As we noted last week (see “how much can a panda bear”), the unwind of enthusiasm for China’s re-opening has been warranted. China’s hard data not only slowed sharply in March and April but is also tracking weak in May. Key commodity prices such as iron ore and copper are 20% and 17% below January (rebound) peak and might reflect both the cyclical and structural challenges in fixed asset and property. Land sales remain 30% below year ago levels. That suggests activity is likely to remain weak on a forward-looking basis.
China, at around 50% of global consumption, is clearly important for the copper price. However, the spot price of copper has often been viewed as a leading indicator of the global cycle for macro tourists. The annual rate of change in spot prices has maintained its correlation with manufacturing (chart 1).
While spot appeared to base year on year in October last year, it has rolled over again since January 2023. The renewed weakness is inconsistent with enthusiasm from the S&P500 (or global risk proxy) over the past few weeks (chart 2). To be fair, peak inflation, interest rates, resilience in the US service sector, excess savings, and money (relative to trend) might explain the divergence with goods and commodity prices. Although historically, the latter has been more reliable as a leading indicator of future growth. The S&P500 appears to be pricing a 1995-style “soft landing.”
The long-term bullish case for copper is the persistent supply deficit in the physical market. In addition, current inventory levels remain low (chart 3). According to Goldman, green uses of copper accounted for 4% of global consumption in 2020, but this is expected to rise to 17% by 2030. Taken to its logical conclusion, the “net-zero emissions” path would require an additional 54% of copper of top that estimate. Of course, that appears improbable based on current known reserves and other factors such as financing for commodities, mine costs, “green” and regulatory (jurisdictional) constraints on supply.
Copper is also an important component to produce semiconductors. However, the recent enthusiasm for companies related to AI is not reflected in the price of the industrial metal and given the physical supply constraints noted above. To be fair, the other factors affecting cyclical demand might overwhelm usage in computers and green energy in the near term.
Nonetheless, it is interesting that the spot price of copper is trading near the lowest level relative to the NASDAQ since 2000 during the last technology bubble and before the commodity super cycle (chart 4). In that episode, the combination of strong cyclical demand and extremely low inventory often led to “pinch point” pricing in copper where the required price to match demand could increase almost exponentially.
The weakness in relative price is also reflected in the valuation of copper miners compared to the NASDAQ. The SOLACTIVE index of global copper miners’ trades on 5.9 times cash flow and 12.3 times 12-month forward earnings. In contrast, the NASDAQ trades at 18.3 times cash flow and 27 times 12-month forward earnings. NVDIA trades at 141 times current cash flow and 27 times next years’ cash flow. In that context, a diversified basket of copper miners with leverage to the same secular opportunity is the antithesis of euphoria. To be fair, industrial manufacturing demand is clearly not the only sector driving total global copper consumption. On the bullish side, the physical supply constraints in copper have parallels with NVIDIA’s semiconductor chips.