Caixin, official PMI divergence caused by sample difference: report

The private Caixin purchasing managers' index (PMI) and the official figure told a different story on China's manufacturing sector in July.

Sample difference and uneven regional performances may explain a divergence between private and official manufacturing data for China in July, a report said Wednesday.

The official data provided Monday by the National Bureau of Statistics (NBS) and the China Federation of Logistics and Purchasing said that the purchasing PMI came in at 49.9 last month, slightly lower than June's 50.

The Caixin General China Manufacturing PMI, based on a private survey, however, added two points to stand at 50.6, the first increase since February.

A reading above 50 indicates expansion, while a reading below 50 reflects contraction, so the two versions disagree on whether the manufacturing sector had returned to expansion, as the Caixin data showed, or was still contracting.

"The official PMI covers a sample of 3,000 manufacturing enterprises, while Caixin covers only 500," said a report by the China International Capital Corporation Limited (CICC).

The CICC played down the influence of sample size while highlighting regional differences as a cause of the divergence.

"The Caixin sample may be skewed toward eastern coastal areas," said the CICC report, adding that "the official PMI covers a larger sample and is more representative."

PMI is based on indicators including production, new orders, employment and raw materials inventories.

In July, the Caixin PMI production and new orders indices rose 4 and 2.2, respectively, pushing the PMI up, while those of the official PMI fell 0.4 and 0.1, causing the divergence, the CICC report said.

Historically, the Caixin PMI has been more volatile, according to the CICC report. Since 2011, the Caixin PMI and official PMI had standard deviations of 1.4 and 0.95, respectively. Endi