Gold's Worst Run in Nearly 40 Years. Has Anything Actually Changed?

Jimmy Wong

Gold finished the week near US$3,890 an ounce, down more than 4% — its worst week in four months. Step back and it is the 10th weekly fall in 12 weeks, a losing run gold has not had since 1989. It now sits about 7.6% below where it was a year ago, and silver had it worse, dropping to about US$48.65, a one-year low.

One number tells a different story. In Australian dollars, gold is around A$5,740 an ounce, almost exactly where it sat three weeks ago. The Australian dollar fell too, and the two moves cancelled out.

Last week gold rose 5%. This week it handed nearly all of it back, and the swing came down mostly to one thing: oil.

On Monday the White House said renewed strikes on Iran were "highly unlikely" if talks progressed. Oil slid, inflation worries eased, the US dollar fell between 0.3% and 0.6%, and gold rose 1.35% to about US$4,126. By Thursday those peace hopes had faded, oil climbed again, and gold fell 3.05% to its lowest level since October 2025 — the same question answered the opposite way, four days apart. Higher oil is not simply good for gold, as the shortcut version has it. Oil feeds expected inflation, expected inflation feeds expected interest rates, and it is rates that gold answers to.

Tuesday did the rest of the damage. Core prices, which leave out food and energy, rose 0.3% against 0.2% expected — a third month in a row at that pace. Hours earlier, speaking in London, Fed Governor Christopher Waller had set exactly that test: core at around 0.3% or higher for a third straight month, and a half-point December rise "should be considered". The score came back at the number, and odds of a half-point move went from nil to about 14% in a day. Because gold pays no interest, what matters most is what bonds pay after inflation, and that figure climbed from 2.63% on 6 November to 2.86% by 12 November. The 10-year yield briefly topped 5.00%, and the US dollar finished near its highest level of 2026. Each of those makes holding something that pays nothing more expensive.

Friday brought a small bounce, but only after short-term indicators had hit extremely oversold levels — that is positioning adjusting, not minds changing.

Is the long-term reason to own gold still there?

The answer starts with being plain about why we hold gold at all. It is an arithmetic problem.

The US government owes a great deal, and at today's interest rates the cost of servicing that debt takes a large and growing bite out of what it collects in tax. Here is the trap. Raising rates to fight inflation makes that bill bigger. A bigger bill widens the deficit. A wider deficit means issuing more debt, much of it short-term, which reprices at the new higher rates within a year or two. Higher rates are meant to be the cure. For the budget, they make the patient sicker.

That cannot run forever. It ends one of three ways: the central bank ends up funding the government, rates are held below inflation so the debt quietly shrinks in real terms, or something more abrupt happens. Each of those leaves a dollar buying less than it did. Gold's attraction is simply that it is nobody's promise — it cannot be printed, and it cannot default.

Which makes this week faintly ironic. Gold fell because traders expect the Fed to raise rates again in December. But every rise makes that arithmetic worse, not better. The market was trading the mechanism. The reason to own gold quietly got a little stronger.

Others are hedging the same risk. China's central bank added 25.8 tonnes in October, its 24th month in a row, lifting its holdings to about 2,435 tonnes. It bought more gold while the value of that pile fell, from US$323.5 billion to US$306.4 billion. It was not buying because the price was going up. Central banks together bought 321 tonnes in the September quarter, 42% more than a year earlier.

Then there is Singapore. This week its central bank opened the national vault to other countries' gold reserves, with room for 500 tonnes. It is also waiving tax on wholesale gold services such as vaulting, refining and broking from April 2027, building a system for settling gold trades between institutions, and targeting physically deliverable gold futures for the second quarter of 2027.

Taken together, that is not a vault. It is plumbing. Vaults, settlement, tax treatment and deliverable contracts are the machinery that lets large institutions hold and move an asset easily, and London and New York have had it for roughly a century. Asia is now building its own. Nobody lays that kind of pipework for an asset they expect to matter less over the next decade, and once it is built it tends to get used for decades. Singapore did try a gold contract back in 2014, and it failed to draw traders and was withdrawn, so the real test is whether volumes build through 2027.

So what would change our mind?

Not a 4% week, and not a quiet month for gold funds. The arithmetic itself would have to improve:

• Budget surpluses, or spending restraint that actually holds. This isn't fantasy — the US ran surpluses in the late 1990s.
• The economy growing faster than the interest bill for years, without inflation doing the work. If the AI investment boom delivers a genuine productivity jump, the problem eases on its own.
• Real yields staying high for years while gold keeps falling — which would tell us the market has chosen bonds over gold as its hedge, whatever we make of the risk.
• Official buying drying up and staying dry. It isn't our reason for holding gold, but the countries with the most dollars to lose are hedging the same risk. If they stopped, we would want to know why.

Being right about the direction tells you nothing about the timing. Japan has carried debt above 200% of its economy for decades without the reckoning its critics kept predicting. This view can be correct and still ask years of patience, with weeks like this one along the way. That is what insurance costs.

Is this a good time to add?

It could get worse before it gets better. A quarter-point December rise now sits near 89%, a bigger one is still on the table, and if oil keeps climbing the pressure can run for weeks. Owning gold through this means accepting some real short-term pain. The US$4,000 line is gone, and Thursday's low near US$3,860 is what short-term traders are watching now, with support generally put around US$3,850. For anyone holding gold for years rather than weeks, though, one week at these levels settles nothing. When the reasons to own it have not changed, buying in stages matters more than picking the bottom. The nearer-term tests are the Fed's October meeting minutes and US retail sales next week, then the decision on 8–9 December.

And our miners?

The ASX gold index fell 3.59% on Thursday but still finished the week up 0.68%, while gold fell more than 4%. The currency explains most of that: miners sell their gold in Australian dollars, and that price hardly moved. The thing to watch is silver, which hit a one-year low and fell harder than gold. ASX gold shares have tended to track silver more closely than gold, because both depend on investors being willing to take risk. Until silver steadies, rallies in the smaller miners will struggle to hold.

If you like our content, and want to learn more about our detailed analysis into gold mining companies, we provide additional content to our members. It is totally free to be our member, just sign up to our service, and we will share it with you through our newsletters.