Why is the iron ore price hard to forecast?
Sector rotation is our dominant investment theme for this market. The process started with investors moving from technology to healthcare. Here we look at iron ore.
Iron ore is the grandaddy of all mineral commodities by value. Only oil is worth more to global trade, but Australian iron ore is top of the pops for our exports.
In the past, the US$80-90/tonne level has been the stress low for this market. It was not always that way. Thew market turned at prices below US$50/tonne in 2015.
One a one-year basis, the price did pause around the US$90/tonne mark.
Price trends alone are no great guide to the future, but levels do relate to metrics like the marginal cost of production. As I explained some years ago, the main factor to set the price at which this market will turn is the cost of Chinese production.
That is the marginal supply side. While it is easy to construct a bear case for Australian iron ore miners, based on a cost of production less than $US50/tonne, the marginal Chinese production is closer to the $US80-100/tonne level.
Chinese attempts to reduce dependence on foreign iron ore imports have tended to support prices at around this mark in previous cycles stretching back five years.
There is also demand, which is again dominated by Chinese steel mills.
In short, iron ore is a China story on pricing, and an Australian plus Brazilian, and emerging African, story on supply of the lower production cost material.
The story we never hear in the press is that of Chinese iron ore production.
The story of Chinese iron ore
Investment banks, central banks, or government economists have hollered “fire” in the theatre of our biggest export commodity, whenever the price got low. Equally, there has been squealing from China when the price of imported iron ore got too high.
Obviously, there is a happy medium, and we should recognize that China has its own mines for iron ore, mines that could easily go broke if prices are too low. The Chinese steel industry wants lower buying prices, while the mines want higher selling prices.
As with all things in China, working out what is really going on can be difficult. Many data sources report an iron ore production profile that looks like this.
You can see that it went gangbusters from 2003 until the slowdown in 2014.
However, as pointed out in US Geological Survey research note from 2017, the picture is distorted by the highly variable grade of Chinese iron ore mines. Once you allow for this factor, the Chinese domestic production profile looks completely different.
The missing factor, in most all official commentary, is the grade of Chinese ore.
This is low, typically assumed to be 30% iron content, but lower again in practice, due to depletion of large ore bodies, and an abundance of non-economic activity.
This adjustment to the data was reported by the USGS in 2017 and was itself triggered by a change of reporting by the China Iron and Steel Association (CISA).
The remainder of this research note fills in the more recent production picture so that we can form a better judgement of the opportunity in ex-China iron ore mining.
Iron ore grade and economic nationalism
The West is just beginning to toy with massive state intervention in the structure and operation of our economies. The motivation is “national security”, the idea that we cannot trust foreigners, who harbor bad motives when trading with us.
China has a large state-owned enterprise system but is also home to private activity.
The concept of the Chinese being Chinese is baffling to Americans, but true.
Capitalism with Chinese characteristics means that there are many competing private enterprises, state-owned enterprises, local government authorities, provincial level governments, autonomous regions, and an overarching central government.
Notice that I used the word “competing” because that is how it works.
China does not operate on a simplistic “top-down” command and control basis. There are elements of that, but also a long and proud tradition, stretching back thousands of years, whereby people do what they want, within limits, when nobody is looking.
You see ambiguity all of the time in the Chinese national statistics.
One set of numbers will not reconcile with another set of numbers because of the differing motivations of competing parties, and what to do about New Year.
The rare earth industry quota system is a prime example of this phenomenon. There have been occasions where different Chinese government agencies have imposed quotas for production and separation that were mutually incompatible.
The logical impossibility of meeting both sets of restrictions simultaneously has never stopped rare earth enterprises from certifying their compliance.
The way that the Chinese get around this is by creating parallel systems of statistical data, when it suits their interest to do so. The biggest and best example of this is the reporting of Chinese iron ore production. In most nations, you would expect to see one official number, a statistic that tallied well with royalty obligations.
In China, there are at least two reported numbers for iron ore production.
The difference between them relates to a single mystery number: the iron ore grade.
The National Bureau of Statistics (NBS) of China, the central government agency akin to our Australian Bureau of Statistics (ABS), reports crude iron ore production.
In contrast, the China Iron and Steel Association (CISA) reports this as a memo item alongside the more economically important useable iron ore production.
The difference is the grade, since crude iron ore is simply the tons of ore dug, while the useable iron ore figure is standardized to 62% iron ore content.
