Chinese PMI and the mysterious belief in numbers

The Chinese official PMI climbed to 53.1 in March, reaching its best level in 12 months. In contrast, a report from HSBC/Markit put the index at 48.3 for March, the fifth straight month of decline. To control information from the microbloggers Chinese authorities have warned 3117 websites, punished 70 Internet companies, arrested 1,065 people and deleted 208,000 harmful messages

I was reading two stories recently. They were positioned right next to each other. Both stories were about China. One was about a specific data point issued by the Chinese government—the Purchasing Managers Index (PMI) for March. The other story was about a crackdown in China on microblog websites. The microblogs were shut down allegedly for spreading rumours about the purge of Bo Xilai, the controversial party secretary of Chongqing. Most investors, economists and analysts would dismiss any connection between economic information and political gossip. They would be wrong, because both stories were about something crucial to markets: information.

The PMI data was not interesting because it went higher or lower than the previous reading. It was interesting because the official information disagreed with information collected and analyzed by a private firm. Of course, the discrepancy was often ignored. Many analysts not only assumed that the information was correct, they went happily on to assume that it actually meant something. Some of the more cautious tried to explain it. The truly analytical tried to correlate and justify the two readings to prove that there was some logic in the differences.

The official PMI improved reaching its best level in 12 months. It climbed to 53.1 in March from 51.0 in February according to the China Federation of Logistics and Purchasing, which issues the data along with the National Bureau of Statistics. Above 50 is supposed to be an indication that the economy is expanding. In contrast, a final report from banking group HSBC and the London financial services consultants—Markit—put the index at 48.3 for March, down from 49.6 in February, the fifth straight month of readings below 50.

Some analysts cheered the results. They simply ignored the private number and used the official number as an indication that the Chinese economy was clearly growing again. The much anticipated soft landing had been achieved. Even if we accept this analysis, one wonders if a growing Chinese economy might also grow more inflation especially since both food and energy prices remain quite high.

Other analysts had a more technical explanation. In past years the March data has shown a big upward bounce due to momentum from the Chinese New Year, which falls in January or February. However, this take on the data is not especially optimistic, because the average rise is over 3% well above this year’s numbers. If the HSBC/Markit numbers are correct then there really is trouble. Any landing won’t be soft.

Analysts also tried to explain away the differences. The private survey was supposed to be inaccurate because it covered small companies in the export business, while the government survey was distorted because of inadequate seasonal adjustment.

The reality is probably simpler. The government or at least their sources have a very strong incentive to publish optimistic figures, the private survey does not. There is reason to be suspicious. The private survey covered 430 companies with a 70% participation rate. The official survey covered 820 companies and their response rate was a perfect 100%. Also, while the Markit methodology is transparent, the government’s is not. The reality is that no one knows exactly what goes into the government survey and the government sees no reason to tell.

Even if we assume that the government is trying to paint an accurate picture, there is no reason to assume that their sources are. According to author and Wall Street Journal correspondent, Tom Orlik, one of the problems is a “recalcitrant sample set”. In short the statisticians are not lying, but the companies reporting to them are.

This is hardly surprising. The veracity of Chinese companies reports are supposed to be guaranteed by their association with international auditing firms. Of course, when the auditing firms discovered that the companies’ books had often been cooked, they resigned. For example, in the past year Deloitte alone has dumped Boshiwa, Longtop and Daquing. The problem is probably going to get worse as the annual reports come due on 30th April.

Other reports from China are also suspect. A US consultancy determined last year that the Chinese under-report steel production of 45 million tonnes about the same as Germany’s total production. The under-reporting was due to the government’s demand to curb production.

Controlling information is the dream of every politician and CEO. The difference is that the Chinese can do it. To control the information from the microbloggers they have warned 3117 websites, punished 70 Internet companies, arrested 1,065 people and deleted 208,000 harmful messages.

The fact that a government’s attempt to control information creates bad information for both the government and everyone else is hardly surprising. What is surprising is the fact that Wall Street’s economists, analysts and pundits actually give it credence. The Chinese are only acting according to their perceived economic incentives. The motivation of investors is less clear.

(William Gamble is president of Emerging Market Strategies. An international lawyer and economist, he developed his theories beginning with his first hand experience and business dealings in the Russia starting in 1993. Mr Gamble holds two graduate law degrees. He was educated at Institute D'Etudes Politique, Trinity College, University of Miami School of Law, and University of Virginia Darden Graduate School of Business Administration. He was a member of the bar in three states, over four different federal courts and has spoken four languages. Mr Gamble can be contacted at [email protected] or [email protected]).

Free Helpline
Legal Credit