Crisil blames divergent growth dataprints for WPI-CPI variance

MUMBAI, Mar 16:  With most analysts refusing to buy into the official growth numbers due to vast differences between GDP and GVA, a Crisil analysis has blamed the variance to the steeply falling WPI and the more-or-less stable CPI.
According to a report formulated by Crisil chief economist Dharmadikari Joshi and his team, “The main reason for the faster growth in manufacturing GDP is that growth in the value of inputs used for production has been slower than the value of the final output.”
As a result, the value-added has grown faster than the volume of output.
The former is used for calculating GDP while the latter for IIP.
“The reason for growth in value of inputs being slower than the value of output is that the prices of inputs fell more relative to the prices of output,” Joshi said.
“This is confirmed by inflation trends where WPI (that captures inflation in commodities used as inputs) has been consistently lower than CPI (that captures inflation in goods sold in retail markets) since the 2008 global financial crisis,” he said.
The report argues that the real manufacturing GDP is obtained by dividing nominal GDP by manufacturing GDP deflater which is composed of the relevant WPI (wholesale price index) for each sub-sector within the manufacturing sector.
“Therefore, the fall in wholesale inflation has had a greater impact on real manufacturing GDP than retail inflation which further induces real manufacturing GDP growth to increase more than the IIP growth,” Crisil said.
It can be noted that many have been questioning the new GDP series for not being aligned with ground-level indicators. The main flaw in the dataprints is that it jacks up the manufacturing GDP over what IIP generally reflects.
The issue has gained notice since the 2008 global financial crisis when manufacturing GDP has been growing faster than IIP. But coinciding with this trend is the divergence in wholesale and retail inflation numbers.
Again since the 2008 crisis, WPI has been trending below CPI (consumer price index). And so is the IIP growth that has trailed GDP growth, the report said.
“The variance may be due to the fact that IIP measures the volume of output, while manufacturing GDP aggregates the value added, which is obtained by subtracting the value of intermediate goods from the value of output,” Joshi said.
“In addition, the basket of goods used to calculate IIP is different from the data used to calculate the manufacturing GDP,” Joshi argued.
But he also noted that of late there have been signs of reversal in this trend as since January this year, WPI has grown faster than CPI, while CPI grew 3.65 per cent on-year in February, WPI rose 6.55 per cent in the same month.
Also, wholesale inflation has moved in tandem with global commodity prices, which started rising in the later part of 2016.
Joshi believes global commodity prices are expected to spike further.
According to a World Bank projection, energy and non- energy commodity prices are likely to rise 26 per cent and 3 per cent respectively this year.
Accordingly, Joshi expects WPI inflation to jump from 3.44 per cent this fiscal so far to 5 per cent in fiscal 2018, while CPI rise will be milder from 4.7 per cent to 5 per cent.
“Correspondingly, the manufacturing GDP deflater will also accelerate from 2.4 per cent in fiscal 2017 to 4 per cent in fiscal 2018, and bring down real manufacturing GDP growth. In such an event, the gap between manufacturing GDP and IIP is likely to narrow,” he said.
He accordingly expects that the new IIP series based on the new base year, are expected to lead to better capturing of ground data.
The current IIP series is based on the industry’s production structure in 2004-05, and thus does not capture the output of the new industries which have sprung up since then.
The new IIP series will correct this anomaly which would in turn further reduce the discrepancy between IIP and GDP numbers for the manufacturing sector, Joshi said. (PTI)

LEAVE A REPLY

Please enter your comment!
Please enter your name here