Back to the cost-curve drawing-board

LONDON: When the going gets tough in commodity markets, technical traders turn to the charts for price signals and everyone else turns to production cost curves for clues as to where prices might bottom out.
In theory, once prices move below the marginal cost of production and stay there, it should only be a matter of time before producers start trimming run-rates or closing operations.
The collective response will, again over time, rebalance supply to match reduced demand, which is what low prices are signaling in the first place.
Such is the current landscape facing most of the industrial metals traded on the London Metal Exchange (LME).
Only copper is currently trading significantly higher than the marginal cost of production, enjoying what is in essence a scarcity premium, a combination of extended historical supply deficit and resulting low inventory cover.
Others such as zinc and lead are fast approaching marginal production cost price areas, while aluminum and nickel prices, it is widely agreed, are already eating into the top end of their respective cost curves.
Yet even apparently clear-cut cases such as aluminum and nickel prove that there is more in the mix than just price and that the cost curve itself is not static but highly dynamic.
Moreover, in the case of zinc and lead, determining a mine or smelter’s profitability may mean factoring in multiple cost-curves, blurring the concept of cost-curve support.
At current aluminum prices, LME three-month metal holding just above the $2,000-per ton level, a significant part of the world’s smelting sector is cash-negative. In theory. In practice, however, things are a little different.
Historically high physical premiums are boosting smelters’ revenues over and above the basis LME price and in doing so offering many operators a financial lifeline.
Producers may not benefit in full from physical premiums.
They may, for example, be selling their metal on term contracts with lower premiums than those quoted in the spot market. Or they may be selling via a trader or merchant, in effect splitting the premium revenue with a third party.
But that said, current physical premiums of around $200 per ton in all major geographic regions are sufficiently high, according to analysts at Macquarie Bank, to potentially lift about 10 million tons of global capacity into cash-positive territory compared to the basis price alone.
For some producers physical premiums are not enough to compensate for specific local factors.
Australian smelters, for example, are faced with a witches’ brew of low outright prices, a strong Aussie dollar (lifting costs) and future uncertainty in the form of the country’s proposed carbon tax.
No surprise then that Norsk Hydro is planning to mothball its Kurri Kurri smelter or that Alcoa is currently running the rule over its Point Henry plant. Australia, in short, is looking like an increasingly unprofitable place to be in the aluminum smelting business.
So, too, is Europe, the focus of most of the cutbacks so far announced.
But here too it is clear that what should in theory close in a low-price environment often doesn’t.
Take for example Alcoa’s intention to curtail its Portovesme smelter in Sardinia by mid-year. The US major has had to backtrack in the face of concerted political and union pressure to keep the plant open.
Under a compromise deal hammered out with both union and the Italian Industry Ministry, Alcoa has agreed to extend the closure deadline until September while it tries to find a buyer for the smelter.
Or consider the Voerde smelter in Germany, placed into administration earlier this month. A key power supply deal has just been negotiated, allowing the 115,000-ton per year plant to continue operating while the administrator tries to find a buyer.
Both cases help explain why Europe is still sprinkled with aging, low-volume aluminum smelters that, in theory at least, should have been killed off during the price collapse of 2008-2009.
In analysts’ jargon, both are examples of why supply can be surprisingly price-inelastic.
Chinese aluminum smelters have shown themselves to be more price-sensitive in the past, not least because many of them define the marginal cost of global production.
National run-rates dropped faster and deeper than those in the rest of the world during the Great Financial Contraction.
Capacity is once again being slashed. Henan, one of China’s provincial centers for aluminum production, has seen around 700,000 tons of annual capacity idled in recent months.
But you’d be forgiven for not noticing since China’s national production of the light metal rose by an annualized 950,000 tons in the first four months of this year.
That’s because older, higher-cost capacity in provinces such as Henan is being replaced by newer, lower-cost capacity in northwestern provinces such as Xinjiang.
China’s collective cost-curve is changing and with it so too is the global cost curve.
Nickel is another prime example of a moving target when it comes to ascertaining an industry’s cost curve.
At current nickel prices, around $17,000 per ton, most of China’s nickel pig iron (NPI) industry should be under water, according to Macquarie Bank.
NPI is defined by its relatively high cost structure. After all, the sector only came into existence in response to the super-charged rally above $50,000 in 2007.
Most believe that it is only a matter of time before NPI producers bite the bullet and rein back loss-making production.
The problem is that if they do they will simply be replaced by a new generation of NPI producer, using more efficient and therefore lower-cost rotary kiln electric furnace technology.
Macquarie estimates such new entrants can survive at prices around $15,000 per ton, possibly even lower.
A prolonged period of price weakness would therefore generate only a short-term response in terms of net supply hit and a more profound medium-term change to the industry cost curve.
Zinc and lead provide a different set of cost-curve problems.
Both tend to be produced together, reflecting the fact they are as often as not found in the same deposit. And both tend to be found in deposits with other valuable by-products, most commonly silver.
As such, ascertaining where any individual mine may sit on the cost curve is no easy matter. Rather, it is a function of at least three, and possibly more, intertwined cost curve dynamics.
Some Western mines, for example, may be almost insensitive to the price of zinc, as long as those of lead and silver fare better. Equally, the price of its by-product sulfuric acid may be as important an arbiter of profitability to a zinc smelter as the price of the metal it is producing.
In both markets the marginal cost of global production is defined by the small-scale, low-grade, low-by-product mines operating in China.
At current prices analysts such as Macquarie assess this strand of global zinc supply is close to or already below the price point at which production will be curtailed.
The first problem is that if it is, we may not notice. This part of China’s mining sector is notoriously opaque, much of it falling beneath the official statisticians’ radar.
The second problem is that China’s small-scale mining sector is almost too price-elastic, primed to restart just as soon as prices allow.
Chinese prices, that is.
China is still sufficiently non-integrated into the global metals market that supply will tend to adjust to local prices and local supply-demand dynamics rather than global ones.
It is yet another way in which what should be a relatively transparent dynamic becomes a highly blurred one.
This is not to argue against the usefulness of cost-curve considerations for medium- and long-term trading strategies. It is just to warn that there is a large gap between the simple theory and complex practice of metal cost-curve dynamics.
Waiting for a statistically significant supply reaction to below-cost prices can require a good deal of patience.
Just ask an aluminum trader.
— Andy Home is a Reuters
columnist. The opinions
expressed are his own.