Wall Street Keeps Cutting Its Gold Forecast. Central Banks Keep Buying Anyway.

Gold touched $4,090 an ounce this morning, up about 1% on the day. Nothing dramatic — just a softer dollar and oil prices easing after the Iran ceasefire held over the weekend. But zoom out and you’ll notice something weirder: gold is climbing in the same month that half of Wall Street told you to expect the opposite.

Several sell-side desks have trimmed their 2026 gold forecasts in the past few weeks. HSBC cut its 2026 average to $4,560 from $4,864 on July 9, citing a hawkish Fed and a stronger dollar. Traders are now pricing real odds of a rate increase before year-end — nine of eighteen FOMC participants penciled one in at the June meeting. That’s the kind of setup that’s supposed to be gold’s kryptonite — higher real yields, stronger dollar, less reason to hold a metal that pays no interest.

And yet. Same month, different institution, opposite conclusion.

The People With the Actual Balance Sheets Disagree

The World Gold Council surveyed 76 central banks last month. A record 45% said they plan to add to their gold reserves over the next 12 months — the highest since the survey began in 2018. 89% expect global reserves to keep growing. Of the 34 planning to add, 31 named reserve diversification as a reason.

China’s central bank just logged its 20th consecutive month of gold buying, adding another 14.93 tonnes in June — its biggest single month since 2023. And the buyer list is getting longer: Guatemala, Indonesia, Malaysia, Cambodia, Uganda and Kenya all showed up in the WGC’s data for the first time this year. This isn’t a few opportunistic buyers timing a dip. It’s the most broad-based accumulation the survey has ever recorded, happening at the exact moment sell-side analysts are downgrading the trade.

I’ve sat in enough strategy meetings to know what’s going on here. Wall Street forecasts get built on next quarter’s rate path. Central banks are building reserves for the next decade, maybe the next crisis. Different time horizon, different math. When a Goldman analyst is wrong, they revise the number next quarter. When a central bank misjudges its reserve mix, the mistake shows up during a currency crisis, and by then it’s too late to fix.

The Number That Actually Matters

Here’s the tell: central banks kept buying through gold’s worst week of the year and through silver’s brutal round-trip from $121 back into the $50s. Institutional buyers who are wrong about direction usually pull back when the price moves against them. These didn’t. That’s not a trading position — that’s a hedge against a world where the dollar’s share of reserves keeps shrinking, which it has, every year, for over a decade.

Retail investors chase headlines — a ceasefire pauses, oil drops, gold pops 1%, everyone feels smart for a day. Central banks don’t operate on that clock. They’re not trying to catch the next 1% move. They’re trying to not be the last G20 country still holding 80% of its reserves in one currency when the music stops.

What This Means for Stackers

You don’t need to predict the Fed’s next move to have a view here. The people managing trillions in sovereign reserves — who have no incentive to hype the trade and every incentive to be right — are accumulating at a record pace, in public, with disclosed numbers, month after month. That’s a better signal than any bank’s Q3 price target.

Keep doing what you’re doing: buy on a schedule, ignore the daily noise, and let the institutions with actual skin in the game validate the thesis for you. If China and dozens of other central banks are still buying after a hawkish Fed pivot, a $50-a-month dollar-cost-average plan into physical gold isn’t the reckless bet. The forecast-flipping is.