ones this desk has mapped since September's hike: high Treasury yields, a firm dollar, and policy restrictive for longer.

Why the timing slipped

The mechanism is the reverse of the usual war trade — the part most gold coverage gets wrong this month.

Bullion's January record was $5,595 (the CFD series at Trading Economics printed $5,608.35), set before the Iran conflict pushed crude higher. Since then the chain ran the other way:

• Brent has hovered near $116 a barrel, with renewed Houthi attacks on Saudi Arabia keeping supply risk priced in, though it slipped more than 1% on Tuesday as fresh diplomacy raised hopes of avoiding escalation (Reuters via Kitco, CNBC).
• Energy there keeps inflation risk in the frame, which keeps the Fed restrictive: markets price a 9–10% chance of a hike on Wednesday and 82–83% of a 25bp increase in December (FXStreet).
• The 10-year sits at 4.99% (Trading Economics) and the dollar is near a two-week high, up 0.3% on Tuesday (Reuters via Kitco).
• Higher real yields make carrying a metal that pays nothing expensive, so October has been liquidation: gold bottomed at $3,975.64 on Friday, its lowest since mid-July.

The awkward part of the war-hedge story: energy-driven inflation is bullish for bullion only if the central bank is behind the curve. In a hiking cycle it is a tax on the trade, not a subsidy. Standard Chartered's Suki Cooper calls prices "at a critical juncture", with technical support at $3,886 an ounce, below which it "would fade quickly" (Yahoo Finance).

The silver line gives the survey away

The poll's silver forecasts are where the consensus stops being internally consistent.

Analysts see silver averaging $70.30 in 2026 and $66.75 in 2027, both cuts, for the same reasons: weakening solar demand, manufacturing substitution, and a market left oversupplied by this year's ETF reversal. StoneX's Rhona O'Connell put it bluntly — with "solar currently oversupplied and AI under pressure", silver "is likely to underperform gold" (Reuters).

Divide the forecasts by each other.

• Spot, Tuesday: $4,060.54 over silver near $53 gives a gold/silver ratio of 76.6.
• The poll's 2026 medians: $4,495 over $70.30 gives 63.9.
• The poll's 2027 medians: $4,650 over $66.75 gives 69.7.

Every implied ratio sits below the current one. A survey that cut silver twice still has silver outperforming gold by 19.8% on its 2026 numbers and 10% on 2027. Either the gold forecasts are too high or the silver cuts are too deep; the poll cannot have it both ways, and its own commentary argues for the silver cut.

The honest reading: the consensus is far more confident about rates than about prices. A high-for-longer yield path puts fair value below $4,495; a December hike followed by a pause bids the metals complex with gold grinding rather than leading.

What the number actually requires

At Tuesday's close the 2026 median, $4,495, sits 10.7% above spot with 65 days left. That is not a forecast of November and December; a two-month average that high needs a rally larger than anything since January. It is a leftover from a year whose first quarter traded above $5,000.

The 2027 median, $4,650, is 14.5% above spot with twelve months to work with. That is the number the survey is really about, and it is the one that moved up.

Against it sits a more aggressive one. Metals Focus expects gold to average $5,330 in 2027 — 16% above its own $4,600 estimate for this year — with a fourth-quarter average of $5,600 and peaks approaching $6,000 (Kitco). The gap between $4,650 and $5,330 is not about the Fed. It is about whether de-dollarisation and fiscal risk re-enter the price at all.

Cooper expects prices to "consolidate and recover in coming months as market focus shifts to de-dollarisation and currency debasement". O'Connell's is subtler — central bank purchases are "sending the markets a message about managing risk, which is more important than the physical tonnage" (Reuters).

The level map decides it, not the forecast

None of this says where Wednesday goes, and the desk carries no position through the FOMC decision or Thursday's PCE. It only says which prices would prove the survey wrong.

• $4,060–$4,062 — Tuesday's close, and the support whose failure on 22 October started this leg down. A daily close above it is the first evidence liquidation has finished.
• $4,100 — the resistance trendline from September's highs. Above it the survey's 2026 number stops looking generous.
• $4,219 and $4,288 — the 100-day and 50-day moving averages. Neither is reachable without a catalyst larger than a hold.
• $3,989.33 — Tuesday's low; $3,975.64 and $3,978.15 — Friday's and Monday's lows; $3,959 — the 17 July low; $3,941 — the year-to-date low.
• $3,886 — Cooper's line. Lose it and the 2026 forecast discussion becomes academic.

The constructive case needs three things in order: a Wednesday hold that does not promise a December hike, oil that stops making new highs, and a close above $4,062 that survives a second session. Two are decided by other people this week.

What to take away

• The consensus stopped cutting 2026 and started raising 2027 — a timing revision, not a change of thesis. Forecasts are not tradeable; $4,062 settles the argument this week.
• Thursday's PCE matters more than Wednesday's statement: the decision is priced, the print is not.
• A consensus that contradicts its own arithmetic is an opportunity, not a signal. Sovereign buying is a 2027 argument; it has never once stopped a 6% monthly drawdown.
• Today's session plan is here. The rest of this month's macro work sits on the news board and under central banks.

Sources

• Reuters — analysts cut the 2026 gold forecast for a second straight quarter, 2027 raised to $4,650
• Reuters via Kitco — gold retreats on a stronger dollar as traders brace for the Fed
• FXStreet — gold reclaims $4,060 as oil retreats, December hike odds at 83%
• CNBC — gold gains as softer oil offsets a firmer dollar
• Kitco — Metals Focus sees $5,330 average gold in 2027
• Reuters via Kitco — the July poll: first gold forecast cut since late 2023, from $4,916 to $4,509
• Yahoo Finance — Standard Chartered's Suki Cooper on the $3,886 line

Not financial advice. Levels here are reference points for reading the session, not recommendations, and gold can gap through any of them.