Has the global central bank "gold hoarding era" come to an end?

null

Original author: Zhao Ying

Original source: Wall Street Gainers

The hottest question in the market recently is: Are global central banks actively selling gold? Has this 15-year official “gold-buying spree” finally come to an end?

According to the Pursue the Trend Trading Desk, UBS strategist Joni Teves gave a clear judgment in the latest precious metals research report published on April 2: the likelihood of central banks making a structural shift and engaging in large-scale gold selling is extremely low. Official institutions will still maintain a net buying stance, but the pace of purchases will ease moderately—UBS expects full-year gold buying in 2026 to be about 800 to 850 tons, slightly lower than roughly 860 tons in 2025.

The report targets one of the most eye-catching recent cases—the news that Turkey “sold about 50 tons of gold within a few weeks.” Teves believes: in Turkey’s official gold data, there are traces of operations such as commercial bank positions and swaps; inferring “central banks have started selling” from headlines alone is high risk. The judgment should wait until more granular split data is available.

At the price level, UBS defines the short term as “a lot of noise”: news cycles around the geopolitical situation will keep gold prices oscillating and consolidating; but the medium-term logic still points to new highs. UBS also lowered its 2026 average annual gold price forecast to $5,000 (from $5,200 previously, mainly due to accounting adjustments for the first quarter), while maintaining its year-end target price of $5,600 (set at the end of January).

The evidence for treating “central banks selling gold” as the main cause of this pullback is not solid. The 800–850 tons look more like a “slowing of the pace.”

The market’s worried scenario is very specific: if the Middle East conflict persists, oil prices push up inflation, growth weakens, and the local currency depreciates, some central banks may be forced to sell gold to cope with pressure. The report does not deny that “individual central banks may sell,” but it emphasizes that this does not equal an official-sector trend reversal.

One key reminder the report offers is: during the past 15 years of official sector entities continuously increasing gold holdings, it is not unusual for “selling” to appear within a single month. The reason may also be quite practical—central banks that bought cheaply early on take some tactical profit-taking outside their core holdings; a surge in gold prices triggers rebalancing; and “natural inflows” into gold-producing countries at certain points convert into overseas shipments. In other words, selling can be an action, not necessarily a stance.

The baseline view is still net buying, but at a slower pace. The detail here lies in the trading habits of official sector entities: they behave more like “physical buyers,” providing a stabilizing force during pullbacks that helps the market settle faster at higher levels. By contrast, official sector entities typically do not chase rallies; they tend to step in when the price is more favorable and volatility is more contained.

This also explains why, when volatility rises, the market suddenly feels that “the central banks are gone.” The observation mentioned in the research is that recently, official sector entities and other longer-term holders are more inclined to wait and watch rather than add to positions immediately during every decline.

The narrative around Turkey’s “50 tons” sell-off has been amplified, and gold prices in the short run are driven more by the dollar and real yields

Turkey’s case is sensitive because it appears to fit the narrative that “central banks have started selling gold.” But Turkey has some special features: some of the change may be swaps rather than direct selling; more importantly, Turkey’s central bank has long treated gold as a policy tool to support domestic banking system liquidity management.

In the total gold amount disclosed by Turkey’s central bank, a portion corresponds to commercial bank positions. When combined with policies after 2017 that allowed banks and other entities to use gold more within the financial system, the “change in total data” does not necessarily mean that the central bank is selling gold to the market. The report’s recommendation is very clear: wait until more detailed data disclosures that can be broken down by definitions are available, and then discuss the trend.

The trading environment in March featured “double uncertainty”: on the one hand, when Iran-related news gained momentum, gold prices were already swinging sharply after the surge-and-pullback in January and February and were searching for a new stable range. On the other hand, the impact of the Middle East conflict on macro conditions and asset pricing is not linear, and long-term funds are unwilling to make bets easily.

When strategic funds that buy on dips are absent, gold is more likely to revert to a traditional framework in the short term: a stronger dollar and rising U.S. real interest rates put downward pressure on gold prices; long positions are further squeezed out, and even some short-selling pressure appears. In addition, Chinese demand during this phase provides support on the downside; after gold stabilized around the $4,500 level, it returned to trading around the $4,700 area.

The underlying logic of central banks holding gold: buy and don’t sell

The World Bank’s “Fifth Biennial Central Bank Reserve Management Survey (2025)” explains a more fundamental question: what do central banks really think about gold. The survey covers holdings as of December 2024; participation from 136 institutions is the highest in the history of the series, and for the first time, it includes a standalone chapter on gold.

A few numbers can clarify the boundaries of central bank behavior: about 47% determine their gold holdings based on “historical legacy,” and about 26% based on qualitative judgments; only about a quarter include gold in a formal strategic asset allocation framework.

More importantly, only about 4.5% make short-term tactical adjustments to gold reserves, and their gold investment style is mainly buy-and-hold (about 62%). This profile suggests that even if the pace of buying slows down, official sector entities are not a group of traders who get driven by news and frequently rotate positions.

On the reasons for adding holdings, more than half list “diversification” as the most important cause; local gold purchase programs account for about 35%, and geopolitical risk for about 32%; reasons related to “liquidity needs” account for only about 6%. Official sector entities’ rationale for gold has not become invalid due to recent price volatility.

Short-term fluctuations are inevitable, but “new highs not finished” remains the main theme

Returning to the trading level, gold is not a straight-line path of rising. In the coming weeks, it may continue to consolidate and see ups and downs as the market repeatedly reassesses geopolitical risk. But it believes that the two lines driving gold allocation in the medium to long term—combined risk from growth and inflation, and the persistence of geopolitical tension—are making “allocation into gold” a more common portfolio action.

Within this framework, the report’s pricing anchor is: average annual gold price of $5,000 in 2026 and a year-end target of $5,600. It also notes that speculative positions have become “cleaner,” while long-term participants still have a relatively low allocation. If pullbacks appear again, it is closer to a “strategic entry window” rather than a signal that the trend has ended.