• In this post, I’m discussing coal, with the ultimate goal of determining whether coal equities are a buy, short or neither. I will describe the coal industry in general and the drivers of supply / demand within the domestic and global coal markets. Based on these discussions, I end the post with forecasts for 2009, 2010 and 2011 coal prices.
• I’m not particularly bullish on coal – declining demand and ample supply should keep coal prices depressed for the next 2-3 years, particularly for met coal. In my next post, I will use these price forecasts to discuss whether certain coal equities are attractive or not.
• While I’m not optimistic about a rebound in coal prices, the equities’ low valuations and the fact that they’ve contracted most of their 2009 supply (and much of their 2010 supply, for thermal coal producers) could make the equities attractive.
• But we’ll analyze individual equities in the next post – discussing the coal industry itself is plenty for one post.
• The relevant PDF materials mentioned in this post can be downloaded here: Coal-Part-1-Materials

This week, I’m writing about coal. We’re going to focus our coal discussion on two stocks, Alpha Natural Resources (ANR) and Foundation Coal (FCL). In this post, we’re going to discuss the coal sector in general, concluding with price projections for 2009 to 2011. In my next email, we’ll analyze ANR and FCL by plugging in our price projections into financial models for the two companies, and discussing valuation levels.

There are two types of coal – steam coal (also called thermal coal) and metallurgical coal (also called met coal or coking coal). Steam coal is used primarily for power generation, while met coal is used to make coke, a key component of the steel-making process. Met coal currently sells for a premium to steam coal, though in the early 2000s this wasn’t necessarily the case. The largest producer of coal is China, with 40% of world production. US is next at 17%, followed by India at 7%, Australia at 6% and Russia at 5%. Within met coal, the largest exporters are Australia (58%), Canada (13%), and USA (11%). The largest importers of met coal are Japan (27%), India (11%) and South Korea (9%). Some of the large consumers of met coal, like China and the US, produce most of their met coal domestically.

The price charts for steam coal and met coal are on page 1 of the attached coal materials at the bottom of this post. Coal prices declined from late 2004 to October 2006, dropping from $60/tonne to $40/tonne for steam coal and $130-ish/tonne to $85/tonne for met coal. The decline was caused by excess supply growth and a weather-related drop in demand; unlike the typical ~2% growth in annual demand over the past 20-30 years, coal-fired electric consumption fell by 1.1% in 2006 due to a warm winter and cool summer. There was also a decline in natural gas prices in 2006 (on page 2, we compare coal prices with natural gas and oil prices). Then, beginning in 2007 and lasting until July 2008, both thermal coal and met coal prices spiked. Thermal coal went from a low of $40/tonne to $130/tonne, while met coal spiked from $85/tonne to $400/tonne. We’ll discuss the various reasons shortly.

The two coal types, met and steam, are driven by different supply-demand dynamics.

Met Coal Demand

For met coal, demand is driven by steel demand, since met coal is used in the steelmaking process. Steel demand has been getting crushed in the past 6 months because steel is used for industrial purposes. China, which makes 35% of the world’s steel, saw production decline to 37,790 kt in December 2008, from 41,314kt in December 2007 and a peak of 46,944kt in June 2008. That’s a drop of 9% year-over-year and 20% from its peak. Production previously dropped 11% y-o-y in November and 16% in October. Page 3 of the materials shows statistics on Chinese steel production and consumption. It’s worthwhile looking at Chinese steel consumption, since Chinese demand has been a central driver behind increasing global consumption. Construction is more than 50% of Chinese steel consumption, followed by machinery at 15%. Other end markets are home appliances and automotive. Construction is slowing down sharply in China, with one analyst estimating that growth in residential / commercial floor space sold has slowed from 30% to 10% year-over-year.

Steel prices are down 40% from their peak in mid-2008 and 10% year-over-year. Page 4 shows a summary breakdown of steel production in major geographic regions from Edelweis Securities, including the analyst’s projections for 2009 and 2010. Edelweis is projecting 17% declines in Europe and North America, a reasonable estimate. But they’re assuming that China stays flat. I’m not convinced of that. I’d assume a 10% decline in China for 2009, and that would bring overall world decline to 10% to 11%. As we mentioned above, Chinese steel production is currently down 10% year-over-year and it’s not unreasonable to expect the trend to worsen in the next few quarters.

What should declining steel production imply for met coal prices, in terms of magnitude of impact? Simmons and Company put together a good historical analysis, shown on page 5. The chart shows the relationship between steel production and met coal production. In 1982, when steel production was down 8.8%, met coal production was down 3.3%. In 1991, steel production was down 4.8%, resulting in met coal production declining 15.3%. The conclusion I draw: a 10% decline in steel production will materially impact demand for met coal. A reasonable estimate is a decline in met coal consumption of 5% to 15%.

