FOR mining investors there is something sinfully alluring about Glencore, an Anglo-Swiss metals conglomerate. It is the world’s biggest exporter of coal, a singularly unfashionable commodity. It goes where others fear to tread, such as the Democratic Republic of Congo (DRC), which has an unsavoury reputation for violence and corruption. It recently navigated sanctions against Russia to strike a deal with Rosneft, the country’s oil champion.

Yet Glencore could still acquire a halo for itself. It is one of the world’s biggest suppliers of copper and the biggest of cobalt, much of which comes from its investment in the DRC. These are vital ingredients for clean-tech products and industries, notably electric vehicles (EVs) and batteries.

The potential of “green” metals and minerals, which along with copper and cobalt include nickel, lithium and graphite, is adding to renewed excitement about investing in mining firms as they emerge from the wreckage of a $1trn splurge of over-investment during the China-led commodities supercycle, which began in the early 2000s. The most bullish argue that clean energy could be an even bigger source of demand than China has been in the past 15 years or so.

Optimism about the mining industry is a remarkable turnaround in itself. In the past four years the business has endured a slump that Sanford C. Bernstein, a research firm, judges to have been as deep as in the Depression. In 2014-15 the four biggest London-listed miners—BHP Billiton, Rio Tinto, Glencore and Anglo American—lost almost $20bn of core earnings, or EBITDA, as commodities plunged. Glencore, which was hit hardest, scrapped its dividend and issued shares to rescue its balance-sheet.

Commodity valuations rebounded last year, and again led by Glencore, mining-company share prices rallied. Recent results show that the four biggest firms not only swung from huge losses to profits but also cut net debt by almost $25bn in 2016. BHP and Rio made unexpectedly large payouts to shareholders. Ivan Glasenberg, Glencore’s tough-talking boss, says the company is now in its strongest financial position in 30 years. “What a difference a year makes,” he exclaims.

Underpinning the turnaround have been curbs on supply—both voluntary, to push up commodity prices, and involuntary, such as strikes and stoppages. Capital expenditure has fallen by over two-thirds since 2013 (see chart). All the firms are reluctant to embark on big new mining projects. Mr Glasenberg says the industry’s pipeline of new copper projects, for example, is shorter than it was before the China boom. Rio’s giant Oyu Tolgoi copper site in Mongolia’s Gobi Desert is a rare exception. The main focus at all the mining firms is on rebuilding balance-sheets and rewarding shareholders who kept the faith.

Even as they promise capital discipline, however, demand for green metals and minerals is tempting them to spend. Last year BHP declared that 2017 could be the year “when the electric-car revolution really gets started”. A recent surge in the prices of battery ingredients, such as copper, cobalt and lithium, has added to the excitement. China, the world’s biggest manufacturer of EVs, is gobbling up supplies. In November China Molybdenum, which is listed in Shanghai, became the majority owner of Tenke Fungurume, a vast copper and cobalt mine in the DRC. Tellingly, the price of platinum, which is used in catalytic converters in internal combustion engines, has lagged behind.

BHP, which has looked closely at EV-related demand, estimates that an average battery-powered EV will contain 80 kilograms of copper, four times as much as an internal-combustion engine. This is split between the engine (the largest share), the battery and the wiring harness. It forecasts that by 2035 there could be 140m EVs on the road (8% of the global fleet), versus 1m today. Manufacturing them could require at least 8.5m tonnes a year of additional copper, or about a third extra on top of today’s total global copper demand.

According to Sanford C. Bernstein, which uses a bold estimate that almost all new cars will be electric by 2035, global copper supplies would need to double to meet demand by then. Finding and digging up all the metals that stand to benefit, plus new smelting and refining capacity, could require up to $1trn in new investment by mining companies, it says. Hunter Hillcoat of Investec, a bank, says the transition could require the addition of a copper mine the size of Chile’s Escondida, the world’s biggest, every year.

Therein lies the rub. By one estimate, it takes at least 30 years to go from finding copper deposits to producing the metal from them at scale. Some of the big ones in operation today were discovered in the 1920s. Because of declining ore grades, community resistance, lack of water and other factors, copper supply will be overtaken by demand in the next year or two. But prices would have to rise considerably to spur the necessary investment in mines.

Sharply higher prices for copper could, however, spur the search for alternative battery and EV materials such as aluminium. When prices of nickel, an additive in stainless steel, soared a decade ago, stainless-steel manufacturers found ways to make products less nickel-dependent.

Another difficulty in supplying a future electric-vehicle revolution is the often inhospitable location of some of the most promising minerals. Cobalt, for instance, is a by-product of copper and nickel. Total volumes are about 100,000 tonnes, and about 70% lies in the DRC. Unregulated artisanal miners produce a lot of it, which has led to worries about “conflict cobalt”.

Indeed, the DRC is likely to be the main source of many of the minerals needed for EVs and batteries. Paul Gait of Sanford C. Bernstein calls it the Saudi Arabia of the EV boom, referring to the kingdom’s role in oil markets. But firms such as BHP and Rio are thought to be reluctant to invest there because of concerns about the country’s stability, transparency and governance.

In the short term the mining industry remains gun-shy about new investments. As Glencore’s Mr Glasenberg notes, it has been fooled before by estimates that demand for copper will double—the latest such misjudgment came as recently as 2008. The very biggest firms, BHP and Rio, have an additional reason to hesitate before splurging on battery materials. Their cash cows are iron ore and coking coal, the raw materials of steel, which are used more heavily in petrol and diesel engines than in EVs. BHP also produces oil, demand for which could one day be affected by battery-powered vehicles. Anglo American has a large platinum and palladium business, feeding demand for diesel and petrol catalytic converters.

All the firms insist that such diverse mineral exposures in fact provide them with a “hedge” whichever way the vehicle fleet develops (though they play up the copper in their portfolio as possibly the best bet of all). Rio is unique among them in also having a lithium-borate project, in Serbia, which it is developing as an option on a batteries boom.

For an unhedged bet, it may be small miners such as Canada-based Ivanhoe that are best placed for a surge in EVs and batteries. Ivanhoe recently said it planned to develop the Kamoa-Kakula deposit in the DRC (pictured on previous page), which it calls the biggest copper discovery ever, containing the highest-grade copper that the world’s big mines produce. Zijin, a Chinese miner, sees the same opportunity and is paying Ivanhoe $412m for half of its majority stake in Kamoa-Kakula. Ivanhoe’s founder, billionaire Robert Friedland, speaks of the metal as the king of them all. “Based on world ecological and environmental problems,” he says, “every single solution drives you to copper.”