Platinum mining costs are likely to drive the long-term price moderately higher

6 Dec 2012 by Jim Fickett.

For a significant fraction of platinum mines, market prices below $1500/oz fail to cover cash costs, and below $1700/oz fail to cover cash costs plus capex. The cost increases of recent years were not a temporary aberration. Therefore the current platinum price of about $1600, while not a screaming bargain, is on the low side, meaning that an investment in platinum is very likely to provide at least modest before-tax returns. Including tax effects, a purchase now makes sense either (1) within a tax-advantaged account or (2) if one believes inflation is likely to remain moderate.

I have mentioned before that the price of platinum is probably reasonable, because it is close to the cost of production (Platinum may be reasonably priced, and Precious metals update). In this post I'd like to examine that statement in more depth.

Mining costs are central for the platinum market

First, note that platinum, though of interest to investors for some of the same reasons as gold, has a very different market from gold. While most gold transactions are between investors, and mine supply matters very little, most platinum transactions are sales from mines to fabricators of catalytic converters and jewellery. Out of total demand of 8 million ounces for 2012, 6 million is for auto catalysts and jewellery; investment net demand is less than 1 million ounces. Total supply, also about 8 million ounces, is split between 6 million newly mined ounces and 2 million from recycling.

Here are supply and demand data for the last few years:

(All data are from Johnson Matthey, who provide a useful twice-annual market overview. The earliest year for which Johnson Matthey breaks out total recycling is 2005.)

South Africa produces, and has, most of the world's platinum

Before getting into any details on mining costs, it is important to also note that most mining takes place in South Africa, and so conditions there tend to have a large influence on the market.

According to the US Geological Survey, in 2011 South Africa produced 72% of all newly mined platinum.

Regarding platinum resources, the best data I can find are rather old, from a 2004 USGS report. These data lump together all the platinum group metals (ruthenium, rhodium, palladium, osmium, iridium, and platinum), but do attempt to quantify the reserve base, which is an estimate of the resources that are of high enough quality to become future reserves.

The 2011 USGS report mentioned at the beginning of this section has only reserves of platinum group metals, not the reserve base. South Africa had 63 million tons of reserves, out of the world total of 66 (over 90%).

In particular, labor relations in South Africa are much in the news whenever investors discuss platinum.

Costs vary considerably between mines

It is not quite enough to say, as some reporters do, that production cost is near the market price. Costs are different for different mines, depending on access to water and electricity, ore quality, past negotiations with miners, etc.

An Oct 2010 investor presentation from Implats gives some idea of the range. In 2010, most of the more expensive mines were about 50% more expensive than the cheapest mines:

And according to then-current projections, the more expensive mines would find in 2015 that their costs were about 100% higher than those of the cheapest mines:

(Note this covers only the cash costs, not capital expenditures.)

At recent market prices, about half of all mines have covered their costs

Here is a five-year history of the platinum price, from Kitco:

Note that for the last few years the price has been mostly in the range $1400 - $1800. In what follows I give the date for each news/analysis quote, and the platinum price at the time, according to Kitco. It is a little difficult to compare the various analyses because some authors talk about extraction costs only (just the current mining expenses), some talk of current cash costs (mining expenses plus royalties and taxes), and some include cash costs plus longer-term capital expenditures. However, piecing together the whole picture, there seems to be a consensus from many sources that

  • below a price of roughly $1500, many mines are not covering their cash costs, and
  • below a price of roughly $1700, many mines would not be able to cover cash costs plus capital expenditures.

18 Oct 2011, $1523. The Financial Times:

At current levels of about $1,530 an ounce, the metal is far below BarCap’s estimate of its “cost floor” of $1,700-$1,800, which means high-cost producers are losing money. If they start cutting production, platinum prices would be unlikely to fall further, even if some of the market’s worst fears about the eurozone crisis are realised.

25 Jul 2012, $1396. Christy Filen at Mineweb analyzed filings from 26 mines, and found that 11 were operating at a loss, with a further four marginal. The results are roughly compatible with splitting the above cost curves down the middle.

