PMFBY Crop Insurance Broking: Commission, State Tenders and Cut-Off Dates

A commercial broking firm approaching the Pradhan Mantri Fasal Bima Yojana (PMFBY) for the first time usually misreads it as a very large placement. It is not. Crop business is allocated by state governments through procurement, and the unit of allocation is not a client but a cluster: a group of districts packaged into one lot and tendered to general insurers for a fixed number of seasons.

A state's agriculture department notifies the crops, the insurance units and the scale of finance that fixes sums insured, then invites actuarial premium rate bids from empanelled general insurers for each cluster. Insurers quote a percentage of sum insured, crop by crop, and the lowest quoted rate takes the cluster. Clusters are increasingly tendered on a three-year cycle, so one bid decision can bind an insurer to a district group across six crop seasons.

Three consequences follow for a broker.

• The premium is fixed at bid opening. Once the L1 actuarial rate is locked, no placement skill moves the premium base. Everything downstream, including whatever an intermediary earns, is a residual inside a number decided before the season starts.
• The buyer is a government department, and the insured is somebody else. The state drives scope, timelines and payment. The farmer holds the cover. Those are different parties with different grievances, and a broker answers to both.
• The obligation is seasonal and unconditional. A cluster award carries enrolment, assessment, claim and grievance duties across every notified crop in every notified insurance unit, whether 4,000 farmers enrol or 4,00,000.

The general dynamics of bidding for government insurance work, and why the lowest-quote format punishes optimistic costing, are set out in government and PSU account broking. This piece does not repeat that argument. Crop differs for three specific reasons: the product's design is statutory rather than negotiated, the enrolment window is measured in weeks, and the money arrives from two governments rather than one client.

The farmer's contribution is capped by the scheme, not by the market. The long-standing caps on the farmer's share of sum insured are:

• 2 percent for Kharif food crops and oilseeds
• 1.5 percent for Rabi food crops and oilseeds
• 5 percent for annual commercial and annual horticultural crops

Everything between that cap and the L1 actuarial rate is subsidy, shared between the Centre and the state. The conventional split is 50:50, with a more favourable central share for North Eastern states. Since the 2020 revamp the central contribution has been restricted to a ceiling expressed as a percentage of the actuarial rate, higher for unirrigated areas than for irrigated ones, with any excess falling to the state if it still wants the cluster covered at the winning bid rate.

That structure produces the single most important number in crop broking economics. On a cluster where the L1 actuarial rate lands at, say, 14 percent of sum insured for a Kharif oilseed, the farmer funds 2 percentage points and the two governments fund 12. Roughly six-sevenths of the premium on that crop is a government receivable, not a customer receivable. The farmer's rupee is collected at enrolment. The other twelve arrive when a budget line moves.

Crop premium is therefore not one receivable with one ageing profile. It is three: the farmer share, certain and collected at enrolment; the central share, released against a portal reconciliation; and the state share, released against a state budget. Modelling crop cash flow as a single premium line models it wrong, because the three tranches do not behave alike and the state tranche is the one that slips.

The reform package addressed this at source. From Kharif 2025-26, states must deposit their premium share in escrow accounts so funds are available when claims settle, and from Kharif 2024 a 12 percent penalty is auto-imposed on insurers for claim delays. The PMFBY reform package covers that architecture in detail.

Commercial broking runs on renewal dates that are negotiable within a few days. Crop runs on enrolment cut-off dates that are not negotiable at all. A farmer who is not enrolled on the National Crop Insurance Portal before the notified cut-off for that season and that crop is uninsured for the season, whatever the merits.

The rhythm is fixed by the crop calendar. Kharif enrolment closes in the weeks after the monsoon's onset, conventionally at the end of July; Rabi closes around the end of December. The operative date for any season, state and crop is whatever the state notification says, and extensions do get issued, but the planning assumption should be a short window inside which data, consent, premium debit and portal upload must all clear.

Loanee and non-loanee are two different operations

• Loanee farmers hold a Kisan Credit Card or seasonal crop loan. Their premium share is debited by the financing bank and uploaded in bulk. Since Kharif 2020 the scheme is voluntary for them, so a bank must record an opt-out rather than assume participation. Volume is high, acquisition cost is near zero, and the failure mode is data quality: wrong survey numbers, wrong crop, wrong area, wrong insurance unit, discovered at claim time.
• Non-loanee farmers enrol individually, through Common Service Centres, field networks, or the portal. Volume is lower, acquisition cost is real, and the failure mode is reach: the farmers who most need the cover are the hardest to enrol before a July cut-off.

PMFBY is not an indemnity contract in the sense a commercial broker is used to. For the main yield cover it is an area-yield index product, and understanding that distinction is the difference between advising on it competently and misselling it.

The insured object is not the individual field. It is the insurance unit, usually a village or village panchayat for major crops and a larger area for others. The contract compares the season's actual yield for that unit against a threshold yield derived from the unit's historical yield series and an indemnity level (commonly 70, 80 or 90 percent). If actual yield falls below threshold, every insured farmer in the unit receives a payout proportionate to the shortfall. If it does not, no farmer in that unit is paid, however bad that farmer's own field was.

