Gold’s $45 Slide Shows Why Central Banks Need a Liquidity Haircut

Gold’s latest retreat is exposing a fault line in the reserve-buying boom: bullion may be a strategic asset, but it is not cash.

Samarium’s gold-price table shows the metal fell $45.10, or 1.10%, over the 24 hours to 07:59 UTC on July 24, sliding from $4,093.56 to $4,048.46. It touched $4,022.80 before recovering part of the loss, while the session high was the opening $4,093.56.

The immediate trigger was familiar but unusually sharp. Brent crude moved above $100 a barrel after renewed attacks on Saudi tankers, lifting inflation expectations and pushing bond yields higher. Reuters reported that traders were assigning an 81% probability to a Federal Reserve rate increase by September. Gold’s haven appeal was therefore offset by the rising opportunity cost of holding a non-yielding asset.

The more consequential development came earlier this month from an IMF Note on reserve management. The authors argued that gold should not be treated as part of a central bank’s liquidity tranche because its price can fall when reserves are needed most. They recommended applying a 6% to 12% haircut to gold’s market value when calculating effective reserve liquidity, depending on market stress.

That is not a theoretical adjustment at today’s prices. A 10% haircut on $100 billion of bullion removes $10 billion from the amount a reserve manager should count as readily deployable. The distinction matters because rising gold prices have mechanically inflated the reported value of official reserves even when central banks have not bought additional tonnes.

The euro area illustrates the scale of the valuation effect. Eurosystem data valued its gold at about €1.23 trillion at the end of June, compared with €1.38 trillion at the end of April, while physical holdings were virtually unchanged at roughly 347.5 million fine troy ounces. Price movement, not reserve transactions, drove most of the change.

That does not invalidate the strategic case for gold. The World Gold Council says central banks bought an average of about 1,000 tonnes annually over the past four years, double the average of the previous decade. Its 2026 survey found 89% of reserve managers expect global official holdings to rise over the next year, while 45% expect their own institution to add gold.

Official buying also accelerated in May, with net purchases of 41 tonnes led by Poland and China. The case remains straightforward: gold carries no sovereign credit risk, cannot be frozen by an issuer and may diversify reserves during geopolitical stress.

But the IMF Note forces a more disciplined question. Gold can strengthen a central bank’s long-term balance sheet while weakening its short-term liquidity profile if its market value is counted at face value. That is especially relevant for emerging economies with large external financing needs, concentrated reserve portfolios or domestic gold-purchase programmes.

The market itself is liquid, but liquidity is not the same as price certainty. LBMA data show weekly gold turnover running near $961 billion on a 12-week moving average, and the World Gold Council estimates global gold trading averaged a record $488 billion a day in the first half. Central banks can sell bullion. The risk is how much value they can realise quickly during stressed conditions.

The current sell-off is modest relative to the IMF’s proposed haircut, but it demonstrates the mechanism. Gold dropped more than 1% in a day even as geopolitical tensions intensified. If reserve managers need dollars during the same shock that pushes yields higher, bullion can deliver less liquidity than its headline market value suggests.

The next catalyst is the Federal Reserve’s policy meeting. A further repricing toward rate increases could pressure gold below $4,000, while any easing in the oil shock could reverse the move. For central banks, however, the longer-term issue is no longer simply whether to own more gold. It is how conservatively to value it when reserves are meant to be available on demand.