Lessons for iron ore from coal downturn: Part 3

So far we have looked at why coal prices (both thermal and met) have been much worse than forecast and some of the unexpected consequences. Both have suffered from stronger than expected supply, especially when prices have pushed many producers into loss making positions.  A big miss on power demand in China has been a huge weight on thermal coal while ex-China demand weakness has had a larger role to play for met coal.

So how come iron ore has so far managed to probably be a bit stronger than expected? Unlike coal, which has seen many parts of the supply chain perform strongly, seaborne iron ore supply has been weakened by the dramatic drop in Indian tonnage as the government seeks to clamp down on illegal mining.

Demand from China has also proven to be much stronger than expected. Crude steel production has been running at ~7.4%YoY, with it clear that steel related sectors like construction, infrastructure and auto production generally doing better than the economy as a whole.

The key to understanding iron ore pricing in the medium term is really in understanding how much high-cost Chinese iron ore is required.  This has proven to be a powerful driver of pricing as this supply has proven to be very flexible.  When this tonnage is priced out, with the consensus that it does so at ~$120/t, it tends to disappear quite quickly, bringing balance to the market.

The top of of the cost curve is estimated to be quite big, with iron ore produced in China over the 225-250mt mark having cash costs of over $120t. And unlike thermal coal, its high-cost is due to very low ore grades rather than transportation costs.  So there isn't a huge amount China can do to displace this ore internally.

The big question for forecasters (when looking beyond stocking cycles that can drive wild swings in spot prices) is when will this high cost ore no longer be needed.  And this seems to be a when rather than an if, given the incentive for seaborne suppliers and the lower growth rates of Chinese steel consumption.

This point doesn't appear to be this year.  Even if steel production is flat for the remainder of the year, China will still need an extra ~45mt of 62% Fe. iron ore in the current year. This is less than additional seaborne availability, but the differential is not going to displace the need for higher cost iron ore outside of inventory cycles.

The inventory cycle though will impact in the second half as steel production cools and iron ore shipments pick up, so prices do look likely to fall, on average, in 2H13.

But the bigger picture question is at what point does China need to absorb an extra ~100mt of seaborne supply beyond any incremental domestic requirements.  The chart on the left tries to when this might be given the balance of risks to demand and supply

The likely increase in supply in the seaborne market is set to outstrip Chinese demand over the next 3 or so years (grey supply projection above blue demand projection range)

It also seems likely to do so by more than 100-150mt.

The red line represents my likely demand forecast of 3-4%pa growth in Chinese steel, although the risks are skewed to the downside.  The space in the orange brackets represents outcomes where supply growth isn't strong enough to displace high-cost Chinese ore. As you can see, it would take relatively pessimistic assumptions on the range of supply projects for us to remain in the orange brackets by 2015/16.  A downside surprise to demand would see us get there much quicker.

So while we aren't there yet, iron ore does seem set to test the same market conditions as coal is currently experiencing in the next 12-18 months.

So what are the key lessons from coal?

The central lesson seems to be that when the high-cost flexible part of supply in China is no longer required, the next part of the equation can prove to be much less flexible and supply growth strong even at lower prices.

So for the many forecasters who have a fairly orderly decline in prices to the top end of the cost curve, this seems unlikely. If anything, the risks are that prices fall much more sharply and are sustained below the cost curve for sometime. Most put the top end of the seaborne cost curve at ~$80-90/t.  In reality when it does fall, it will go through these levels much faster and for longer than many expect.

So for investors, I don't think it will pay to jump in on the assumption we are through the top end of the cost curve, with it better wait for industry capitulation, which may take years.

It also seems very unlikely that iron ore producers will be able to manage supply risks to prices because the market is highly concentrated.  Iron ore is a much more concentrated market than met coal, with Rio Tinto, Vale and BHP Billiton accounting for 60% of the market.  But even commodities like platinum, which is also very concentrated, has seen no effort or co-ordination to manage supply risks.

If anything, like coal, the push will be to improve efficiencies and push production higher rather than lower, to get cash costs down.  Given the large amount of capex spent on infrastructure to increase capacity, producers aren't going to given up on projects that have a mine life of 20-30 years, because cash flow becomes constrained in a given quarter or even a given year. So while returns on assets from higher cost producers may be poor, that will not be enough to stop them producing.

For a country like Australia, the macroeconomic risks to growth won't come through less exports than expected.  Rather it will come through the margin squeeze affecting all other parts of the growth equation.  Fortunately for Australia, the coal collapse didn't happen at the same time as iron ore, which would have been truly disastrous.