How to deal with the GDP growth variations in forecasts
The Financial Express [
Hyderabad
]
Bengaluru
,
Chandigarh
,
Delhi
,
Mumbai
,
Kochi
,
Pune
,
Kolkata
Jun 4, 2024 | Page No.: 19 | Position: Top Left | Source: Madan Sabnavis | Sq Cm: 827 | AVE: 310147 | PR Value: 1550735
Record : 1 / 1
View Text
Merge JPEG
JPEG
Tiff
Word Doc
PDF
1
All Text
<br>How to deal with the GDP growth variations in forecasts releases. These final numbers the final numberwas much lower can overstate or understate the comes before the year comes to reduce ambiguity. can be seen when the second than the initial estimates while in true picture. Presently, the only INTERESTING OBSERVATIONS an end. This forecast can be dis- The problem is that since advance estimates are brought FY21 there was improvement in discernible use of this number is First pensed with as these numbers data has been volatile and sub- out in February and hence the growth. FY22 was different as in the Budget where revenue advance Provisional would not be used in any of the ject to base effects, most expla- FY22 growth number would be afterbeing revised downwards in numbers are based on how the FY20 5.0 4.2 3.9 major policies of the govern- nations would be erroneous if anchored at this point of time. May 2022, it went up sharply in economy has grown in the pre- ment. This would mean that the final numbers deviate to this The table (see right) gives three the final estimates. Hence, the ceding year and the projection FY21 -7.7 -7.3 -5.8 after the first advance estimates, extent. The challenges for NSO MADAN SABNAVIS of these estimates starting from sharp revision in the upward made for the coming year. This FY22 9.2 8.7 9.7 there would be only provisional are known and hard to solve the first advance estimates to the direction between January, May is necessary as there is need for FY23 7.0 7.2 7.0 estimate in the month of May. immediately given that the flow THE SHARP UPWARD revision final numbers which come after andfinalforecasts canbeverysig- some basis for making these The divergence between the of data is uneven.As a large part in the GDP growth number two years.There are interesting nificant,which can be 1% higher. projections. Here, the first FY24 7.3 8.2 7 P final and provisional numbers is of the economy is unorganised, from 7.6% to 8.20/o raises a few observations to be made from the The first issue here is advance estimates has use. Source: MoSPI a concern. This is so because for getting authentic data is a chal- pertinent issues relating to data table.There appearto be substan- whether there is any merit in In the absence of such data, all purposes, the provisional lenge. Under these circum- availability and quality of fore- tial differences between the first having an estimate in January the ministry of finance would number is used to explain what stances, it would be prudent for casts.The practice is to have the advance estimates, which come even before the financial year have to make independent fore- get for FY25 was presented, the bias. And the final numbers has happened in the economy policy makers and corporates to first advance estimate in Janu- well before the year ends, and the has ended. This is important casts of both the previous year as GDP for FY24 was assumed to be would increase the denominator which is then used for policy for- workwith GDP growth numbers ary followed by a second one in final numbers, that arrive after because data used would be well as the coming year. The call at a certain level based on real for the fiscal deficit ratio (in mulation. But once the final with a range of up to 25 bps on February. There are provisional some iterations. The provisional available generally up to on the year gone by would then growth of 7.3%. Now, with nominal terms),thus lowering it. growth numbers vary by as both sides. Meanwhile, efforts estimates provided in May, numbers lie somewhere between November (industry, inflation) be even more subjective and growth being 8.2%, there is an The opposite would hold when much as 1% from the provi- must be on to narrow down the which show that the economy the ones that are released in the or December (GST) and is based hence the NSOs estimate is upward movement which in denominator comes down. sional estimates, there are ques- time period before the final has done very well last year (as month ofMayjust afterthe finan- on several extrapolations. Hence required. This could run the risk turn will provide a higher base The next issue is whether tions of comparability that numbers are out. per the table). And over time, cialyearends.As canbe seen,there pushing up numbers for 8-9 of skewing budgetary numbers, for making a proj ection for FY25. there should be second advance come in. Ideally, the final num- (The writer is chief economist, there are final estimates, which is no definite pattern and the months to 12 months runs the but there is no alternative. For But in FY20 and FY22, there estimates. Here the number is a bers should be publishedwithin Bank ofBaroda. Views are captured in subsequent waves vary across years. In