This the grade of iron ore that you can happily shovel into a blast furnace to produce steel, while the actual grade of iron ore that you dig may be much lower.
You might think, in a rational world, that most market commentators and forecasters would be all over this fact. I won’t speak for the market in general, but our Reserve Bank of Australia (RBA) certainly was not cognizant of this fact in 2010.
I don’t blame them for that, as I only found out last week!
That is why it has been so long between articles. The problem is the prevailing view that Chinese iron ore production grades can be proxied at 30-33%. This is false.
Let us explore this obscure topic more deeply as it relates to the craziness that results from excessive nationalism in the reporting of statistical data.
Obviously, if you are in the Chinese government, detailed information on the strength of domestic iron ore production is a sensitive topic. Larger numbers look better.
Conversely, if you are a CISA steel producer, that needs to fill a blast furnace with ore, you will want to know how useful Chinese ore is compared to imported product.
Lest you think that this is just me having a swing at China, I would refer you to a prior piece I penned on the travesty that is Australian official trade data, from the ABS.
Due to ABS censorship, under the 1905 Privacy Act, this data is worse than useless.
Even our own Department of Foreign Affairs and Trade (DFAT), has pointed out that the Australian official trade data is a national liability in trade negotiations.
We live in a time of widely propagated official nonsense.
Sit back and enjoy the story!
The World Steel Association dataset
The US Geological Survey is respected for the depth of their expertise, and general quality of their statistical reporting. However, they are US-centric, and not always focused on the industry picture as any producer or consumer might be.
Whereas the USGS recognized the problem in 2017, the World Steel Association (WSA) began making adjustments to their Statistical Yearbook in 2005.
In that year, they started parallel reporting of crude iron ore production in China with a second series of useable iron ore, standardized to roughly 62% iron ore grade.
In the 2005 edition of the yearbook, they backdated this to the year 1995.
They based this on an approach pioneered by the UN agency UNCTAD, which ran a program called the UNCTAD Trust Fund on Iron Ore Information.
If you follow that link, you will get to a Wayback Machine archive of the now dead webpage which described this former data gathering effort. Yup. It is gone!
There is no conspiracy here, just standard issue social neglect.
Anyway, here is the WSA data alongside the Chinese NBS crude iron ore production data and the grade-adjusted WSA useable iron ore and the UNCTAD equivalent.
There are a couple of points to note here. Firstly, the Chinese reported NBS data is the same as the WSA data, when the latter agency first reported it. However, there were many revisions to the WSA series subsequent to publication. Secondly, the UNCTAD series of useable iron ore production predates the WSA series and differs later.
Let me explain what I think is going on here.
The UNCTAD project died in 2017, but it was started to improve the data quality for iron ore production and trade statistics. The UNCTAD series started in 1990 and was picked up by the WSA in the year 1995, backdated from their 2005 issue yearbook.
When UNCTAD ended, the calculations were continued by RMG Consulting.
I don’t know the details there, but if you bruise this dataset a few things become very obvious. Firstly, the useable iron ore data appears to be derived by self-consistency arguments between Chinese pig iron production and iron ore usage. Secondly, the frequent subsequent revisions appear to target grade consistency.
The exercise of discombobulating the data is an academic exercise. I will do that out of interest, but not here! It will suffice to show you the implied grade chart.
Notice that I have based this on the post 1995 data series and used the WSA number that was first reported in 2005. The 30% figure for grade is generally accepted as the appropriate number for Chinese iron ore. However, you will notice that it dropped a great deal post 2006. I have not worked this out, completely, but I think it is due to diverse ore bodies under exploitation, and not always for iron ore.
For instance, the giant Bayan Obo Rare Earth Element (REE) mine in Inner Mongolia also produces Iron and Niobium, with a raw grade of 65% iron and 70% recovery.
Where the iron ore comes from is unclear.
We do not know how much of the crude iron ore production had a viable iron ore concentrate as a byproduct of other metals production.
The huge uplift in volume of crude iron ore production was not accompanied by a large output of useable iron ore. It may have been a useful byproduct.
There is a tendency among Western analysis to presume that everything that China does is due to some nefarious plan. I do not approach this important topic with a mindset that all Chinese data is wrong. I think the place is complicated.
There is a huge array of mines, of all sizes, in China. From the geological surveys we know that they have roughly 80 billion tonnes of material that grades around 30%.