Steam Coal Demand

For steam coal, demand is driven by global energy needs, since steam coal is used mainly for coal power plants. Power demand will suffer in a recession, but materially less so than an end market like steel. Page 6 shows the relationship between US GDP growth and power consumption. Steam coal consumption declined in 1982 and 2001, but not by a lot (-0.5% in 1982 and -2.2% in 2001). In many cases, steam coal demand is more impacted by weather than the economy. Other steam coal consumption declines occurred in 1986 and 2006, when GDP growth was healthy, while steam coal consumption didn’t decline in the 1990-1992 recession. Nevertheless, early predictions are that electricity demand will decline in 2009, by 1% to 3%. One analyst told me that the rule of thumb for electricity consumption is that it would decline by two-thirds of overall GDP decline. The EIA writes:

“The economic slowdown in 2009 will lead to a decline in electricity consumption, and this factor combined with projected increases from other generation sources (nuclear, petroleum, and wind) will lead to a 1.2-percent decline in electric-power-sector coal consumption.”

As well, Chinese electricity production growth has slowed materially in the past few months, dropping from its 10% to 15% growth rate to a more benign 3% to 7% growth rate. As things get worse, the growth may turn into a decline.

Over the long-term, steam coal is somewhat substitutable for natural gas, since both are used for power generation. And nuclear energy, renewables, and hydroelectric power also compete with coal. In the US, coal accounts for 50% of the country’s electricity generation, while natural gas accounts for 20%. As natural gas prices decline, coal prices will decline too, not necessarily proportionally, but directionally. Natural gas is down 50% since July. So we also need to have an opinion on where natural gas is headed. Unfortunately, that would require a separate post.

I should add that while we’re interested in global trends with met coal, we’re mostly concerned with domestic trends for steam coal. Except in anomalous years like last year, when coal prices went abnormally high, US steam coal is generally too expensive to export. South Africa, Colombia and Russia can export coal more cheaply to Europe and South America, while Indonesia and Australia are more competitive exporters to Asian countries. So when we’re analyzing supply-demand dynamics for steam coal, we can focus most of our efforts on the domestic supply-demand situation.

Met Coal and Steam Coal Supply

On the supply side, the supply tightening in 2007-2008 was caused by a variety of factors. US coal miners trimmed production levels following the 2006 / 2007 price drops (overall US coal production fell by 1.4% in 2007); there were a handful of mine outages; and the weather returned to normal, bringing at least US demand back to its steady 1%-2% growth rate. Coal supplies also tightened globally, and countries like China, Indonesia, Poland, Russia and South Africa began to encourage or mandate that internally produced coal be used for domestic electricity generation. Infrastructure bottlenecks became a problem, particularly amongst ports and rail transportation in Australia and South Africa. And most importantly on the met coal side, there was severe flooding in Australia, which exports nearly two-thirds of the world’s coal, putting several mines out of commission. The floods occurred in Jan / Feb 2008, and that’s when met coal prices jumped from $150/tonne to $400/tonne.

The implications of this are that, especially on the met coal side, as the infrastructure bottlenecks and nonrecurring supply shocks like the Australian floods disappear, there may be plenty of coal to go around. The Baltic Dry Freight Index, which tracks shipping, is down 90%+ in the past 6 months. Yes, 90%. Demand to some degree has halted for shipped goods, which very much includes coal. So I’m making the assumption that infrastructure bottlenecks are no longer the problem they once were. That means that supply is going to increase, for both met coal and steam coal.

I will caveat that met coal is far less in abundance than thermal coal. Global reserves are concentrated in specific regions, notably the Bowen Basin in Queensland Australia, the Appalachians of the eastern USA, and North Eastern British Columbia, Canada. Other regions with large known met coal reserves include Mongolia and Mozambique. I do think it’s been the infrastructure bottlenecks that have been the major driver behind the price spikes, but the limited amounts of met coal, or the rising cost of extracting the coal, could be a factor in supporting met coal prices.

So we’ve got clearly weakening demand and increasing supply. Coal prices have been dropping, and they won’t pick up until the global economy somewhat rebounds. The million-dollar question though is how low will coal prices go.

There’s two ways to attack that. First, at what price will high-cost producers halt production? Second, just how severe is the supply-demand imbalance going to be?

