How to deal with the GDP growth variations in forecasts
The Financial Express [
Kochi
]
Bengaluru
,
Chandigarh
,
Delhi
,
Hyderabad
,
Mumbai
,
Pune
,
Kolkata
Jun 4, 2024 | Page No.: 19 | Position: Top Center | Source: Madan Sabnavis | Sq Cm: 390 | AVE: 76068 | PR Value: 380340
Record : 1 / 1
View Text
Merge JPEG
JPEG
Tiff
Word Doc
PDF
1
All Text
How to deal with the GDP growth variations in forecasts INTERESTING OBSERVATIONS FY20 5 0 4 2 3 FY21 -7 7 -7 3 5 8 FY22 9 2 8 7 9 7 FY23 7 0 7 2 7 0 FY24 7 3 8 2 7 qi • -astri ¦ MADAN SASHAY S 3I4ARPUPWARDrevision In the GDP growth number from 7 6 to82aoraises a few pertinent issues tenting todata availability and qualityof foie- casts The practice is tollave the first advance estimate In ame ary fotlowed tYf a second one in lebruansThereare provisional estimates provided in May which show that the economy has done yery well last year as per the table And over time there are final estimates which are captured in subsequent releases These final numbers can be seen when the second advance estimates arc brought out in February and hence the F1 22growth numberwoukl be anchored at thn point of time Thetablelscerighfigimsthrte of thew estinutes smiths from theft nitschsuice esti tut fit° the fowl numberswhich own after twoyears-Mereare interesting utwovannwtobe made fionithe table lbereappentobesubstm• tialdifferennw between the tint advance nainsates whIch come swIlbefore the e trends and the final numbeekthat atrnv after some iterattonsThe prmwaonal numbers bcsonsewherebetween the ones that are released in the mendio Mayttmafterthefinan cialrarendtitscanbeseenthere is no definite pattern and the vases vary across yews In FY20 thetinalnumberwasmuchlower thandwininalcstimatenstulem IY2 I there was improsvment m growth FY22 was different as atterbeingresisaldownwanbin May 2022 it went up haply in the final estimates Hence the sharp revision in the upward direction Iniween lanuary May aticIfinal forecamscanbewrysig- nifkantashichcanbe 1 111g wt The first Issue here is whet bet there is any merit in having an estimate In filtily even before the financial year has ended this Is important because data used would be wadable generally up to November industry inflation or ficcember GST and is based onseeralextrapolanons Hence pushing up numbers for S-9 months to 12 months tuns the risk of seasonal factors which can enwrstate or WI the true picture 14csen Cy the only discernible use of tlus number is in the Budget where revenue numbers are bawd on how the economy has grown in the pie- ceding year and the projection made for the coming yearilas ire necessary as there is neat for sonw basis for making these projections Ilere the lint advancetalmates has use In theabsence of much data the ministry of finance would haw tomake independent fore- casts of bolwprevcars year as well as the coining war Finical on tlw rarenneby would then be even more suhject me and hence the NSO s estimate is requires This conk nin thensk of skewing budgewynumbers but there is no alternative For example when the Union Bud get for 1125 was precentalthe GI PforFY24wasassumed Base at a certain level based on real growth of 7 3 Nov with growth being 8 2 there can upward movement which in turn will provides higher base formakingaprojectionforll25 Rut in FY20 and FY22 there would have been a downward bias And the final numbers would Increasethedenominator for the fiscal deficit ratio in nominaltemukt has loweringit The opposite would hold when denominatorcomes down the next issue is whether there should be second advance estimates Bens the numberis a little more updated but still corms before the yearcomes to an end This forecast can be dis• pensed with as these numbers scald not be used Si any of the major policies of the govern mein This would mean that afterthcfirstadvanceestimates there would be only provisional estimat tilemonthof May The divergencebsiween the thialandprociStonal numbers is a concernfists iNNO because for all mellows the provisional number Is used to explain what has happened in the economy winch is then used forpolgyfor mulatto° But once the final growth numbers vary by as much as Iwo from the pmv ssonalectimates thereareques- t ions of comparability that conic in Ideally t he final nuns bersshould bepublished within three months by August to rod• ambiguity The problem is that since data has been volatile and sub Wet to base effects 111051 expla nations would be erroneous if the final nungwrsdcviate to this ement Thcchallenges for NSO are known and hard to solve inunediatelygiven that the flow of data isunnersAs a large pswt of theeconomy is u norganised getting authentwdata uadal• lenge limier thew circum- stances it would he prudent for policyMAXISand corporates to workwith GI •growth numbers with a range of up to 25 bps on both sides ktearnchsle efforts must be on to narrowdown the time period before the final IltallbOISAMOUL file Uglier a thuf economise Rank of Itanvia Vieses erpn-ssed ore personal