Using Iron Ore To Track China Growth

Bullish Equities, Bullish Euro

• China now competes with the US as a key driver of global growth. Yet the quality of Chinese growth data, especially GDP, is poor.
• Alternative measures such as electricity production, PMI surveys and consumer sentiment provide better visibility on the Chinese cycle, but commodity prices could provide some of the best real-time measures of Chinese growth.
• Iron ore is China’s second largest commodity import (after oil). We find that since 2012, iron ore is providing a timely often leading indication of Chinese growth. It provides additional insights that other markets cannot.
• Therefore, iron ore prices should be used along with other markets such as copper, oil and Baltic dry to track Chinese growth in real time.

The Problem With China GDP

China’s economy has grown substantially since the 2008 global financial crisis. On a purchasing power parity basis, it has overtaken the US to become the world’s largest economy (Chart 1). It is also much larger than the next largest economies, India, Japan and Germany.^[1] Both investors and policymakers are therefore increasingly focusing on Chinese growth in addition to US growth to determine the path of global markets. The Fed now regularly incorporates developments in China into its policymaking process with more mentions of China than any other country in its FOMC meetings.^[2]

The trouble with this new focus is the poor quality of official Chinese growth data, especially GDP. For a start there’s the lack of volatility – notwithstanding this year’s lockdown-related drop in growth, China GDP appears to follow an unusually smooth path (Chart 2). Then there’s the fact that Chinese policymakers use GDP as a policy instrument, so almost inevitably GDP will print in line with the stated GDP goal.

Chart 1: China Has Surged to Become Largest Economy on PPP Basis	Chart 2: China Has Unusually Smooth GDP

What is China’s True GDP Growth?

This inevitably leads to the problem of what growth to track and forecast – do we stick with the officially reported GDP or is there some other, more realistic, measure of growth? Many have attempted to get to the “true” growth measure. Even China Premier Li Keqiang reportedly favours looking at rail cargo volumes, electricity consumption and bank loans rather than GDP. Other approaches include using nominal GDP (which exhibits more volatility than real GDP), using a broader set of Chinese growth numbers, using other countries’ export data to China and even using alternative data like satellite data.

For the purpose of this paper, we pick three alternative measures:

1. Nominal GDP. Unlike real GDP, this measure exhibits more volatility. It suggests that adjustments to the price deflators could be the way real GDP ends up being so smooth.
2. The Federal Reserve’s China Cyclical Activity Tracker (or C-CAT). This recently developed index expands on Premier Li’s idea and looks at eight Chinese indicators (Fernald, et al., 2019), each of which are published monthly: the consumer expectation index, electricity production, Chinese exports, floor space started, railway freight traffic, retail sales, industrial production and fixed asset investment.
3. Chinese imports derived from other countries’ export data. We have a high level of confidence in data produced by other countries; therefore we can calculate the exports to China and HK. This would reflect Chinese imports, which would measure Chinese demand.

While none of these measures are perfect, they do exhibit more volatility that real GDP (Chart 3).

Chart 3: Alternative Measures of China Growth Cycle

Using Iron Ore Prices

While these three measures are good, we can go one step further by capturing the Chinese cycle on a timelier basis. So rather than using monthly or quarterly data, we can use commodity and financial market prices. And given that China is one of the largest commodity importers in the world, shifts in commodity prices are obvious candidates to reflect Chinese demand.

In terms of what comprises Chinese imports, integrated circuits are largest, followed by oil, iron ore and then copper (Chart 4). While oil and copper prices are widely followed and often used to track Chinese growth, iron ore prices have been less well studied. Part of the reason is the poor quality of data up to 2008 due to the now defunct benchmark pricing system and the start of iron ore swaps in 2009), and part is a case of old habits dying hard. The good news is that iron ore futures have become much more liquid and are more reflective of demand patterns than before.

Chart 4: Iron Ore is China’s Third Largest Import

How Does Iron Ore Compare to Chinese Growth?

We use the generic 2^nd month SGX Iron ore 62% Fe Fines contract as our preferred measure of iron ore prices^[3]. When we compare annual changes in iron ore prices to our three measures of Chinese growth, we find strong correlations (Charts 5-7). The relationship appears especially tight from 2009 to 2018. Last year, though, iron ore prices went up in the first half even with Chinese growth falling. This could partly be due to China bringing forward production or inventory building that required iron ore to offset weakness from US tariffs on Chinese trade.^[4]

But it could also suggest that iron ore could be measuring growth not captured by our three main measures. Certainly, Chinese stock markets believed something was improving – the China A50 rallied 40% in the first half of 2019 (Chart 8). Other economic time series also rose in 2019H1, notably China non-manufacturing and consumer expectations (Charts 9 and 10).

Chart 5: Nominal GDP vs Iron ore (y/y)	Chart 6: C-CAT Index vs Iron ore (y/y)

Chart 7: Chinese Imports vs Iron Ore (y/y)	Chart 8: Stocks and Iron Ore Rallied in 2019H1

Chart 9: IP and Services PMI Jumped in 2019H1	Chart 10: Iron Ore vs Consumer Expectations

How Do Other Market Prices Compare?

