The Short Answer
Gold fixed at $5,405 in London on 29 January 2026, the highest price in its history. Its low since then was $3,994 on 16 July 2026, a 26% correction, and on 4 September 2026 it fixed at $4,415, 18% below the peak and 25% above where it was a year earlier ($3,546 on 4 September 2025). The record was the end of a run that saw gold nearly double in sixteen months. What followed is the largest correction since 2013, but so far a correction, not a reversal.
Prices are the LBMA PM fix. The figures and table on this page are regenerated from the LBMA series after every fix, so they are current as of 4 September 2026; the live chart is on the gold price page.
The Path
| Date | Gold (LBMA PM) | Change from peak |
|---|---|---|
| Aug 2025 | $3,429 | — |
| Sep 2025 | $3,825 | — |
| Oct 2025 | $4,012 | — |
| Nov 2025 | $4,191 | — |
| 23 December 2025 (2025 high) | $4,449 | — |
| Dec 2025 | $4,368 | — |
| 29 January 2026 (all-time high) | $5,405 | — |
| Jan 2026 | $4,982 | −8% |
| Feb 2026 | $5,222 | −3% |
| Mar 2026 | $4,608 | −15% |
| Apr 2026 | $4,611 | −15% |
| May 2026 | $4,546 | −16% |
| Jun 2026 | $4,026 | −26% |
| Jul 2026 | $4,027 | −26% |
| Aug 2026 | $4,563 | −16% |
| 4 September 2026 | $4,415 | −18% |
The shape matters. Gold spent February above $5,000, then fell in two legs, to $4,600 by March and to $3,994 by 16 July 2026, and has since recovered about 30% of the loss. Silver, which peaked the same day, fell roughly twice as far; the silver correction is covered separately.
Why Gold Rose to $5,405
The drivers of the 2024–2026 bull market are well documented on this site and did not change in January. Central banks bought gold at a record pace for a fourth year, led by emerging-market reserve managers diversifying away from the dollar; the central bank gold buying analysis and the reserves by country data track that. Falling real interest rates lowered the cost of holding a non-yielding asset. ETF inflows returned after two years of outflows. And in the final months, the same momentum and retail participation that sent silver parabolic pushed gold from $4,000 to $5,405 in four months, far faster than the underlying buying justified.
Why It Fell
Three forces, in roughly this order. First, positioning: by late January, speculative long positions in futures and ETF holdings were at extremes, and a market that has already bought has nothing left to buy with. Second, the macro backdrop stopped helping: gold is most sensitive to real yields and the dollar, and as expectations for rate cuts were pushed out during the spring, the opportunity cost of holding gold rose. Third, silver’s collapse dragged sentiment across the whole complex; when the most speculative metal falls by half, gold holders sell too.
What did not happen is a change in the structural buyers. Central bank purchases continued through the first half of 2026 at a pace consistent with recent years, and the buying that put the floor under gold in every dip since 2022 remains in place. That is the main difference between this correction and 2013, when central banks were still net sellers in some quarters and gold fell 28% in a year and kept falling.
Is This 2013 or 2008?
In 2008, gold fell 30% in eight months during the financial crisis, then made a new high within a year. In 2013, it fell 28% and did not recover its high for seven years. The difference was what happened to the buyers: in 2008 the drivers of demand (negative real rates, monetary expansion) intensified after the fall; in 2013 they reversed as the Fed began tapering.
Today’s setup has elements of both. Real yields are the risk. Central bank demand and fiscal deficits are the support. The inflation-adjusted gold chart shows that even at $4,415 gold sits well above its 1980 and 2011 peaks in real terms, which argues for caution about how much of the 2025 rally was durable. The gold vs S&P 500 comparison shows gold has still outperformed equities over one, three and five years.
What It Means for Holders
- Physical buyers are being paid to wait. Dealer premiums over spot on Gold Eagles and bars have compressed from the January highs. A lower spot price and a lower premium compound.
- Do not confuse the record with the trend. A 26% correction after a near-doubling is normal bull-market behaviour; gold corrected 20% or more three times between 2001 and 2011 on its way from $250 to $1,900.
- The ratio trade has closed. With the gold-silver ratio back at 66, there is no longer an obvious case for favouring silver over gold on relative value.
- Rebalance, don’t react. If gold’s rally pushed it past your target allocation, the correction did some of the rebalancing for you. The portfolio allocation guide covers how to size it.
We do not forecast prices. What we can say from the data is that the structural demand that built the bull market is intact, the speculative layer on top of it has been removed, and the price is now where it was in the autumn of 2025, before the final surge.