FY20, risk of seasonal factors which example, when the Union Bud- would have been a downward little more updated but still three months by August to expressed are personal <br>How to deal with the GDP growth variations in forecasts releases. These final numbers the final numberwas much lower can overstate or understate the comes before the year comes to reduce ambiguity. can be seen when the second than the initial estimates while in true picture. Presently, the only INTERESTING OBSERVATIONS an end. This forecast can be dis- The problem is that since advance estimates are brought FY21 there was improvement in discernible use of this number is First pensed with as these numbers data has been volatile and sub- out in February and hence the growth. FY22 was different as in the Budget where revenue advance Provisional would not be used in any of the ject to base effects, most expla- FY22 growth number would be afterbeing revised downwards in numbers are based on how the FY20 5.0 4.2 3.9 major policies of the govern- nations would be erroneous if anchored at this point of time. May 2022, it went up sharply in economy has grown in the pre- ment. This would mean that the final numbers deviate to this The table (see right) gives three the final estimates. Hence, the ceding year and the projection FY21 -7.7 -7.3 -5.8 after the first advance estimates, extent. The challenges for NSO MADAN SABNAVIS of these estimates starting from sharp revision in the upward made for the coming year. This FY22 9.2 8.7 9.7 there would be only provisional are known and hard to solve the first advance estimates to the direction between January, May is necessary as there is need for FY23 7.0 7.2 7.0 estimate in the month of May. immediately given that the flow THE SHARP UPWARD revision final numbers which come after andfinalforecasts canbeverysig- some basis for making these The divergence between the of data is uneven.As a large part in the GDP growth number two years.There are interesting nificant,which can be 1% higher. projections. Here, the first FY24 7.3 8.2 7 P final and provisional numbers is of the economy is unorganised, from 7.6% to 8.20/o raises a few observations to be made from the The first issue here is advance estimates has use. Source: MoSPI a concern. This is so because for getting authentic data is a chal- pertinent issues relating to data table.There appearto be substan- whether there is any merit in In the absence of such data, all purposes, the provisional lenge. Under these circum- availability and quality of fore- tial differences between the first having an estimate in January the ministry of finance would number is used to explain what stances, it would be prudent for casts.The practice is to have the advance estimates, which come even before the financial year have to make independent fore- get for FY25 was presented, the bias. And the final numbers has happened in the economy policy makers and corporates to first advance estimate in Janu- well before the year ends, and the has ended. This is important casts of both the previous year as GDP for FY24 was assumed to be would increase the denominator which is then used for policy for- workwith GDP growth numbers ary followed by a second one in final numbers, that arrive after because data used would be well as the coming year. The call at a certain level based on real for the fiscal deficit ratio (in mulation. But once the final with a range of up to 25 bps on February. There are provisional some iterations. The provisional available generally up to on the year gone by would then growth of 7.3%. Now, with nominal terms),thus lowering it. growth numbers vary by as both sides. Meanwhile, efforts estimates provided in May, numbers lie somewhere between November (industry, inflation) be even more subjective and growth being 8.2%, there is an The opposite would hold when much as 1% from the provi- must be on to narrow down the which show that the economy the ones that are released in the or December (GST) and is based hence the NSOs estimate is upward movement which in denominator comes down. sional estimates, there are ques- time period before the final has done very well last year (as month ofMayjust afterthe finan- on several extrapolations. Hence required. This could run the risk turn will provide a higher base The next issue is whether tions of comparability that numbers are out. per the table). And over time, cialyearends.As canbe seen,there pushing up numbers for 8-9 of skewing budgetary numbers, for making a proj ection for FY25. there should be second advance come in. Ideally, the final num- (The writer is chief economist, there are final estimates, which is no definite pattern and the months to 12 months runs the but there is no alternative. For But in FY20 and FY22, there estimates. Here the number is a bers should be publishedwithin Bank ofBaroda. Views are captured in subsequent waves vary across years. In FY20, risk of seasonal factors which example, when the Union Bud- would have been a downward little more updated but still three months by August to expressed are personal