Dvara E-Registry
- Ashu Sikri (Unlicensed)
Notes from Mar 17 call w Sandeep (CEO). If we continue down the financing path, could be a good partner to engage with though they are pretty locked in to using their own tech platform (Dvaara e-registry) as they feel its comprehensive in what it collects and necessary for driving FPO ops and building lender trust
Dvara is Banking Correspondent (BC) that provides First Loss Default Guarantee (FLDG) and capacity building to FPOs / borrowers
Dvara invests heavily in the orchestration role; coordinate between FPO, buyer and lender; they need to do this as the FPOs capacity to do so themselves is limited
Two products that Dvara is working towards:
“Insurance-like” loan: This is priced like insurance with an up-front “premium” or commitment payment by the farmer and if there is an adverse event, the product pays out enough to cover a farmers’ existing loan (so they don’t end up in default). Its basically a committed pool of money to be drawn when a bad event happens. Apparently this is at a design phase right now with some partner.
Sandeep feels an actual insurance offering can’t work in India bc PMFBY is a bad but heavily subsidized product that distorts the market: its yield-area index based which relies on crop cutting experiments (which are rife with lots of issues) and payouts are set at a village level rather than tied to the experience of an individual farmer.
Further farmers only pay ~2% of the cover amount as premium which means other products which are more tailored but expensive cannot get a foot in the market. Dvara piloted an insurance product whose premium is at 7% of coverage amount but found it cannot compete
Re-wire payment flows so that cash goes from lenders directly to farmers (say, for input purchases) rather than lending to the FPO which is institutionally weak. This is more aspirational, right now, Dvara is managing movement of cash between lenders / FPOs / buyers.
What lenders want to see from Dvara / FPO to issue a loan?
Which farmers am I funding? Over what time period? What’s the price? How secure is the buyer connect (need to see a history of purchases)? How much can the FPO put up as collateral? How much is the FLDG?
Dvara’s take on the value of data to various “data consumers”. Headline is that these actors don’t have the sophistication to work with data but need products that are built using the data that solve a problem:
Lenders: Regional Rural Banks (RRBs), PSUs are not in the picture; Ag focused NBFCs and small biz finance banks don't know how to use this data, want a readymade product. Dvara creates a “khet score”
Insurers: dysfunctional market (see above); govt feels the opportunity lies in digitizing crop cutting experiments using remote sensing which doesn’t address the core issues
Input sellers: have established chains already
Buyers: Data itself is not useful; want help finding/product, make sure grading/sorting is done, get at least a truckload at each visit. This is the “orchestration” work that Dvara is doing right now and its expensive
Is there value in asset financing or is opportunity entirely in working capital?
For cereals, warehouses + sorting / grading infra is used sparingly and utilization will be low so doesn’t make sense for FPO to own
For horticulture, cold-storage has value but Dvara doesn’t currently work with in this space
Notes from Feb 2 demo w Tarun Katoch
What really drives member stickiness with an FPO?
Farmers are most interested in how can you help me sell at better price?
Getting to the first crore of output sales is huge; farmers see that the FPO adds value and have seen groups that go from 1 to 6cr in terms of offtake in a single season (as more members come on-board)
Credit
Inputs
Instilling a sense of ownership and high member engagement are critical for an FPO to succeed
cooperatives in dairy work well bc there are daily transactions
FPOs need to work on internal comms and getting members to participate in governance; Dvara is still figuring this out
Farmers meet local traders regularly and have long standing relationships; how can FPOs re-create this when they are large / diffuse?
Dvara’s model is super hands on; they have teams of people focused on inputs and output marketing including doing tie-ups with suppliers / buyers. Dvara steps in and serve as a mediator when there are disputes over quality btwn FPO and buyer (which happens often)
Currently promoting 25 FPOs in Rajasthan, MP, Odisha, Maharashtra
Critical to have a Dvara person on-site to handle data collection. Tried leaving this to the CEO/BOD/Accountant at the FPO but the data needs to be accurate and timely and they were not able to deliver.