The commodity markets most closely tied to China, apart from iron ore, would be oil, copper and the Baltic dry index. Oil and copper should correlate highly to growth because China is the biggest importer of both, while Baltic dry would capture China’s dominance in trade and hence shipping. The other two candidates for market prices would be Chinese 10y bond yields and Chinese stocks. Again, we look at annual changes in each, and we use SGX futures prices where applicable.

Broadly speaking, all these measures appear to correlate to varying degrees with measures of Chinese growth. In charts 11-15 we show them compared to the C-Cat index. At one end, bond yields, oil and copper appear to capture most of the ups and downs well. At the other end, Chinese stocks and Baltic dry exhibit a less stable correlation.

Chart 11: C-CAT Index vs China 10y	Chart 12: C-CAT Index vs A50 y/y

Chart 13: C-CAT Index vs Copper y/y	Chart 14: C-CAT Index vs Oil y/y

Chart 15: C-CAT Index vs Baltic Dry Index y/y

To clarify the relationship between each of these market prices to growth, we can run correlations. We perform them against the Fed measure of Chinese growth, the C-CAT index, and look at three time periods capturing various cycles (Table 1). The results show that stocks have a low and often negative correlation with growth so are unlikely to be a reliable measure of Chinese growth. However, all the other measures show strong correlations with growth, including iron ore. All the correlations, except for stocks, are statistically significant too.

We can also assess the leading properties of the market prices by lagging them by one or two quarters. This could indicate whether they signal Chinese growth three or six months later. With a one-quarter lag we find that, at least from 2012 onwards, iron ore, copper and oil continue to have high correlations (Table 1). With a two-quarter lag, iron ore sees the most stable and high correlations. This suggests that iron ore could have important forward-looking properties that other markets do not capture.

Table 1: Market price correlations with China growth proxy (Fed’s C-CAT)

The C-CAT index is computed quarterly, but we can look at monthly data too. Observing electricity usage, we find that again most of the market prices correlate positively (Table 2). And again, Chinese stocks are the exception. Oil has tended to show the strongest correlation, though iron ore and copper also have strong correlations as well.

Looking at the leading properties, we run the correlations lagging the market prices by six months. Here we find the correlations remain positive for most, but they all fall in magnitude – except for iron ore, which from 2012 onwards shows a stronger correlation. This confirms some of the leading indicator qualities of iron ore.

Table 2: Electricity production vs market prices

What Are the Measures Currently Telling Us?

Looking at annual changes in SGX iron ore price, they have been going sideways at low levels for the past few months (Chart 16). However, looking at three-month changes, which captures the shorter-term growth pulse, we find a noticeable jump in prices. This suggest that the growth picture has been improving – at least since April – despite global lockdowns associated with COVID-19.

We can combine the five best growth indicators, iron ore, bond yields, copper, oil and Baltic dry, to arrive at a China tracker. Here, we find a similar story of a surge in prices and hence in likely Chinese growth (Chart 17). Part of the surge is associated with the dramatic rise in oil prices. That said, the levels of the short-term tracker look elevated, so there could be scope for a pullback.

Comparing iron ore prices to the overall growth tracker (the three-month change version), we find that iron ore prices have moved up with the overall tracker (Chart 18). Importantly, iron ore markets didn’t fall as much as the overall tracker during the initial COVID period, which suggested a more optimistic take on growth, which later turned our correct (Chart 19).

Since July, iron ore prices have been rising while the overall China tracker has been going sideways (Chart 19). This might mean that iron ore could have limited gains, or that iron ore markets are more optimistic on Chinese growth than the overall tracker. Either way, including iron ore broadens our understanding of the direction of Chinese growth and along with a few other Chinese-centric markers provides a real-time tracker of growth.

Chart 16: Short-term Pulse in Iron Ore Positive on China Growth	Chart 17: Short-term Aggregate China Tracker Has Surged in Recent Months

Chart 18: Iron Ore Markets Moving Up Like Overall China Tracker	Chart 19: This Year, Iron Ore Fell by Less During COVID Period Than Overall China Tracker

References

Ahmed, S. et al., 2019. Global Spillovers of a China Hard Landing. FRB International Finance Discussion Paper No. 1260.

Clark, H., Pinkovskiy, M. & Sala-i-Martin, X., 2017. China’s GDP Growth May Be Understated. NBER Working Papers Series.

Fernald, J. G., Hsu, E. & Spiegel, M. M., 2019. Is China Fudging Its GDP Figures? Evidence from Trading Partner Data. Federal Reserve Bank of San Francisco Working Paper.

Hoenderop, S., Hoenderop, J. & Williamson, P. J., 2018. An Alternative Benchmark for the Validity of China’s GDP Growth Statistics. Journal of Chinese Economic and Business Studies, 16(2), pp. 171-191.

Kerola, E., 2018. In search of fluctions: Another look at China’s incredibly stable GDP growth. Bank of Finland – BOFIT Institute for Economies in Transition Discussion papers.

1. On a nominal GDP basis, the US remains the largest economy with a 25% share of world GDP, followed by China (16%), Germany (6%), Japan (4%), and the UK (3%) ↑
2. See FOMC minutes over 2019 and 2020. See also (Ahmed, et al., 2019) ↑
3. See “Iron Ore: A Primer on the ‘New Crude’” (forthcoming) ↑
4. US tariffs were first imposed on imports from China in 2018 and were later increased sharply in 2019 ↑