For the first question, generally speaking, “marginal cost” is a metric to watch closely in the commodity world. A piece of copper is different from a can of coke. Only the Coca Cola company can make Coke, and the brand cachet of Coke gives Coca Cola pricing power such that Coke will always be sold materially higher than the cost of producing it. It probably costs 10 cents to manufacture and distribute a can of Coke. But Coca Cola sells the can for $1. A piece of copper from one company, on the other hand, is no different from a piece of copper from another company. Over the long-term, therefore, the price of a commodity should converge to the cost of the highest-cost producers. If there’s a limited amount of a commodity and demand is increasing, it will cost more and more to mine new amounts of that commodity, and therefore the price should go up over time. It’s not that there’s a limited amount of oil in the world. It’s just that the cheap sources of oil are running out, and as we resort to drilling for oil from more expensive sources, the price of oil will increase. But if oil prices shoot materially above the cost of the highest cost producers, that’s the product of a short- / intermediate-term dislocation. Ultimately, supply will catch up to demand, and the price should converge to the cost of the “marginal cost” or highest-cost producers. Conversely, when there’s too much supply, the commodity price can go below that of the marginal cost producer. High-cost plants or mines will shut down, demand may increase, and ultimately when supply and demand balance out, the price will converge back to the cost of the resulting marginal cost producer.

So who are the marginal cost or high-cost producers in the global coal market? The US and Canadian coal miners, for one. Interestingly, when prices spike, as they did in late 2007 to mid-2008, we see US coal exports jump, because it starts becoming economical for US producers to sell coal abroad. When prices drop, as they have this quarter, US exports drop materially. So what’s the marginal cost of production in the US / Canada? The easy answer is that the marginal cost of steam coal is somewhere between $40/ tonne and $60/tonne and for met coal, it’s somewhere between $70/tonne and $95/tonne (FOB mine). Those are broad ranges though. And let’s be specific about what we mean by price / cost. It costs one thing to produce coal at the mine, and another to transport it to where it needs to go. The most common used pricing metric is “FOB”, which stands for “Free on Board” and refers to the costs for mining the coal and transporting it onto the ship. “FOB mine” refers to the cost of merely mining it out of the ground, and not transporting it from the mine to the ports. In Appalachia, the cost of transporting the coal to the port is about $25 to $35/tonne. Additionally, US coal is priced in short tons, while international coal is priced in metric tons (also called “tonnes”). There are ~1.1023 short tons in a metric ton. We’ll be referring to metric tons (tonnes) throughout this email, unless specified otherwise.

For the marginal cost of thermal coal, GMP Securities estimates that average 2007 FOB operating costs for thermal coal were $34/tonne in 2007 and that the 60th percentile for 2008 is around $42/tonne. It adds a further US$5.70/tonne, which is the average capital cost/per tonne of recently announced projects assuming a 10 year mine life. That gets it to a long-term global thermal coal price of $50 / tonne FOB.

The marginal cost of met coal is higher, because met coal is more scarce than thermal coal. One high-cost producer is Western Canadian Coal, whose cash costs are $85-$95 per tonne FOB, and $60 / tonne FOB mine. AME has the following discussion of where marginal costs could be in 2013: “By 2013, we predict contract prices will be close to their long-term, stabilized levels, which are likely to be determined by the operating cash costs of marginal producers. AME predicts a contract price forecast of US$149/t FOB for premium hard coking coal in JFY2013. This price will likely be determined by cash operating costs of met coal mines. At this time, we estimate the individual marginal producers of semi-soft coking coal and low-volatile PCI will be operating at around US$75-85/t and operating at around US$80-90/t for hard coking coal. In estimating a long-term price forecast, we have factored in depreciation costs (those not included in cash operating costs) and a small profit margin.”

In terms of the second question, it’s important to get a sense of just how severe the supply-demand imbalance is going to be because prices could drop well below marginal cost. Figuring out supply-demand dynamics is tough. Part of the problem is getting the relevant information – the main consulting firms that put together detailed supply-demand forecasts for coal focus on the import-export market, or how much coal is being imported and exported, rather than total global production and supply. GMP Securities has a chart whereby in 2009, total thermal coal exports increase by 7% while imports decline by 1%, implying a material rise in inventories. For met coal, consultant AME Mineral Economics forecasts an 8.6% decline in met coal imports, driven by CIS (-22.5%), North America (-18.4%) and EU-15 (-10.2%).

The problem with tracking only imports / exports, as opposed to global consumption / production is that it doesn’t take into account consumers that are able to produce most of their met coal internally, such as China (and the US, to a lesser degree). In 2007, China produced 477Mt of met coal, compared to imports of 5Mt. Until the downturn, it appeared that those imports would continue to increase, since it had been only in 2006 when China was a net exporter. Now, the economic downturn could potentially cause its exports of hard coking coal to increase in the future. China is such a large producer and consumer of met coal that shifts in its met coal demand relative to its met coal production can have a material impact on global prices. There are internal government mandates that essentially require Chinese coal to be used internally. So those mandates would have to be relaxed in order for Chinese coal to flood the market. Until China becomes a material exporter or importer, it’s Japanese, Indian and South Korean demand that are most influencing seaborne met coal prices, since they are currently the largest importers.