Dvara is charging for their services but not sure exactly how much
Khet score, which is an output of the analytics suite (see below) combines two perspectives: (i) potential of the land (based on nutrient soul condition derived, exposure to flood/drought and other risks from RS) and (ii) discipline of the farmer (track whats happening on the field over multiple years)
Score ranges from 20 to 100; above 50, OK to lend
Some screenshots from their dashboard below. While in theory, this is for the CEO and BOD, reality is that they work on intuition and Dvara has fortnightly calls with the CEO and BOD to go through everything captured in the dashboard and drive things forward
Farmer Profile
List of Nearby Warehouse Providers
Landing Page
Incl a 5 day weather forecast; futures and spot prices
FPO Input Shop
(Apparently this also accessible via farmer facing app as well)
Digitized Plot Boundaries
Notes from Feb 1 call w CEO Sanjay Mansabdar
“FPO stack” demo scheduled for Feb 2
Dvara serves as a BC (banking correspondent) for lenders that want to meet their Priority Sector Lending (PSL) requirements, which is a real challenge for banks
They are also a Cluster Based Business Org (CBBO) and promote / support FPO themselves
They have a digital platform for FPOs and the hook to get farmers to load data is access to credit
Three components to their “Doordrishti” platform
FPO stack: Static (eg, geo-referenced land parcels) and dynamic (crop, sowing and harvest dates) data at the farmer level, aggregated for the FPO and easy to share w input companies and buyers (sounds pretty similar to KDE). They also so some services like soil testing and the results are captured here.
Analytics stack: Layer in remote sensing and imaging data to build a credit perspective; what is the risk / reward of lending to these farmers. The output of this layer is a farmer credit score (“khet score”
Lending stack: This is the BC bit; Dvara acts as a BC for banks (PSUs, regional banks, etc) to help them build a portfolio of farmer loans to meet their PSL targets. Dvara handle various operational bits like documentation, KYC, e-sign, collections, etc.
Two main products (note, didn’t get a sense for total loan volume but Sanjay indicated huge demand from lenders and farmers for this product, will explore further in follow-up call)
Farmer loans
Initially 35-40k rupees check size; increases to 75k by third cycle
Interest rate of 23-25%
Without credit, farmer’s entire decision is driven by local trader who bundles inputs with access to credit and offtake (but at low prices), access to capital enables choice in what inputs to buy/where and where to sell
FPO loans
Still nascent as FPOs as an institution are weak (no systems or governance, poor capitalization, no credit guarantees) and lenders are wary
Dvara leans in heavily, orchestrates the entire transaction and provides a first loss guarantee
The funding is short term and transaction based. Example: Lender pays an input supplier on-behalf of an FPO who sells the inputs to its members and Dvara remits payment back to the lender
Developed an insurance product that automated the claims process via farmer photos and remote sensing but not able to get traction
For context, the PMFBY scheme charges farmers 2% of their cost of cultivation as premium and payout is super “coarse” as it based on delta of yield from 4 random sites in the region relative to historical standard; not tied to the individual farmers realization at all
Dvara’s approach was to tie the payout to the farmers actual realization which they tracked via remote sensing (if you grow paddy in x region, this is the standard yield, if you suffer, payoff based on a ratio) and imaging (farmer self-submit geo-tagged photos, track changes over time)
Dvara not able to offer this product for the 2% floor set by PMFBY (their pricing is closer to 7%) so its stalled
On equipment loans, there is interest in cold storage from FPOs that are working with perishables; for non-perishables, utilization for most equipment is just a couple times a year so not worth owning directly
Sanjay feels a pure digital business is not possible w small/marginal farmers
Dvara model is to place a “Krishak sakhi” at the FPO to ensure effective data collection for its FPO stack otherwise it just doesn’t happen
I asked if Dvara offers FPOs data portability for the info captured in FPO stack; Sanjay claims not much demand for this at the moment but in theory they are fine with it. Their beleif is that the data is useful when tightly integrated (eg, farmer info + analytics + on-ground people to help banks lends into this space)
Also, data monetization requires major scale; an you aggregate across 30 FPOs in a region and present that to an input company; individual FPO is not that interesting