That is basis risk carried by design, and it is what farmers complain about most. A broker cannot engineer it away. What a broker can do is be precise about it in every conversation with the state, the bank and the farmer, because a farmer who understood the index at enrolment lodges a different grievance from one who thought he had bought damage cover on his own plot.

The CCE is the measurement, and the measurement is the claim

Actual yield for the insurance unit is established through Crop Cutting Experiments (CCEs): physically harvested sample plots, conducted by the state agriculture department under prescribed protocol, with a minimum number of experiments per insurance unit per crop. The CCE is therefore the operative event of the entire contract. A CCE conducted late, conducted in the wrong plot, or not conducted at all does not produce a small procedural problem. It produces an unassessable insurance unit.

This is what YES-TECH, the remote-sensing yield estimation system, was introduced to reduce dependence on. It became mandatory for paddy and wheat from Kharif 2023 with soybean added in Kharif 2024, moving part of the yield basis from a harvested sample to a technology-derived estimate. The point for anyone advising in this line is that the yield basis, whether harvested or modelled, is a state and scheme process. The insurer carrying the cluster does not control it, and neither does an intermediary.

Here is the honest position, and it is less flattering than the premium numbers suggest.

Farmer-facing distribution of PMFBY is not broker territory. Loanee enrolment runs through the financing bank. Non-loanee enrolment runs through Common Service Centres, the portal, and insurer field staff. A broking firm is not the natural channel for a 2 percent farmer contribution on a one-hectare holding, and the scheme was not built to fund one.

Where a broking firm does appear is on the state's side of the table, and the mandate there is consultancy-shaped rather than placement-shaped:

1. Tender design and evaluation support: cluster construction, bid document drafting, evaluation of actuarial rate bids, and comparison against the state's own subsidy exposure at each bid level.
2. Scheme administration oversight: monitoring enrolment against the cut-off, reconciling portal data, tracking CCE completion by insurance unit, and chasing settlement timelines against the insurer.
3. Grievance and audit support: the state carries the political cost of unpaid farmers, and it is the party that most needs an independent reading of why a cluster's claims did or did not trigger.

Firms that do this well treat crop as an advisory practice attached to a state relationship, resourced by people who can read a yield series and a CCE schedule, and costed against a three-year cluster cycle. Firms that do it badly treat it as premium volume and discover in season two that the volume was never theirs.

Whatever the payment basis, the timing problem is the same, and it is worse in crop than anywhere else in Indian insurance.

In a private placement, commission ages against a single premium payment from a single client under Section 64VB discipline, so the outer bound of the wait is knowable. In crop, the premium is assembled from three sources across months, and any payment that is a percentage of premium inherits the slowest of the three. The escrow requirement from Kharif 2025-26 exists precisely because states were not funding settlement on time.

The implications for working capital:

• Age crop receivables separately. A crop line inside a general commission ageing schedule distorts the whole schedule's days outstanding and hides which accounts are actually slow. The mechanics of that schedule are in commission receivable ageing; the crop-specific instruction is to give it its own bucket.
• Do not treat a season as a year. Two crop seasons with different cut-offs, subsidy release patterns and claim timelines produce a cash profile that has nothing to do with April-to-March. Model it by season and let the financial year fall out.
• Fee mandates age better than percentage mandates. A defined fee invoiced against milestones (tender award, enrolment close, CCE completion, settlement sign-off) is collectible on the state's procurement cycle. A percentage of premium is collectible on the subsidy cycle. Those are not the same cycle, and one is materially longer.
• Price the three-year cycle, not the first season. A three-year cluster mandate is six seasons of enrolment reconciliation, CCE tracking, grievance handling and claim chasing. Cost all six before quoting on the first.

A short pre-bid discipline for any firm considering crop work, sequenced the way the season is.

1. Read the state notification, not the scheme summary. Notified crops, insurance units, scale of finance, indemnity level per crop, cut-off dates, and the cluster's district composition are all state-level facts. Two neighbouring states run the same scheme with materially different parameters.
2. Pull the cluster's yield history. Threshold yield derives from the unit's historical series. A cluster whose recent seasons sit near or below threshold is not a mandate with occasional claims; it is one with claims every season and a grievance queue to match.
3. Confirm the payment basis in writing. Percentage or fee, payable by whom, against what event, and whether both are permitted. Confirm before the bid, because nothing about a government mandate reprices after award.
4. Cost the enrolment window honestly. The July and December weeks are where the year's labour is concentrated. Staffing for the average month is how firms miss cut-offs on somebody else's data.
5. Map the CCE dependency. Which department conducts them, on what schedule, and what the last two seasons' completion rate looked like by insurance unit. Chronic CCE gaps mean contested claims regardless of how the season went.
6. Decide the strategic value explicitly. A crop mandate can be worth thin economics for the state relationship it builds and the agri-adjacent private business (agribusiness property, cold chain, food processing, contract farming) that follows. That is defensible. It should be a written decision with a rupee figure attached, not a discovery made in season two.

Crop is one of the largest lines in Indian general insurance and one of the least like the rest of it. The premium is set by a tender, the claim is set by a measurement nobody in the contract controls, the money comes from two governments, and the deadline is a date on an agricultural calendar. A firm that understands all four can build a durable practice around state relationships. A firm that understands only the premium number will bid on the wrong thing.