Report: Gold — The Fall Began Before the War: What Actually Moved Gold, and Where It Goes Next
Gold’s 2026 round trip is usually blamed on the Middle East conflict that began on 28 February, but the timing does not support that reading. On the LBMA Gold Price PM benchmark, gold peaked at $5,405.00 on 29 January and bottomed at $4,001.80 on 25 June — a 26.0% drawdown — before recovering 14.0% to $4,562.75 by 28 August, leaving it 15.6% below the January fix. The single largest move of the whole decline came a month before the first shot: a 7.8% drop at the 30 January fixing, which alone accounts for 27.1% of the fall measured in logarithmic returns.
Gold then rallied through February. With 28 February falling on a Saturday, the last fixing before the conflict was the month’s close of $5,222 — just 3.4% below the all-time high, with only 11.5% of the eventual drawdown yet realised. The war did not begin with gold collapsing; it began with gold within touching distance of its record. The bulk of the decline came afterwards, and for reasons that were not primarily geopolitical: the largest block, 46.9% of the drawdown, fell between April and the June trough, after the acute phase had passed, as US 10-year real yields rose.
The report’s central argument is that war is not a gold factor in its own right but a transmission mechanism — and in 2026 it transmitted the wrong way. Rather than the disinflationary shock that strengthens gold’s safe-haven appeal, the conflict delivered an energy and inflation shock: flows through the Strait of Hormuz fell below 10% of pre-conflict levels, and the International Energy Agency released 400 million barrels from emergency stocks, its largest coordinated action ever. That kept policy restrictive — the Federal Reserve held at 3.50–3.75% on 29 July, with three members preferring a hike — and higher real yields raise the cost of holding an asset that yields nothing.
Quantifying that channel gives the report its key diagnostic. Between end-March and end-July, gold fell 12.6% while 10-year real yields rose 47 basis points — a sensitivity of about 2.7% of price per 10 basis points. Applied to August, when real yields fell just 5 basis points, that relationship would predict a 1.3% gain; gold rose 13.3%. Roughly 12 percentage points of the August recovery are therefore unexplained by real yields, resting instead on a weaker dollar, the US Treasury’s enlarged buyback operations, renewed geopolitical uncertainty and a technical break above the 200-day moving average.
The structural bid from central banks is real but slower than the market often assumes. The World Gold Council estimates official buying of 288.9 tonnes in the second quarter, but revised its first-quarter figure down sharply, from 244 tonnes to 56.5 tonnes. Official demand builds a floor over years; it did not prevent a 26.0% drawdown in five months, and it is reported too slowly to serve as a real-time signal.
Looking ahead, The Edge’s scenario distribution to end-2026 centres on $4,400–5,000 at 50% probability, with a bull case of $5,200–5,800 at 30% and a bear case of $3,700–4,200 at 20% — a judgement-weighted midpoint near $4,790. No single variable decides the outcome; the operative signal is the interaction of real yields and the dollar, confirmed by investment flows. The report also stresses a discipline often lost in commentary: gold has no single price, and any drawdown figure is meaningless without naming its basis — three published benchmarks recorded three different 2026 peaks on the same day.
The full report sets out the phase-by-phase decomposition, the benchmark and grading of every figure, and the evidence behind each scenario, drawing on data from the World Gold Council, the LBMA, the US Treasury, the IEA and the Federal Reserve.
Download the full report
⬇ Download the report in English
⬇ تحميل التقرير بالعربية
