It's A Long Way Down

Includes: BHP, RIO, VALE, VLO
by: Paulo Santos

I'm going to make an incredible prediction in this article. Basically, what I am going to say is that an industry, whose leaders trade for an aggregate of hundreds of billions of dollars, will, in the next few months and years, suffer a brutal revaluation, losing a good part of the value it has today.

The reason I'll make this prediction, is because I think we are presently at a crossroads for this industry, and the industry has characteristics that make such an outcome possible and, in the present case, even likely.

That industry is iron ore. A few days ago, I wrote that "The Iron Ore End Is Nigh." This is not in the sense that the iron ore industry would forever be gone, but that on the short term iron ore prices would be pressured (like they were) and the companies providing it would see diminished profitability. But today, I am going a bit farther - I will be making the case that such price implosion and diminished profitability will not be short term in nature, quite the contrary. The industry in question will be challenged, in terms of valuation and profits, for the medium to long-term.

Iron ore is used to make steel, and steel is consumed mostly in investment, in making structures - in adding capacity. Now, industries that add capacity have a fundamentally different behavior from those that do something that gets consumed.

For instance, let us look at the difference between the usage of crude, the making of automobiles and the making of machines to make automobiles. If China were to make 10 million automobiles in year 1, 15 million in year 2, and 12 million in year 3, you'd always get an increase in automobiles circulating, so you'd get an increase in crude usage. However, for the auto makers, from year 2 to year 3 you would see the disappearance of 20% of your market … auto makers are already a cross between an industry that sells capacity (autos) and consumption (autos get replaced on a given cycle). Now, the machines that make autos are replaced on an even longer cycle, so imagine what happens to the demand for those machines in the scenario above! The auto maker might need to buy some new machines from year 1 to year 2, but his demand for those machines can fall as low as zero, from year 2 to year 3…

This illustrates how different the behavior of demand can be from industry to industry, depending on how the end product is used and consumed. Industries selling capacity not only need growth in their end markets, they need exponential growth. That's the reason why these industries tend to be highly cyclical in nature, and to get awarded lower earnings multiples.

And what does this have to do with iron ore? A lot. Iron ore is a commodity that provides capacity (is used in investment to provide capacity) yet iron ore companies today are valued as if its consumption can only grow into the future, which is helped by reliance in plans to increase steel production in China, and by the fact that China had been showing the required exponential growth for a while. Investments in new capacity are ongoing on every large player, based on the assumption that such exponential growth will carry on - when obviously exponential growth at high rates always dooms itself, and given what is happening in the residential construction market in China, it seems such fate has arrived. Steel production and thus iron ore consumption in China can be expected to stagnate, or even drop, in the coming 1-2 years due to the developments in residential housing in China (as well as auto production in China and shipbuilding worldwide, but those are less relevant), and this will collide with growing supply capacity.

In a way, iron ore companies are valued much the same way as crude/energy producers. This is what makes it possible for Rio Tinto (NYSE:RIO), VALE S.A. (NYSE:VALE) and BHP Billiton (NYSE:BHP) to have market capitalizations of $107.5 billion, $128.3 billion and $202.6 billion, respectively. But as we've seen, the dynamics of iron ore (and steel) usage can be deeply different from those of crude.

Indeed, steel is mostly used to add capacity, instead of being consumed and lost in a regular manner as happens with energy products. This means that, once a phase of exponential investment ends, steel use can actually decline from year to year. And with it, iron ore will also decline - this against plans of further production expansion, and on a sector whose main suppliers all have high ROE's (again, RIO, VALE and BHP have ROEs of 10.9%, 29.6% and 38.3%, respectively). High ROEs are obviously incompatible with a sector dealing in fungible commodities that suddenly finds itself oversupplied. The kind of consequence to be expected can be seen on what happened to refiners like Valero (NYSE:VLO) from 2007/2008 to the present - these companies, too, were selling a commoditized product in a market that was tight for a while, and then became oversupplied.

Now, mines have considerable investment / fixed costs, and the high ROEs mean a lot of room to cut prices. Those are the ingredients for a steep fall in iron ore prices, much as they were for the recent historic rise pictured below.

Click to enlarge

While crude might have had a similar rise and no one expects it to become as cheap as it was in the past, the dynamics on the iron ore market are different, as explained above. This means that iron ore can, indeed, get back to the prices of yore. And if it does, as now seems reasonably likely, when you have residential prices falling in every large city in China, and such will probably dictate the decrease in Chinese residential construction activity - a sector that presently is building around 48 million dwellings (compare that to less than 3 million at the top of the U.S. RE bubble).

The counterpoint

Although the logic explained here is solid, it might be that it's still too soon to see its effects. Right now, there's little doubt that iron ore will be pressured, but it might still not be the start of the medium/long term trend I am speaking about.

The reason I'm making this counterpoint, comes from the excellent "Chartbook" that Rio Tinto compiles every year. In this document, two charts stand out. These charts are the reasons why even though I am certain of the theory I described, I also have some doubts regarding the exact timing. The charts also illustrate the theory, in that several advanced economies have already gone through the predictable process I am trying to explain here. (bear in mind that the charts end in 2010, the process is a bit farther along already, and the U.S. is already consuming less steel per capita than China)

Click to enlarge


Given the dynamics affecting the iron ore market, as explained, I'd expect iron ore prices to gravitate towards prices similar to those hit at the bottom of the 2009 crisis - around half what they are today. But worse still, I'd expect the prices to stay at those levels for years, such will mean a deep revaluation of the main iron ore suppliers, which will certainly shed a great deal of market capitalization in the process.

Given this expectation, the stocks in the sector should be sold as soon as possible, as the process is already ongoing (iron ore prices are already dropping). For those mode adventurous, short positions could even be considered, since the expectation is for a substantial drop in market capitalizations in the iron ore industry (along with prices, revenues and profits).

Disclosure: I have no positions in any stocks mentioned, and no plans to initiate any positions within the next 72 hours.