The nickname "Dr. Copper" has been around for decades. The idea is that copper, because of its pervasive role in real economic activity, effectively holds a PhD in economics, it anticipates turning points in growth before the official data arrives. Like most useful market aphorisms, it is partly true, frequently misapplied, and worth understanding carefully rather than taking on faith.
What makes copper different from most commodities is the sheer breadth of its demand base. Oil is heavily influenced by geopolitics, OPEC discipline, and the energy transition. Gold is a monetary asset as much as an industrial one. But copper, copper shows up in construction, manufacturing, power transmission, transportation, and increasingly in the electronics and electrification infrastructure driving the next decade of capital spending. It is the connective tissue of the physical economy.
That breadth is exactly what makes its price signal worth paying attention to. A commodity whose demand is concentrated in one sector will tell you about that sector. Copper tells you about most sectors at once.
What copper actually measures
Global copper demand breaks down roughly as follows: construction accounts for approximately 40% of end-use, driven by residential and commercial building activity, plumbing, wiring, and HVAC systems. Industrial machinery and manufacturing takes around 25%, factories need copper for motors, transformers, and production equipment. Electrical infrastructure, power grids, and transmission lines account for around 20%. Transport, vehicles, rail, shipping, takes the remaining 10% or so, a share that is rising as EV penetration increases.
Notice what is on that list: housing construction (the most interest-rate-sensitive sector in any economy), industrial production, power grid investment, and transport. When all four slow simultaneously, copper knows. It has no central bank. No quarterly earnings call. No CFO managing expectations. The price moves because physical buyers are either booking forward contracts or cancelling them.
Construction: ~40%, the most rate-sensitive sector in any economy
Industrial / manufacturing: ~25%, factory output, motors, transformers
Electrical / power grid: ~20%, transmission, infrastructure build-out
Transport: ~10% and rising, EVs require 3–4x the copper of an ICE vehicle
China share of global demand: ~55%. The copper price without context of Chinese PMI and fixed asset investment data is half a signal at best.
The China problem, and why it matters for interpretation
Any serious discussion of copper as a macro signal has to confront the China dependency. Approximately 55% of global copper demand comes from China, its property sector, its manufacturing exports, its grid expansion, and its state infrastructure programmes. This concentration creates a significant interpretive challenge.
A copper rally driven by Chinese fixed asset investment and infrastructure stimulus tells you something very different from a copper rally driven by a synchronised upturn in US and European manufacturing PMIs. The first is a policy-driven demand spike in a single economy with questionable data transparency. The second is genuine global industrial expansion.
This is why I always read copper prices alongside China's official PMI, the Caixin PMI (the private-sector-weighted alternative that is often more candid), and import data from Chinese customs. When Chinese copper imports are rising alongside rising LME prices, the demand signal is real. When LME prices are rising but Chinese imports are flat or declining, the price action may be driven by financial positioning rather than physical demand, and that is a much weaker signal.
The copper-gold ratio: the only ratio worth carrying in your head
In isolation, copper's price is interesting but noisy. The copper-to-gold ratio is considerably more informative. Here is the basic logic: copper rises when the world is investing in growth, building things, expanding capacity. Gold rises when the world is scared, when investors want a store of value that has no counterparty risk and no dependence on industrial activity.
When the copper-gold ratio is rising, it means copper is outperforming gold. Growth expectations are being revised upward. Risk appetite is expanding. This is a risk-on signal that has historically correlated well with rising equity markets, tightening credit spreads, and upward revisions to global growth forecasts.
When copper rises and gold falls, the market is pricing in a world that builds things. When gold rises and copper falls, the market is pricing in a world that hoards things. These are very different worlds.
When the copper-gold ratio is falling, gold outperforming copper, the signal reverses. Growth expectations are softening, safe-haven demand is rising, and the macro environment is turning cautious. The ratio fell sharply ahead of the 2008 financial crisis, well before equities peaked. It fell again in 2015 as Chinese growth fears mounted and EM economies corrected. In both cases, copper was sending a warning that equity markets had not yet priced in.
The divergence signal, when copper and equities disagree
The most actionable signal copper generates is when it diverges from equity markets. Equities can be driven by multiple expansion, share buybacks, rate-cut optimism, and narrative momentum that has little to do with actual economic activity. Copper cannot. It only moves when someone is buying it to use.
In the second half of 2007, the S&P 500 was still near all-time highs. Copper peaked in May 2007 and began a quiet decline through the summer and autumn. By October 2007, equities had also topped, but the divergence had been running for five months. Copper knew something about credit conditions and construction activity that equity prices had not yet acknowledged.
A similar pattern played out in 2015. US equities were grinding higher into mid-year records while copper fell from $2.90/lb in January to below $2.20/lb by August. The copper decline was tracking Chinese property slowdown and overcapacity in steel and aluminium. EM equities eventually corrected sharply. Copper was ahead of the equity market's acknowledgement by several months.
The rule of thumb: when copper sells off while equities make new highs, someone needs to be wrong. Historically, it tends to be the equities. Not always, sometimes copper overshoots to the downside and equities are right. But as a flag to raise your scrutiny, the divergence has a good track record.
LME warehouse inventories, the confirmation tool
When I am trying to determine whether a copper price move is driven by genuine physical demand or financial speculation, I look at LME (London Metal Exchange) inventory data. This is publicly available, updated daily, and tells you directly whether physical buyers are pulling copper off the exchange.
When LME copper inventories are declining, it means physical users, manufacturers, construction companies, cable producers, are drawing down exchange stocks to cover real consumption. This is a bullish confirmation of demand. When inventories are rising while prices are also rising, it means financial buyers are accumulating positions without the corresponding physical drawdown. That is a more fragile rally.
The inventory signal is not infallible. Chinese bonded warehouse stocks (which sit outside LME) can complicate the picture, and traders sometimes cycle metal between registered and unregistered storage in ways that distort the headline number. But as a quick cross-check on physical demand, it is one of the most transparent data points in commodity markets.
The structural complication: EVs and the electrification floor
There is a legitimate critique of using copper as a pure cyclical signal going forward, and it is worth taking seriously. The electrification megatrend, EVs, grid upgrades, data centre power infrastructure, renewable energy buildout, is creating a structural demand tailwind that did not exist in previous cycles.
A typical internal combustion engine vehicle contains roughly 20kg of copper. A battery electric vehicle contains 60-80kg. Data centres require enormous amounts of copper for power distribution and cooling. Grid modernisation programmes in the US, Europe, and India are capital-intensive in copper. This structural demand floor means copper may not fall as far in future cyclical downturns as it did in 2008 or 2015, because even when construction slows, electrification spending does not stop.
The practical implication: if copper falls less sharply in the next downturn than historical precedent would suggest, that does not mean the cyclical signal is broken. It means structural demand is absorbing some of the cyclical weakness. You need to separate the two components, use PMI data and Chinese import figures to isolate the cyclical signal, and track EV sales growth and grid capex announcements to understand the structural floor.
Dr. Copper's PhD has not been revoked. But the thesis has a new chapter. Read the whole thing.