There is a version of the 10-year Treasury yield that trades in the metals market, and I find it slightly ridiculous how few equity and crypto people ever look at it. Divide the price of copper by the price of gold, plot the result against the 10-year, and the two lines track each other closely enough over long stretches that you would assume someone is arbitraging them together. As far as I know nobody is, the relationship just falls out of what each metal is for.
The mechanism is simple enough to explain over a beer. Copper goes into buildings, power grids, cars, machinery, wiring. When the world expects more construction and more manufacturing, copper gets bid, and the price responds quickly because inventories are thin relative to consumption. Gold does roughly the opposite job. It is what people hold when they distrust growth, distrust currencies, or want insurance against everything going sideways at once. So when you divide copper by gold, you cancel out a lot of what moves both of them, the dollar, broad commodity swings, general liquidity, and what remains is a fairly clean gauge of growth appetite versus fear. The 10-year yield is pricing a similar blend of growth and inflation expectations, which is why the two series rhyme. Jeffrey Gundlach spent years putting this overlay in his webcasts and treating the ratio as a rough fair-value guide for the 10-year, and the idea stuck because the chart really is uncanny across long periods.
Where the ratio has actually led
What makes this useful rather than just cute is that the ratio has sometimes turned before yields did. In early 2016, after the commodity bust and the China devaluation scare, copper against gold based out and started climbing months before the 10-year broke out of its range later that year. In March 2020 the ratio bottomed within days of the equity low and then ground higher for the rest of the year while the 10-year sat pinned near record lows. Anyone watching it had a growing pile of evidence that the recovery trade was real well before the bond market repriced in early 2021. Copper was sniffing out the stimulus and the reopening while Treasuries were still trading the emergency.
The 2008 episode ran the other direction just as cleanly. The ratio fell apart as the financial crisis spread, and yields collapsed along with it, the same signal with the opposite sign. In both directions the logic is identical. The metals market is a real-time vote on physical demand, and physical demand tends to move before official data, and sometimes before rates traders are willing to commit.
How I build and read it
Building it takes about a minute. Take front-month COMEX copper in dollars per pound, divide by gold in dollars per ounce. The raw number is tiny, a small fraction of one, so multiply by a thousand or index it to 100 at some starting date if the decimals annoy you. The units are arbitrary and only the shape matters. Weekly closes are plenty, since daily data mostly adds noise from inventory reports and options expiries. Any charting platform will plot one symbol divided by another, and both series are available free from FRED or exchange sites if you want the thing in a spreadsheet.
Reading it, I keep the process deliberately boring.
- Trend first. Is the ratio above or below its 40-week average, and has it been moving in one direction for more than a few weeks. A single week of movement means very little, copper is jumpy.
- Compare it to the 10-year. If the ratio and yields agree, the growth picture is coherent and I do not learn much, because the information lives in the gaps between them.
- Compare it to whatever risk asset I actually hold. If equities or crypto are pushing highs while the ratio is rolling over, I treat that as a reason to tighten risk or stop adding, and never as a reason to get short. Divergences can run for months.
- Ask why yields are moving before trusting any of it. This is the step most people skip and the one that decides whether the signal means anything.
For crypto specifically the ratio earns its spot as a confirmation input. Bitcoin trades like a leveraged risk asset in most regimes, whatever the narrative of the month says. A rising copper to gold ratio alongside rising yields usually means yields are going up for growth reasons, which risk assets can typically absorb. Rising yields with a falling ratio is the ugly mix, tightening or inflation fear without growth behind it, and that combination has historically been rough for tech equities and crypto alike. I would rather know which of those two worlds I am in before sizing anything.
The regimes where it stops working
This part matters more than the party trick, because the ratio has failed in well-defined ways and will fail the same ways again. The clearest breakdown was 2021 into 2022. Yields exploded higher as inflation forced the Fed into an aggressive hiking cycle, and copper against gold went sideways and then down. The bond market was pricing policy and inflation, the ratio was pricing real growth, and both were right about their own subject. If you had used the ratio as a fair-value model for the 10-year through that stretch, you would have concluded yields were wildly overshooting, and you would have been steamrolled buying duration on that logic. When yields move because of central bank policy rather than growth expectations, the linkage loosens badly.
Then there are distortions in each leg. Copper demand is heavily Chinese, roughly half of the global total historically, so a stimulus push or property bust in China moves the ratio even when nothing has changed in the rest of the world. Supply matters too, since a major mine strike or smelter outage lifts copper for reasons that have nothing to do with demand. On the other side, gold sometimes gets bid for reasons unrelated to fear in Western markets, with sustained central bank accumulation being the obvious one, and that suppresses the ratio while risk appetite is perfectly fine. And when a central bank is pinning long rates through QE or yield curve control, the ratio can push higher while the 10-year is simply not allowed to respond. The gap is still information about where yields might go if the pin comes off, but it is useless as a timer.
So my rule of thumb ends up being a filter. Use the ratio when the question is about growth, ignore it when the question is about policy, and check the copper-specific news before reading anything into a sharp move. In practice this is a five-minute weekly check. Pull the ratio, glance at the 40-week trend, note whether it agrees with yields and with whatever I am holding, and write one line about it. Most weeks the line says nothing changed. The handful of weeks a year where the ratio quietly stops confirming everything else are the ones that have paid for all the boring ones.