Recently, the market has been asking: Have global central banks started selling gold? Has the trend of desperate "hoarding" of countries in the past 15 years come to an end?
In response, UBS precious metals strategist Joni Teves clearly responded in his latest report on April 2: a large-scale sell-off is almost impossible. The central bank as a whole will continue to buy gold net, but at a slower pace.
Actual gold purchases in 2025: about 860 tons; Estimated gold purchases in 2026: 800 to 850 tons. In other words, I bought a little less, but I didn't stop, let alone sell.
So why are there rumors of a "central bank sell-off"? The trigger was Turkey – some news said it "sold about 50 tons of gold in a few weeks." However, UBS pointed out that the "official gold data" released by Turkey is actually mixed with the operations of commercial banks, such as gold swaps, short-term transactions, etc., and does not represent the true intentions of the central bank.
Using such fuzzy data to assert a "global central bank turn" is very risky. It's best to wait for clearer breakdown data before drawing conclusions.
"The central bank is selling gold"? Don't jump to conclusions
The recent correction in gold prices has attributed the reason to "the central bank has started selling gold". But this claim is not well documented. The UBS report pointed out that even if global central bank gold purchases fall to 800–850 tons in 2026 (slightly lower than 860 tons in 2025), it will only slow down, not turn around and sell.
What is the real concern of the market?
If the conflict in the Middle East drags on for a long time, oil prices soar, inflation rebounds, the economy deteriorates, and the local currency depreciates, some countries may be forced to sell gold for foreign exchange emergencies.
The report admits that individual central banks may indeed sell, but this does not mean that the entire official sector has turned. Just like in the past 15 years, the central bank has been buying gold in general, but it is normal to sell occasionally in a single month.
The reasons include: countries that bought at low prices in the early years did some "pocket safety" outside of their core positions; After the price of gold rose sharply, it reduced its holdings slightly in order to balance asset allocation; Gold-producing countries (such as South Africa, Russia) sometimes export new gold directly, which looks like a "sell", but is actually a natural flow.
The key point is that selling is not equal to a policy shift.
Moreover, "official buyers" like central banks have a characteristic: they do not like to chase higher, and rarely rush in when gold prices skyrocket; likes to buy while the decline falls, especially when the market panics and prices fall, to provide "support"; But they don't take action immediately every time they fall - if the volatility is too great, they will wait and see and wait until the situation is more stable and the price is more suitable before acting.
Therefore, the recent market feels that "the central bank is gone", probably just that they are waiting for a better time, rather than not buying.
Turkey sells 50 tons of gold" is over-interpreted
Recently, the market has been talking about Turkey "selling about 50 tons of gold in a few weeks", which seems to confirm the statement that "the central bank has started to sell". But it's not that simple.
Turkey is very special: its "official total gold" actually includes gold held by commercial banks; Since 2017, Turkey has allowed banks to use gold for collateral, lending, liquidity management and other operations; Therefore, data changes may not necessarily mean that the central bank is really selling gold in the market, but may be internal account adjustments, gold swaps or circulation within the banking system.
UBS clearly advises: Don't rush to conclusions, wait for more detailed data that can distinguish "central bank vs. bank" positions come out.
Why has the price of gold fallen recently?
The real reason is here: the market faced "double uncertainty" in March: the situation in the Middle East was repeated (such as news related to Iran), but the price of gold has risen and fallen sharply in January-February, and is in the stage of finding direction; The impact of geopolitical conflicts on the economy and assets is complex and unpredictable, and long-term funds do not dare to act easily.
The result is that the big buyers (such as central banks and sovereign funds) who usually "buy on dips" are temporarily absent.
The price of gold has no support force, so it returns to the traditional logic: the US dollar strengthens, gold becomes more expensive (for non-US dollar buyers), which is bearish; Real interest rates in the United States have risen, and the opportunity cost of holding gold has become higher, and more people are selling.
This led to the bulls being "washed" and even short-selling forces.
However, Chinese demand acted as a buffer. The price of gold once fell to around $4,500 and then stabilized, and recently returned to around $4,700.
The central bank buys gold, not for speculation, but for "pressing the bottom of the box"
The World Bank's latest Reserve Management Survey 2025 (covering 136 central banks, the most in history) specifically analyzes how central banks view gold for the first time. The results are clear: for most central banks, gold is bought and basically not sold.
Several key data illustrate the problem: 47% of central banks hold gold because it "has been left like this in history" and has no intention of moving; 26% feel that they should hold some based on experience or judgment; Only about 25% formally include gold in their asset allocation model; and only 4.5% can really do short-term trading operations.
What's even more illustrative is that 62% of central banks regard gold as an asset to "buy and hold for a long time" - they are not traders and will not adjust their positions as soon as the news comes out.
So why increase your holdings?
More than half of the central banks said: mainly to diversify risks (don't put eggs in one basket); about 35% is because of domestic gold, which is bought on the spot to support the local market; About 32% mentioned geopolitical tensions (e.g., war, sanctions); Only 6% use gold as "emergency cash".
In other words, the recent fluctuations in gold prices have not shaken the fundamental reason for central banks to hold gold.
What does it mean for investors?
Short-term gold prices will definitely fluctuate - geopolitical news comes and goes, market sentiment is repeated, and prices are inevitably bumpy.
However, in the medium and long term, the two major trends have not changed: the combination of "high inflation + weak growth" in the global economy is unsettling; Geopolitical conflicts (the Middle East, Russia and Ukraine, etc.) may be prolonged.
This will lead more and more institutions to treat gold as a standard safe-haven asset rather than a dispensable option.
UBS maintained its forecast accordingly: the average price for the whole year of 2026: $5,000 per ounce; Year-end target price: $5,600 per ounce. And at present, speculative funds have been significantly reduced ("clean positions"), while long-term investors are still underweighted in gold. Therefore, if the price of gold pulls back again, it is more of an opportunity to buy strategically than a signal of the end of the bull market.
For Physical/ETF Investors:
No need to panic sell: the central bank's bottom position is stable, and systematic selling pressure does not exist; It can be deployed in batches when there is a pullback: if the price of gold steps back to the range of $4,500-4,600, the cost performance will be significantly improved; Focus on gold ETFs with good liquidity such as GLD and IAU, or the main contracts of Shanghai Gold and COMEX.
For macro traders:
Short-term gold prices are highly negatively correlated with the US dollar index and 10-year TIPS yields; If the Fed starts cutting interest rates in June, or the situation in the Middle East escalates again, the price of gold may quickly break through $5,000.
