PMI or GVA – Which is the true manufacturing sector indicator, anyway?

Representative image

Is India's manufacturing sector doing well? It depends on which indicator is used to gauge its performance.

India's GDP growth more than halved to 6.3 percent in July-September, according to official data released on November 30. While this was largely expected, the 4.3 percent contraction in manufacturing gross value added (GVA) was a shock.

A day later, the S&P Global Purchasing Managers' Index (PMI) for India's manufacturing sector came in at 55.7 for November. This was above 50, indicating manufacturing activity had expanded last month from October.

Although comparing two different sets of data for different periods is inappropriate, go back further and the PMI numbers remain impressive.

In July-September, the manufacturing PMI averaged 55.9, up from 54.4 in April-June, a period when manufacturing GVA grew by a three-quarter high of 4.8 percent. India's manufacturing PMI has been above 50 from July 2021, meaning the sector has grown consistently ever since.

Meanwhile, manufacturing GVA growth averaged a meagre 1.2 percent over the same period, with the GVA contracting in two quarters.

Although the deviation between manufacturing PMI and the sector's GVA growth is not new, there has been "higher than normal divergence" between the two in the post-pandemic phase, according to Vivek Kumar, an economist with QuantEco Research.

"Prima facie, since manufacturing PMI is survey-based data, while manufacturing GVA is based on actual activity data, differences are bound to creep up," Kumar said.

The seasonally adjusted nature of PMI adds to the divergence. But perhaps the most important factor is that however strange it may sound, manufacturing PMI and manufacturing GVA do not measure the same thing.

Sub-index activity

Manufacturing PMI is made up of five sub-indices: new orders (30 percent weight), output (25 percent), employment (20 percent), supplier delivery times (15 percent), and stocks of purchases (10 percent). So, manufacturing output itself accounts for only a quarter of the headline PMI.

The reasons for the wider divergence between manufacturing GVA and PMI in the post-pandemic phase could include a significant base effect and the demand and supply shock over the past two-and-a-half years, which may have influenced the sub-indices in ways that could distort the headline PMI.

Consider supplier delivery times. Normally, delivery times would increase if demand is high. However, during the pandemic, they increased because of restrictions on activity. This pushed up input costs.

"In the current situation, the role of order book, employment status, and positive business sentiment could be buoying manufacturing PMI even as elevated input costs might be weighing on margins and hence manufacturing GVA," Kumar of QuantEco Research said.

Framing the narrative

India's impressive PMI numbers have helped take away some attention from the dismal performance of industry. In his December 7 announcement of the rate-setting Monetary Policy Committee's (MPC) decision, Reserve Bank of India (RBI) Governor Shaktikanta Das noted that India's manufacturing and services PMIs for November "are among the highest in the world".

However, non-survey data continues to be disappointing. Industrial production contracted by 4 percent in October, making it the worst monthly performance in over two years. Admittedly, October was hamstrung by fewer working days on account of festivals. But the numbers are hard to ignore, with the manufacturing sector shrinking by 5.6 percent.

"The industrial production data indicates that the global manufacturing slowdown is also being felt in India, as negative spill overs from a slowing global economy take their toll through both the trade and financial channels," noted Rahul Bajoria, chief India economist for Barclays.

Economic data has not been easy to decipher over the past couple of years. Add in the fact that national accounts data can undergo multiple revisions, and the real state of the economy may not be known for some time.