However, what we don’t know is how many of these mines are producing metals concentrates metals other than iron ore. This point matters.
What is clear from this data exercise is that domestic iron ore mining is not a growing sector of production for the pig iron or steel industry. It looks tapped out to me.
China imports share of iron ore usage
Analysts that use the crude iron ore data from the NBS and compare this to world iron ore production are going to get the wrong answer. This is the data anomaly that was first addressed by UNCTAD in the 1990s, later by the World Steel Association in 2005, not recognized by the RBA in 2010, but picked up by the USGS in 2017.
There are some data sources, including the hugely expensive one I subscribe to, that have not figured this out this yet. They still report China has having 40% share of world iron ore production. The real figure is more like 10%, as shown below.
The big four are Australia, Brazil, China and India, in that order. The Chinese import share of iron ore now appears to have stabilised at circa 80%.
This is consistent with China being around 50% of world steel production, and 10% of iron ore production. Their domestic ore is lower grade, and so needs to be blended with higher grade material from Australia (circa 62%), Brazil and India (circa 66%).
Geostrategic Analysis
In this note, I have focused on the bigger picture of Chinese iron ore production and consumption. I will devote a later note to specific stock opportunities.
There are two key investment conclusions I draw:
The beneficiation required for Chinese ore is around 4x to lift grade
Australia, Brazil and India are critical for Chinese supply
Let us start with the first point to address the cost of production.
Cost of production implications
Due to the average China grade being around 15%, they need to mine about four tons of ore to produce one ton of useable 62% grade material. Bulk mining cost structures are largely driven by the cost to blast and ship rock.
Beneficiation of lower grade ore also adds cost.
On simple math, we can conclude that marginal costs of production for Chinese domestic iron ore is likely more than four times that of a Pilbara operation.
With C1 cash costs in the Pilbara at around US$15/tonne Chinese costs are likely more than US$60/t. I have never seen a cost-curve for Chinese iron ore, but we can safely assume that marginal production costs are above $US60/t.
Allowing for beneficiation costs, I think US$80 to 100/t is more realistic.
Strategic implications
There is clear strategic competition between China and the USA. Even though China is the largest trading partner for Australia, our trade policy aligns with the USA.
US-China tension will lead to bad outcomes for the Australian mining industry.
Due to the dominant global market share for Australian iron ore, it is hard for China to substitute away from Australian imports. The USA is no destination worth considering.
The US consumption of iron ore is 1.5% of the global total.
Australian politicians, and our media, are woefully ill-informed.
Unfortunately, we only have one country, and it seems to be very badly led.
The solution for Australian investors is to realize two things:
Australian iron ore will continue to be immensely profitable
Government policy will successfully shrink our industry.
These conclusions seem baked in by the geology and the geopolitics.
Similar conclusions likely apply to exports of iron ore from India, due to that nation being aligned with the QUAD to contain Chinese economic development.
We expect both Australia and India to remain vital sources of iron ore imports to China, but the relationship is no longer one of partnership.
It is now borderline toxic.
I do not make Australian national policy or write idiot editorials in our financial press.
I am an investor whose responsibility lies to my clients to keep them informed.
Investment conclusion
Continue to hold quality iron ore producers in Australia, especially those which have higher grade product, and operations that can flex production. China will continue relying on Australian imports but will invest its own capital elsewhere.
The USA and Europe are ex-growth steel markets. They do not matter.
Expect Australian politicians to make announcements, with photo opportunities, about cooperative investment activities with Western partners to combat China.
Pay attention if the large steel markets of Japan, India, South Korea or Taiwan are part of the announcement. Otherwise, ignore it as it won’t make any difference at all.
For Australian companies, such as Rio Tinto, that are involved in greenfield provinces like Simandou, in Africa, consider allocating the bulk of your growth capital.
Rio Tinto has hedged its exposure to poor Australian government very well.
The other wild card is Brazil.
Brazilian iron ore is of high grade, and China is investing in manufacturing plants across Latin America. Brazil is non-aligned and a likely global winner.
I have one rule for my conduct in business:
The customer is always right.
China is the biggest customer in the global iron ore trade and Brazil understands this in a way that Australia and India do not.
Don’t give up on Australian and Indian iron ore miners.
They will still make suitcases of cash out of China.
However, for the growth business, Bet on Brazil, and dabble in Africa.
Simandou will come good, but it may take a while to show up at scale.
As always, happy investing!