Our earlier discussion of the rapid slowdown in Asia trade volumes and the anecdotal evidence of growthiness issues across many industries brings up the seemingly dichotomous relationship between top-down 'data' such as GDP or PMI and bottom-up sector-level activity. As BofAML points out, there has been a significant improvement in data collection in this activity data which enables 'outsiders' to cross-check macro data and potentially obtain leading information. As markets have become skeptical of China's macro data, so the effort to search for alternative measures such as power output, container throughput, and rail transport seems worthwhile. Though not perfect by any means, the higher frequency data mapping flowchart below and a comprehension of the upstream vs downstream activity flows seems to go a long way towards building a credible view on the real state of the Chinese economy - for better or for worse.
BofAML: A guide to China’s activity data
Every day, I still look for the price of No. 1 heavy melt steel scrap.
-- Alan Greenspan, The American Iron and Steel Institute's annual meeting, 1997
In the financial sector, economists attempt to spot and forecast turning points of business cycles as well as to predict government policies. Economists used to almost exclusively rely on macro-level data such as GDP, Industrial Production (IP) and Fixed Asset Investment (FAI). This is especially true in emerging markets like China where sector-level activity data are not easily available. However, with significant improvement in data collection and rapid economic growth, a large number of sector-level data are available for us to both cross-check macro data and obtain leading information on China’s economic activities.
In this report, we wish to introduce some of these activity data with a practical and balanced approach. Markets have been increasingly skeptical of China’s macro data when the Chinese economy is in a downturn, so we think it might be worth the effort to search for some alternative measures such as power output. However, activity data are no panacea. Actually, they could have as many problems as those macro data and in some cases their quality could be even worse while analysis of those data could be much trickier, so we should always avoid accepting these data without prior examination. We believe activity data, if carefully analyzed, could be a good complement for macro data, but they could never be a replacement.
Advantages of activity data
First, activity data could be subject to less distortion. For some activity data, collections and compilations are easy and transparent, and therefore they may be considered more reliable; for some other activity data collectors may not have any reason for distorting the data. Some activity data have higher frequencies than those commonly used monthly macro indicators, giving them a special advantage in predicting macro trends. Third, sector-level activity data could provide rich information on how different parts of the economy perform at different stages of business cycles. This information might be particularly useful for stock pickers.
Disadvantage and traps of activity data
Though macro data could be occasionally misleading, sector-level activity data are definitely not crystal balls. First, individual sectors may have their unique cycles and may not be representative of the whole economy. For example, pork supply fluctuates due mainly to bounded rationality of Chinese farmers and diseases, which has little to do with the general consumption demand. Second, most activity data are just coincident or lagging indicators instead of leading indicators, although sometimes higher frequencies of those data give people the impression that they act as leading indicators. Third, activity data are not free from potentially being distorted. .Fourth, high frequency (daily, weekly) activity data are technically difficult to analyze as they are subject to seasonality that is hard to adjust in China which has its own calendar compared to western countries.
Upstream vs. downstream
We also group those selected in two categories: downstream and upstream. Usually change of downstream activities such as exports, infrastructure/property FAI and auto sales could be quickly passed on to upstream sectors such as power, coals, steel, cement, non-ferrous metals and construction machinery. In those upstream raw material sectors, prices (if not regulated) and inventories could be quite sensitive to downstream demand. The transportation and telecommunication sectors connect those upstream and downstream sectors, so we will also cover them briefly. In Chart 1 above, we outline the activity data covered here and their relationship with macro data.
Price vs. volume
In the compedium of charts below we compare the correlation of these data with more aggregate economic data, especially IP, which we believe is probably the best monthly indicator for GDP growth in China. In raw material upstream sectors such as steel, cement and nonferrous metals, price is more important. In sectors such as transportation, construction machinery, auto and property, sales volumes are more important. Electricity is an exception as power tariff is regulated while inventory is technically impossible, so we will just monitor power production and consumption.
and where do all these data come from...