12 Aug 2012, $1399. Geoff Candy at Mineweb:

As Barclays writes … “From a cost perspective, the bulk of PGM production is under water at current prices after accounting for sustaining capex. …”

5 Sep 2012, $1564. Bloomberg:

Nic Johnson, who helps manage $30 billion of commodity assets at Pacific Investment Management Co. in Newport Beach, California. Platinum “moved very close to the marginal cost of production which should put an upward pressure on prices. Any type of outages in South Africa will make it more attractive.” …

Wage disputes are adding to mining companies’ struggle to maintain output. They are now digging as deep as 1.3 miles to find ore, and temperatures at the rock face of Northam Platinum Ltd. (NHM)’s Zondereinde mine in South Africa can reach as high as 162 degrees Fahrenheit (72 degrees Celsius). Lonmin got 4.4 grams of platinum-group metals from every ton of ore last year, 5.4 percent less than in 2010.

Global extraction costs are now about $1,430 an ounce, according to Thorsten Proettel, an analyst at Landesbank Baden Wuerttemberg in Stuttgart, Germany. The average annual price has exceeded that only three times in the past quarter century.

10 Sep 2012, $1593. Peter Major at the Financial Times:

Platinum mines employ around 150,000 of South Africa’s 500,000 mine industry workers, who on average have the second-highest wages in the country, after government employees.

South Africa has nearly 80 per cent of the world’s known PGM [platinum group metals] reserves (over 2bn ounces) and produces over 4.5m oz of platinum a year (75 per cent of world mine supply) and about 4m oz a year of the other PGMs.

… At $1,590 oz today, platinum is nearly double its 100-year ‘average real price’ of $830 oz. …

Between 2001 and Aug 2012, [the platinum price in South African rand rose] under 9 per cent a year. …

Meanwhile, South Africa’s inflation rate averaged 6 per cent a year, wages increased over 11 per cent a year and other costs by more. …

if you combine ‘all’ of SA’s platinum mines, $1,550 oz is needed just to maintain production and net profitability. This allows for no capex (another $200 oz would be needed for that). So, roughly half of SA’s 4.5m oz pa platinum production generates no returns below $1,500 to $1,550 an oz.

25 Sep 2012, $1633. Leo Liu at Gold Seek gives a 2011 all-in cost supply curve (based on data from GFMS) and deduces that, as of the date of writing, about half of all platinum mined is not recovering the full cost of production.

Costs seem unlikely to drop in the future

The big question for long-term investors is whether the rise in costs is ephemeral or permanent. Some of the most frequently mentioned drivers of costs are labor, increasing depth, electricity prices, and decreasing ore grades. Let's look at these in turn.

Regarding labor costs, it seems the mines are not in a strong negotiating position (due to government pressure), and hence cost rises in this area are unlikely to be undone:

“A terrible precedent has been set,” said National Union of Mineworkers spokesman Lesiba Seshoka. “We hope this is the end of labor unrest but don't think it is. Miners see that they can strike illegally, get rehired and demand a wage increase.”

Anglo Platinum already dismissed 12,000 employees at its mines as a result of the dispute as workers continue to push for higher wages. Despite the dismissals, the company said employees can return and get a 2,000 rand bonus.

The June 2010 issue of Modern Mining gives a nice overview of the depth issue. It seems clear that mines continue to go deeper, which means higher expenses for both excavation and refrigeration:

Lesego Platinum’s Phosiri, is now well into the study phase … The only negative is that the currently defined inferred resource lies at a depth of between 1 and 2,3 km – which is relatively deep by the standards of most current platinum mines. Dorian Wrigley, the CEO of Lesego Platinum, however, makes the point that platinum mining at these depths will be the norm by 2017 – the year when Phosiri is targeted to start production. …

