RBI economists to CSO: Your GDP projection methods are outdated
An analysis of GDP data revisions by a team of economists from RBI shows a general bias towards upward revisions in growth rates relative to first advance estimates releases
New Delhi: The Reserve Bank of India (RBI) has advised the Central Statistics Office (CSO) to use state-of-the-art methods like “Nowcasting” and high frequency indicators, such as commercial vehicle sales, domestic and international air passenger traffic and foreign tourist arrivals to better forecast gross domestic product (GDP) growth estimates in order to avoid significant revisions in the data relative to the first advance estimates (AE).
An analysis of GDP data revisions by a team of economists from RBI shows a general bias towards upward revisions in growth rates relative to first AE releases. For example, the first AE of GDP growth in 2015-16 said the economy grew at 7.6%. While the provisional estimate kept the GDP growth at 7.6%, the first revised estimate and the second revised estimate took the GDP growth to 7.9% and 8.2%, respectively.
Data revisions from 2003-04 onwards show the CSO revised real gross value added (GVA) growth estimates relative to AEs upwards in 10 years, and downwards in the remaining four years.
“Average upward revision was of the order of 70 basis points, while it was only 27 basis points in the case of downward revisions. In the case of real GDP, AEs were revised upwards in twelve years (average of 81 basis points), and only in two years were there revisions in a downward direction (average of 204 basis points),” noted the RBI report titled “Examining Gross Product Data Revisions in India”.
The RBI economists said that they observe a bias in revisions at the turning points of growth cycles. “We observe that at the ‘upturns’ in the Indian economy during 2005-06 and 2009-10 and the downturn of 2008-09, which coincided with the global financial crisis (GFC), there have been huge revisions in the subsequent estimates,” the report said.
In 2005-06, the AE of real GVA growth at 8.1% was subsequently revised upward to 9.5%, an underestimation of around 140 basis points. Similarly, in 2009-10, the CSO revised the AE of real GVA growth of 7.2% to 8.6% in its final estimates which is an underestimation of around 140 basis points. On the other hand, for 2008-09, the CSO revised real GDP growth downward from 7.1% in AE to 3.9% in final estimates, an overestimation of around 320 basis points.
The RBI economists advise that, instead of only taking into account the sectoral indicators like the index of industrial production and crop production data among others—which are released with a lag, CSO should also look at high frequency indicators, such as commercial vehicle sales, domestic and international air passenger traffic and foreign tourist arrivals.
They said CSO should examine relationship of these indicators with GVA of respective sectors frequently and use the revised coefficients for computing first AE.
“Furthermore, these sets of new information can be incorporated in state-of-the-art methods such as “Nowcasting” to produce better (potentially unbiased) estimates of economic growth relative to simply employing the reported first AE,” the report said.
Nowcasting, which is a cross of Now and Forecasting, is often employed in real time weather forecasting as well as for up-to-date macro-economic forecasts.
Editor's Picks »
- Same-store sales growth trips at Future Retail
- Cipla Q4 FY18 results no reason to reverse stock underperformance
- Dr Reddy’s Q4: It’s a wait and watch, share price spike notwithstanding
- What SBI Q4 results say about the Indian economy and the bank
- Patanjali’s slowing growth does not mean that Colgate’s is accelerating