There is a popular theory in crypto markets that the ratio of copper to gold prices can act as an early warning signal for Ethereum’s performance relative to Bitcoin. The intuition is elegant, and it has resurfaced in recent weeks as copper has staged one of its strongest rebounds in years while ETH/BTC approaches the closely watched 0.03 level. A new quantitative analysis from Sandmark puts the theory to a rigorous test — and the results are considerably more cautious than the narrative suggests.
The underlying logic is worth understanding before the data complicates it. Copper is an industrial metal whose price tends to rise when markets expect stronger economic activity. Gold, by contrast, tends to attract demand when investors are seeking safety. When copper outperforms gold, it is often read as a signal of growing risk appetite and confidence in the global economy. A similar framework can be applied to Bitcoin and Ethereum: Bitcoin has a more monetary character within digital assets, while Ethereum is more sensitive to network activity and the expansion of decentralised applications. On that basis, periods when copper outperforms gold might be expected to coincide with, or even precede, periods when Ethereum outperforms Bitcoin.
The analysis, covering data from January 2017 to July 2026, finds that this relationship is real in some circumstances but far too unstable to be used as a reliable timing tool. The headline result from the primary statistical test finds no meaningful positive effect from copper/gold momentum on subsequent ETH/BTC performance over the 13-to-26-week lag range most commonly cited by proponents of the theory. When exploratory tests do identify a positive association — specifically using 26-week momentum windows — the relationship fails to clear the significance threshold once statistical adjustments are made for testing multiple specifications simultaneously. More tellingly, when the 2020-to-2021 period is removed from the data entirely, the positive association disappears almost completely. The period that gave the theory its credibility turns out to account for the overwhelming majority of its apparent predictive power.
The analysis surfaces one genuinely interesting conditional finding, however. When the US dollar is in a weakening trend, the relationship between copper/gold and ETH/BTC becomes considerably stronger — in one specification, a one percent rise in copper/gold is associated with a subsequent rise in ETH/BTC of more than one percent over the following 26 weeks, with a high degree of statistical confidence. When the dollar is strengthening, that association falls away entirely. This matters for the current moment: as of late July 2026, copper/gold has rebounded sharply but the dollar has also strengthened, which is precisely the regime in which the historical relationship has been weakest. The conclusion the analysis reaches is measured but important: copper/gold can serve as a useful piece of macro context when reading the ETH/BTC relationship, but it is not a standalone signal and treating it as one carries real analytical risk.
