The True Busi­ness Ac­cel­er­a­tor

Power breakfast-cum-Round­table

Banking Frontiers - - Conference Report -

Bank­ing Fron­tiers or­ga­nized a round­table of bankers and in­sur­ers re­lated to the agri sec­tor to dis­cuss the adop­tion of weather re­lated tech­nolo­gies:

Hi­man­shu Goyal: There is a big ex­cite­ment in IBM af­ter the ac­qui­si­tion of The Weather Com­pany about 18 months ago. It will be in­te­grated in a year’s time. This is the world’s sec­ond largest IoT plat­form known to mankind. We gen­er­ate around 26 ter­abytes of data ev­ery day from var­i­ous sources to give a 15-minute up­date on ev­ery half a square kilo­me­ter of plot on the earth’s crust. This sys­tem caters to 26 bil­lion API calls in ev­ery 24 hours.

There were 900 em­ploy­ees in this com­pany. A few of them are the world’s best­known me­te­o­rol­o­gists or math­e­ma­ti­cians with PhD de­grees from the Ivy League in­sti­tutes. These me­te­o­rol­o­gists can take data from any­where, use nu­mer­i­cal and phys­i­cal mod­els and give you an ad­vi­sory on weather. That makes us very con­fi­dent about our ac­cu­racy or fre­quent up­date bet­ter than any­one in the in­dus­try. We are most ac­cu­rate on tem­per­a­ture com­pared to our com­peti­tors.

Awadesh Ku­mar: The big­gest prob­lem in the agri­cul­ture sec­tor is in­for­ma­tion asym­me­try. We don’t know much about the plot. If it is not good for the farmer, it is not good for the fi­nan­cial sec­tor. We can take care of that fi­nan­cial asym­me­try in some way. We can rate each and ev­ery plot in terms of risk and the crop that can grow on it. We can give you a bal­ance sheet of an agri­cul­tural as­set ev­ery 15 days.

Su­mantra Mukher­jee: In the last 18 months, we have im­proved our prod­ucts and have started de­liv­er­ing so­lu­tions to our farm­ers much be­yond in­for­ma­tion. The prod­ucts that we have in­tro­duced are es­sen­tially tech­nol­ogy so­lu­tions, pick­ing up mul­ti­ple lay­ers of data for ev­ery sin­gle farm and pro­vid­ing per­son­al­ized in­ter­ven­tion points and then giv­ing so­lu­tions for those in­ter­ven­tion points to the farm­ers in their own lan­guages. What we do is we hand hold the farmer from the stage where he is se­lect­ing the crop to grow, pro­vide cli­matic zones, de­mand sup­ply chains that have been ob­served for that par­tic­u­lar com­mod­ity and list the most suit­able crops that he can grow. The weather de­tails are fed into the sys­tem and the sys­tem tells us what time and what level of in­ter­ven­tion is re­quired from fer­til­iz­ers to pes­ti­cides, when can a pest at­tack hap­pen and then fi­nally help­ing farm­ers in get­ting max­i­mum re­turn by help­ing them iden­tify the best pos­si­ble mar­ket. We have mapped over 1000 mar­kets in the coun­try and get com­mod­ity price data from them ev­ery sin­gle day on al­most 86 com­mod­ity va­ri­eties, which farm­ers can ac­cess and then de­cide where do they want to go and sell.

Mod­er­a­tor: Do banks have tra­di­tional wis­dom or tech­nol­ogy for weather fore­casts?

P.C. Pan­i­grahi: No tech­nol­ogy for de­ci­sion mak­ing has been taken; only field data and tra­di­tional method are used, es­pe­cially in pre-lend­ing and post lend­ing stage. Type of crops will be based on the type of ge­og­ra­phy and the im­pact of cli­mate. De­ci­sions re­gard­ing shift­ing of crop are taken at the grass­roots level and no weather data is taken.

How gran­u­lar and strin­gent are the bank’s lend­ing strate­gies?

The strat­egy is based on the dis­trict level credit plan­ning pol­icy. It is also based on the Po­ten­tial Link Plan (PLP) de­vised by NABARD, and at the dis­trict level, DLCC de­cides the dis­trict credit plan and it is re­viewed at quar­terly in­ter­vals by the lead bank. Vul­ner­a­bil­ity of the area, in­clud­ing nat­u­ral calami­ties, is also taken as data and used to pre­pare the strat­egy of the credit plan.

What weather guid­ance tech­nol­ogy can help growth in agri-lend­ing?

Tech­nol­ogy such as weather satel­lites and Dop­pler radar are help­ful. Dop­pler radar is ad­di­tion­ally able to in­di­cate the wind di­rec­tion and wind speed. The big data is sourced from satel­lites and weather sta­tions that pro­vide broad scale fac­tors to fore­cast weather changes.

Manoj Rawat: Cou­ple of banks have started us­ing on pi­lot ba­sis fore­cast­ing tech­niques spe­cially for the crops which are long term in na­ture and which are of pre­mium qual­ity, where you can af­ford to have this kind of fore­cast­ing. Banks have started build­ing an­a­lyt­ics but not at the farmer level.

