- Published on
The Curious Case of Gold - Part 4: The Drawdown, and the Mechanism I Had Backwards
- Authors
Part 3 was written on 7 March with gold at $5,172, silver at $80.62, and a gold-to-silver ratio of 62.8. I laid out a bubble case and a repricing case, said I would give my honest view, and asked what would have to be true for gold to reach $6,000.
It did not reach $6,000. As I write this on 3 November, gold sits at roughly $3,962 and silver near $52, with the ratio blown out to 78.8.
That is a drawdown of about 23% in gold and roughly 35% in silver from where Part 3 was written. The bubble case had more in it than I credited, and silver was the more dangerous of the two holdings - which Part 3 did at least flag. But the price call is the less interesting error. The more interesting one is that I had the mechanism backwards, and the Gulf crisis is what exposed it.
Table of contents
- Marking Part 3
- The week gold fell on a war
- What gold actually is
- Silver did the thing silver does
- What I got right, narrowly
- What I am watching now
Marking Part 3
| 7 March 2026 | 3 November 2026 | Change | |
|---|---|---|---|
| Gold | $5,172 | ~$3,962 | -23% |
| Silver | $80.62 | ~$52 | -35% |
| Gold-silver ratio | 62.8 | 78.8 | +16 points |
Part 3 framed this as bubble versus repricing and treated them as competing explanations. Looking at that table, I think the framing itself was the problem. Both were partly true, and they operate on different clocks: a speculative premium that unwound over eight months, sitting on top of a structural official-sector bid that has not gone anywhere. Asking which one was correct produced a less useful answer than asking which one dominates over what horizon.
The week gold fell on a war
Here is the part that genuinely surprised me.
On 30 October, explosive devices struck a tanker in the Strait of Hormuz, part of the sustained campaign against commercial shipping that has slowed traffic through the strait to a trickle. Oil rose. And gold fell, dropping below $4,000, with silver tumbling alongside it.
If you hold the standard retail model of gold - a safe haven, a geopolitical insurance policy, the thing you own for exactly this scenario - that price action makes no sense. A shooting war threatening a fifth of the world's seaborne oil is close to the platonic case for the hedge, and the hedge went down.
Then on 2 November, as Iranian diplomacy progressed and oil eased back, gold and silver rebounded.
So the sequence was: war escalates, gold falls. War de-escalates, gold rises. That is the exact inverse of the safe-haven story, repeated twice inside a week, which makes it a mechanism rather than noise.
What gold actually is
The transmission runs through the Federal Reserve, not through fear.
An oil shock raises headline inflation expectations. Higher inflation expectations raise the expected path of policy rates, because the central bank is expected to respond. A higher expected path of nominal rates, with inflation expectations already in the price, raises expected real rates. And gold is, mechanically, a long-duration asset with no cash flow - its price is extremely sensitive to real rates, because the opportunity cost of holding a non-yielding asset is precisely the real yield available elsewhere.
So the chain is: Hormuz attack, oil up, inflation up, Fed tightening expectations up, real rates up, gold down. Geopolitical risk entered the gold price as a monetary variable with a negative sign, not as a fear premium with a positive one. The reversal on 2 November ran the same chain backwards: diplomacy, oil down, tightening concerns ease, gold up.
This is not a new insight in the literature, and I want to be clear that I am not claiming to have discovered it. But Part 3 was written in a register that implicitly treated gold as geopolitical insurance, and that register was wrong in a way that would have cost money. Gold is a real-rates asset that sometimes behaves like a crisis hedge - specifically when the crisis is disinflationary, or when it threatens the monetary system itself rather than the oil supply. An inflationary supply shock is close to the worst case for it.
The practical version: if you own gold as protection against war, check whether the war you are imagining pushes inflation up or down. Those are different trades and gold only wins one of them.
Silver did the thing silver does
Silver fell 35% against gold's 23%, and the ratio widened from 62.8 to 78.8.
Part 3 noted the ratio at 62.8 with silver lagging, and read that as the speculative premium from January not having fully returned. In hindsight that was the correct observation attached to the wrong conclusion - I treated the lag as room to catch up rather than as the early signal of a market losing its appetite for the more speculative of the two metals.
The ratio above 80 is conventionally read as silver being cheap relative to gold, and at 78.8 we are nearly there. I would treat that signal with some suspicion. Silver's industrial demand component means the ratio can stay elevated through a soft patch in manufacturing, and "historically cheap" has been a poor timing tool in both directions.
What I got right, narrowly
Two things from Part 3 survive.
The structural official-sector bid was real and has not reversed. Central bank accumulation has continued through the drawdown, which is the clearest evidence that the repricing case was not pure narrative - a price can fall 23% while the long-horizon buyer keeps buying, and those two facts are about different participants with different time horizons.
And the parabola observation held up. Part 3 argued the January crash had revealed how fragile the rally's structure had become, even as the price recovered. That fragility is what the last eight months were. The recovery above $5,000 in March was the rally's last strong move, not the resumption Part 3 half-expected.
What I am watching now
Three things, in order of how much they would change my view.
The real rate path. Given the mechanism above, this is the variable that matters most and the one I underweighted. If the Gulf situation keeps oil elevated and the Fed stays tight in response, the headwind persists regardless of how frightening the headlines are.
Whether official buying slows. The structural bid is the floor under the repricing case. If central bank accumulation decelerates while price is falling, the floor is softer than it looks.
The ratio above 80. Not as a buy signal, but as a read on whether this is a precious-metals drawdown or an industrial-demand story wearing one. Gold holding while silver keeps sliding would say the latter.
Part 3 asked whether $5,000 gold was a bubble or a repricing. Eight months on, the answer appears to be that it was a repricing with a bubble on top, and the bubble came off first. The part I would most want back is not the price target. It is the sentence where I treated gold as the thing you own when the world gets dangerous, without specifying which kind of dangerous.
Sources and further reading:
- Kitco - gold drops below $4,000, silver tumbles as Hormuz attack lifts oil prices
- Kitco - gold, silver rebound as Iran diplomacy sends oil lower, easing Fed tightening concerns
- USAGold - daily precious metals market report, 3 November 2026
- OilPrice - oil price forecasts jump as Hormuz disruption drags on