Anecdotally, I suspect that supply will materially outstrip demand in 2009 and 2010, particularly for met coal. Demand is down dramatically – that can be seen by plummeting steel and pig iron demand. When we see Japanese GDP decline 3.3% sequentially in Q408, there’s cause for concern. And in terms of supply, (i) the Australian floods are gone, (ii) infrastructure bottlenecks have disappeared and (iii) coal miners haven’t yet announced as dramatic production cuts as steel and iron ore.

So I think supply will materially outstrip demand in 2009, and very possibly in 2010. The implication? Prices should drop down to marginal cost. Indeed, they may drop materially below marginal cost.

For thermal coal, we can focus on domestic supply-demand, since global prices are not going to become high enough anytime soon to spark an active US export market. Fortunately, US thermal coal statistics are easier to come by than global met coal statistics. Page 7 shows historical US supply / demand statistics, and my projections for 2009 and 2010. On the supply side, I’m assuming 2009 comes in 55mm tons lower than 2008 production. These are mostly due to production cuts in response to the reduced demand. So far, 30 – 35mm production cuts have been announced in the US, about 25% of them in met coal. On the demand side, I’m assuming electric power demand drops 2%, and since electric power comprises 85% of demand, that number is a key one to watch. I assume most met coal demand drops much more dramatically, but note that it’s a smaller piece of the overall coal demand pie. It’s also not directly relevant to our discussion, since we’re trying to get a sense of thermal coal supply-demand in the US; unfortunately, I don’t have statistics that separate U.S. met coal supply / demand from U.S. thermal coal supply / demand. My forecasts show a material oversupply of U.S. coal in 2009, and a smaller oversupply in 2010.

So let’s now make some price estimates for 2009, 2010 and 2011. We’ll use these price estimates in our projection models when we analyze ANR and FCL. ANR produces coal in Central Appalachia (CAPP) and Northern Appalachia (NAPP). FCL mainly produces in CAPP, NAPP and the Powder River Basin (PRB)in Wyoming, where coal is abundant and much cheaper than in the East (but transporting it is much more expensive).

All prices below are FOB mine ($/short tons).

2009
Spot thermal coal – CAPP: $50.00
Spot thermal coal – NAPP: $57.00
Spot thermal coal – PRB (8400 BTU): $10.50
Spot met coal (premium hard coking coal): $75.00 ($110 FOB/metric ton)

2010
Spot thermal coal – CAPP: $52.00
Spot thermal coal – NAPP: $59.00
Spot thermal coal – PRB (8400 BTU): $10.50
Spot met coal (premium hard coking coal):$70.00 ($105 FOB/metric ton)

2011
Spot thermal coal – CAPP: $54.00
Spot thermal coal – NAPP: $61.00
Spot thermal coal – PRB (8400 BTU): $10.50
Spot met coal (premium hard coking coal): $85.00 ($120 FOB/metric ton)

This post has already become quite long. So I’m going to discuss the actual coal companies and stocks in my next post. Basically, we’ll take our coal price assumptions and plug them into the projection models of the stocks we’ll be looking at.

LEGAL:

THIS COMMUNICATION IS FOR INFORMATIONAL AND EDUCATIONAL PURPOSES ONLY AND SHALL NOT BE CONSTRUED TO CONSTITUTE INVESTMENT ADVICE. NOTHING CONTAINED HEREIN SHALL CONSTITUTE A SOLICITATION, RECOMMENDATION OR ENDORSEMENT TO BUY OR SELL ANY SECURITY OR OTHER FINANCIAL INSTRUMENT OR TO BUY ANY INTERESTS IN ANY INVESTMENT FUNDS OR OTHER ACCOUNTS. THE AUTHOR HAS NO OBLIGATION TO UPDATE THE INFORMATION CONTAINED HEREIN AND MAY MAKE INVESTMENT DECISIONS THAT ARE INCONSISTENT WITH THE VIEWS EXPRESSED IN THIS COMMUNICATION. THE AUTHOR MAKES NO REPRESENTATIONS OR WARRANTIES AS TO THE ACCURACY, COMPLETENESS OR TIMELINESS OF THE INFORMATION, TEXT, GRAPHICS OR OTHER ITEMS CONTAINED IN THIS COMMUNICATION. KERRISDALE CAPITAL MANAGEMENT, LLC OR AFFILIATED ENTITIES MAY OWN SECURITIES OF OR OTHERWISE HAVE AN INVESTMENT RELATED TO ANY COMPANIES MENTIONED IN THIS COMMUNICATION. THE SENDER EXPRESSLY DISCLAIMS ALL LIABILITY FOR ERRORS OR OMISSIONS IN, OR THE MISUSE OR MISINTERPRETATION OF, ANY INFORMATION CONTAINED IN THIS COMMUNICATION.