How can weather data im­prove farm out­put?

Manoj Rawat: All we talk about weather is rain­fall. IMB will give it on taluka ba­sis. What we want is weather in­sur­ance. What hap­pened when the hu­mid­ity changes? What hap­pens if the qual­ity of soil changes be­cause of the ex­ces­sive use of fer­til­iz­ers? There are 23 weather re­lated pa­ram­e­ters which the in­sur­ance com­pa­nies have to take into ac­count. Con­sider cumin, which is high value crop. At the har­vest, if sud­denly the hu­mid­ity in­creases the en­tire crop is lost. There­fore, we have to go be­yond rain­fall. We have to go for weather based in­dices.

Which other busi­nesses have weather re­lated risks?

Manoj Rawat: In­dian econ­omy is agrar­ian econ­omy. A sin­gle bad mon­soon and the en­tire econ­omy de­rails. Ev­ery sec­tor gets im­pacted by mon­soon. If a sin­gle mon­soon can de­rail the econ­omy, then we are still an agrar­ian econ­omy.

What lessons have in­sur­ers learnt af­ter los­ing $2 bil­lion in nat­u­ral calami­ties?

Dr Ajay Verma: Agri­cul­ture in­sur­ance has def­i­nitely started tak­ing the note of the weather data, weather in­for­ma­tion, weather fore­cast­ing and re­al­time weather data what­ever is avail­able in the coun­try.

Hi­man­shu Goyal: If we can pre­dict whether for at least 3 months in ad­vance by ma­chine learn­ing and some fac­tual data, that will ben­e­fit many in­dus­tries.

Where do you source weather fore­casts from?

Dr Ajay Varma: In fact, as of now, pre­dom­i­nantly most of the in­sur­ances are de­pen­dent on the data given by IMD in In­dia. To some ex­tent there are also pri­vate data providers. The only thing is how gran­u­lar are these data. In­sur­ance com­pa­nies would like to look at max­i­mum pos­si­ble weather in­for­ma­tion and cov­er­ing the ma­jor weather risk par­al­lels as far as pos­si­ble.

Are weather re­ports a po­tent tool to drive your busi­ness devel­op­ment?

Dr Ajay Varma: Weather re­ports are avail­able in the pub­lic do­main. We take cog­nizance of all those re­ports be­cause we are risk car­ri­ers and are very much sen­si­tive to all. We have to rein­sure all those risks. And rein­sur­ers are watch­ing very closely on all those weather fore­casts. We can­not af­ford to avoid all those re­ports and we would def­i­nitely wel­come more reg­u­lar in­for­ma­tion avail­able to fi­nal­ize our strat­egy.

Which other busi­nesses have risks con­nected to weather?

Pan­i­grahi: Sec­tors af­fected are pri­mar­ily agri­cul­ture, forestry, marine and al­lied ac­tiv­ity. Among in­dus­tries, mostly af­fected are automobile sec­tor be­cause 75% of the peo­ple live in ru­ral In­dia and they have a ma­jor im­pact on automobile pur­chases. Next is tex­tile in­dus­try as it is re­lated to cot­ton. Tele­com, FMCG, banks, in­sur­ance, MFIs NBFCs, ce­ment, con­struc­tion, com­mod­ity mar­ket are the sec­tors which are se­verely af­fected.

How ca­pa­ble are we in fore­cast­ing droughts catas­tro­phe us­ing his­tor­i­cal data?

San­jay Ku­mar Dora: Cli­mate change has emerged as a risk mul­ti­plier, es­pe­cially in the agri­cul­ture sec­tor. We are see­ing catas­tro­phe, which, based on tra­di­tional anal­y­sis or mod­el­ing, is likely to hap­pen once in 100 years. But we are wit­ness­ing drought of that sever­ity or in­ten­sity in 2-3 con­sec­u­tive years. How ca­pa­ble are we to fore­cast the catas­tro­phe, based on this type of cli­mate change?

Hi­man­shu Goyal: There was a dis­cus­sion that all our coastal cities face the dan­ger of go­ing un­der wa­ter. We have data for In­dia for the last 5 years, and for world­wide and US we have data for around 25 years. There are meth­ods by which you can at least pre­dict 6-7 months in ad­vance. But, be­yond weather, there are lo­cal is­sues like wrong trees planted on wrong soil. In Chennai, all the dev­as­ta­tion hap­pened to the trees which were not sup­posed to be on that soil. We are cre­at­ing a huge amount of clut­ter by put­ting con­crete on these cities. If you fly down into Delhi, you don’t see trees from Noida to the point you touch down; you just see some trees in Houz Khas area. It is a man-made is­sue that des­per­ately needs so­lu­tions.

Par­ti­ipants lis­ten to a pre­sen­ta­tion

An in­tense dis­cus­sion in progress

Newspapers in English

Newspapers from India

© PressReader. All rights reserved.