Market Insights: Steel yourself, Chinese demand for iron ore may have already peaked | BetaShares

Market Insights: Steel yourself, Chinese demand for iron ore may have already peaked

BY David Bassanese | 22 March 2016

The recent rebound in iron ore prices has breathed renewed hope into the commodities market, particularly in light of suggestions that China may further boost housing and infrastructure activity.  That said, new research into the longer-term outlook for Chinese steel production – and hence iron ore demand – paints a more sobering picture, which is all the more reason to suggest the Australian dollar is overvalued at current levels.

China’s Per Capita Steel Production May Have Already Peaked

In a recent post I suggested that despite the rebound in spot iron ore prices of late, the immediate supply-demand fundamentals for the commodity were still not favourable.  That’s because  China still faces excess steel capacity and a glut of newly built residential properties, while global export supply is also still rising.

That said, a post back in April last year nevertheless suggested there might still be a more optimistic longer-term picture, given that Chinese steel production per capita could still rise to the levels seen in Japan and South Korea.  In particular, as seen in the chart below, it was noted that Chinese steel production per capita was around 600 tonnes, which was similar to that of Japan and Korea at a similar stage of per capita income, but less than the 900-1000 tonnes per capita that have been achieved in these countries more recently at much higher levels of GDP per capita.  Even assuming only modest Chinese population growth, that would still be consistent therefore with rising levels of steel demand.


New research from the staff at the Reserve Bank of Australia, however, proposes a more qualified view – which is significant as the RBA’s outlook for China will no doubt affect its attitude toward interest rates and the Australian dollar.

In a recently released paper* from an RBA hosted closed-door conference on China, the authors noted the outlook for Chinese steel demand remained quite mixed. On the one hand, optimists point to the possible further growth in Chinese urbanisation, which would require more steel-intensive residential construction and infrastructure development.  But the exact level of Chinese urbanisation is statistically clouded and, on some plausible estimates, a ‘relatively high’ 55% to 60% of the population could already be urbanised. What is clearer is that motor vehicle use remains relatively low by developed world standards, as is the level of infrastructure more broadly – such as paved roads, sanitation and water facilities.

                                                 Motor Vehicles per 1000 People

                                                             source: see footnote

Against this, however, are some negative forces.  For starters, China’s use of steel scrap in steel production remains relatively low by global standards – and to the extent this is increased over time (especially as ageing steel structures are replaced), it will reduce the need for iron ore inputs.  Growing environmental concerns may also lead to cleaner and less energy intensive production technologies that force greater rationalisation among the many small and inefficient steel producers still dotted around the country.  It also remains the case that there is a glut of Chinese steel productive capacity, estimated at around 400 million tonnes, or 50% of current production.

                                      China – Crude Steel Production and Capacity

                                                                 source: see footnote

China can’t export like South Korea and Japan

But to my mind, the real clincher in this debate is that suggestions China could reach the per capita steel production levels of Japan and South Korea may be too much to ask, as the latter have relied on developing major steel export industries that would be hard for China to replicate given its sheer size.  The world could simply not absorb that extra level of steel exports, without either causing prices to plummet and/or the erection of much greater trade barriers.

Indeed, using statistical analysis to map the historical relationship between steel production per capita and economic development across a broad range of countries, the RBA’s paper estimated that Chinese per capita steel production (at around 600 tonnes) is already above its theoretical peak of 560 tonnes. According to the authors, “steel production therefore falls gradually, in per capita and aggregate terms, from current levels in our projection.”

                                                             Steel Production per Capita
                                                                    (5 year moving average)
                                                         Source: see footnote

Commodity Prices and the $A are over stretched

This relatively sobering view of longer-run Chinese steel production is in stark contrast to many still optimistic predictions that China would be producing 1 billion tonnes of steel per year within the next decade or so – a view often expressed by major iron ore producers such as BHP-Billiton and Rio Tinto.  If it’s true that Chinese steel production has already peaked, it adds to downside risks to iron ore demand and prices in coming years.  It also suggests at around US75c the Australian dollar is overvalued, with more downside than upside over the next year or so.

Investors with a view that the Australian dollar may be in for fall versus the U.S. dollar may look to express their view through the BetaShares US Dollar ETF, which aims to go up when the Australian dollar goes down (and visa versa).

*China’s Evolving Demand for Commodities, by RBA Staff economists Ivan Roberts, Trent Saunders, Gareth Spence and Natasha Cassidy. Paper presented to the RBA 2016 Conference, Structural Change in China: Implications for Australia and the World


  1. Paul  |  March 23, 2016

    Thanks for your frequent and detailed analysis David, but some graphs occasionally are not useful on account of the shape of the ‘curves’ contained therein. For example, Graphs #1 and #4 above have some curves which ‘wind back’ so that there are multiple figures on the vertical axis for a given value on the horizontal axis. The most obvious example is for Japan and USA in graph #1: for a $USGDP figure of about 35,000, there appears to be at least two figures for Japan’s Steel Production (about 660 and 900). Graphs would be more readable maybe if the ‘curve’ joining the dots simply wasn’t drawn. (Perhaps you have no control over that if the charts are taken from another source?).

    1. David Bassanese  |  March 24, 2016

      Hi Paul
      Thanks for your comment. Yes it may seem a little confusing as GDP per capita sometimes falls during major economic downturns.. as in 2009. I have changed the chart to be at 5 year intervals only and also found data to extend back US history to 1900. The US data is now very interesting also as it shows how similar to that of China it was at a similar level of development.

Leave a Reply