South Africa’s current deepest platinum mine is Northam, whose No 2 shaft goes to a depth of 2 100 m (with a further extension of 300 m planned), but generally most platinum is still mined from underground working areas extending from close to surface to around 1 000 m. But Wrigley argues that this is changing fast, with at least 10 shafts in the country already going to below 1 000 m with several more planned. “Although the BFS will give us clarity, the likelihood is that Phosiri will have a vertical shaft system down to about 1 500 m with subsequent declines accessing even deeper levels of the orebody,” he says. “By today’s standards, this would count as a deep-level mine. Remember, however, that it will be some years before Phosiri will be producing – in fact, our target date is around 2017. By this time, Phosiri will fall well within the new industry norm in respect of depth. We estimate that by 2017 South Africa’s mines will be producing around 4 million PGM ounces from sub-1 000 m levels – compared to about half a million ounces from these depths in 2000.” He adds that in 2017 there will probably be at least six shafts in the country deeper than 1 500 m. …

Wrigley, who is also now COO of Village, points out that Phosiri, if developed into a mine, would – given its depth – require refrigeration, making it a relatively substantial consumer of electricity.

Electricity costs have risen at an average of 27%/year in the last four years. You might think they could fall back. But in fact they have only been catching up with reality – current charges are still cheap by global standards. And the consensus in the industry is that electricity costs will continue to rise for some time. Nic Brown at Natixis (an asset manager) says:

Another big factor for South African miners is energy costs - this represents maybe 15% of their total costs in South Africa and Eskom we believe are looking to increase electricity prices by somewhere between around 15% to 20% per annum in coming years. So if we just take these cost increases alone, you're looking at 7.5% to 8% rise [in total costs] - we would expect this to play out year after year for the South African miners

The continuing decline of ore grades is frequently mentioned as one explanation of an increase in mining costs. An analysis by the fund manager Stanlib shows the average grade at three of the large miners decreasing from about 4.5 grams per ton in 2002 to about 3.5 g/ton in 2011. This decrease has been associated with an increase of about 25% in processing costs. While new discoveries are always possible, there has been little change in known platinum resources for some time, and it seems likely that, as with many other metals, the best deposits have been used up, and ore grade will continue to decline.

Overall, then, with all four of the main factors driving recent cost increases, it seems likely that the situation is unlikely to improve and, in fact, costs are likely to continue rising.

Possible surprises

There are always possible surprises. Here are some possibilities:

  • Zimbabwe has surface deposits which are cheaper to mine, and lower labor costs, so an expansion of mining in Zimbabwe might provide lower costs; however a history of nationalizing private property, and current regulations requiring majority ownership by locals, restrict investment, so big changes are probably not imminent
  • As mentioned in a previous post (Platinum, gold, and asteroids), it is possible that mining of asteroids could provide platinum at a lower cost; however many hurdles remain to be overcome, and there is no change on the medium-term horizon
  • If the world ever moves toward hydrogen-fueled cars, the fuel cells converting hydrogen and oxygen to water and electricity will likely increase platinum demand significantly; on the other hand, if lithium-battery-powered electric cars and trucks become more common, perhaps the need for platinum in catalytic converters could drop significantly


Currently many platinum mines are losing money. That can go on for a while, but not for the long term.

While there are many uncertainties, the most likely scenario is that demand, averaged over the business cycle, will change only gradually; that South African costs will be a limiting floor on the long-term platinum price; and that these costs will not fall, and will most likely rise.

In sum, then, it is quite likely that in the longer term, the platinum price, currently in the neighborhood of $1600/oz, will need to rise to at least $1700/oz. If one purchases platinum now, there is no strong reason to expect great returns. However it is quite likely that one will at least preserve capital, and probably have some positive return.

As long as inflation is not too bad, some positive return may be good enough. If inflation gets very bad, then even if one has only a very small positive real return, the large nominal increase in price will be taxed, and could result in a net loss. So at current prices platinum is only attractive, on the above analysis, if (a) one believes that inflation will remain moderate, or (b) the platinum can be bought within a tax-advantaged account.