Rabi Lending Starts With Water: What Reservoir Storage Tells a Bank Before Sowing Begins

Rabi Lending Starts With Water: What Reservoir Storage Tells a Bank Before Sowing Begins

AgribusinessBusinessDataDataStoreIndiaRabiWeatherYield
Oct 06, 2026
18 views

A Rabi crop loan is sanctioned in October, weeks before the crop it finances is fully established. The borrower's repayment capacity, however, is largely set by something the sanction note rarely mentions: how much irrigation water will actually be available between November and March. After a deficient monsoon, that quantity is not a matter of opinion. It is published, reservoir by reservoir, before sowing begins — and this year it is materially tighter than the year a lender's portfolio last performed against.

The national storage number, read properly

The Central Water Commission monitors 178 major reservoirs across India. Its weekly bulletin for the start of October put combined live storage at nearly 72% of capacity — 20 percentage points below the same point last year and 12 points below the ten-year normal. In volume terms, the bulletin summarised by SANDRP records 132.734 billion cubic metres (BCM) of live storage, against 166.856 BCM a year earlier and a ten-year average of 151.118 BCM. Only 18 of the 178 reservoirs were full, and 51 were half-empty or worse.

For a credit team, the percentage-point gap is the operative figure, not the headline. A portfolio priced and provisioned on last season's water position is carrying roughly a fifth less stored irrigation buffer into this Rabi season — before a single new loan is disbursed. The India Meteorological Department has additionally forecast below-normal rainfall for October, the first sowing month; that is a forecast, not an outcome, but it removes any prudent basis for assuming an early refill.

Why the average is the wrong unit for a loan book

Storage is not distributed the way portfolios are reported. In the same CWC bulletin, three of India's five regions held storage above 80% of capacity, while the southern region remained below 50% and the northern region had slipped to 60.5%. The central region sat below last year yet above its own long-term average, and eastern-region reservoirs held more than both last year and normal. A state-level or national average blends commands with adequate water and commands in genuine deficit into a single reassuring number.

A crop loan, though, is not repaid by an average. It is repaid out of a specific command area, a specific borewell cluster, or residual soil moisture in a specific rainfed district. The correct lending unit this season is therefore the reservoir command and the district, not the state. Two branches in the same state can face opposite Rabi risks if one lends into a command fed by a reservoir at 85% and the other into a command fed by one below half.

The government is planning Rabi on a water budget — lenders should underwrite on one too

This is not only a risk-team inference; it is now the stated planning basis for the season. At the National Agriculture Conference for the Rabi campaign on 28 September, the Agriculture Ministry asked every state and district to assess irrigation water availability — a water budget — before finalising crop plans, with limited-water areas directed towards less water-intensive crops such as pulses and oilseeds and assured-irrigation areas planning other crops accordingly. The same release records two points that belong in a lender's Rabi review: Kisan Credit Card coverage remains inadequate in several areas, particularly among small and marginal farmers, with states asked to examine bottlenecks in loan sanctioning at the banking level; and states were asked to monitor crop conditions, sowing progress and market prices weekly through the season.

Read together, those points define the lender's task precisely: expand timely Rabi credit where the water budget supports the crop plan, and differentiate — by crop, district and command — where it does not. Undifferentiated growth and undifferentiated caution both misprice the season.

Four underwriting adjustments for this Rabi

Map the book to commands, not just districts. Overlay branch-level Rabi exposure on the reservoirs and canal commands that serve it, using the CWC position as the opening state. Exposure concentrated in below-50% commands gets flagged before disbursement, not after the first missed irrigation.

Sanction the crop the water can finish. A water-intensive crop plan in a low-storage command is a higher-risk proposal than the same borrower growing pulses or oilseeds on the same land, because the failure mode is mid-season water exhaustion after most input costs are already sunk. Crop choice is observable at sanction; it should move the assessment, the limit, or both.

Stage disbursement against the season, not the calendar alone. Reservoir drawdown, October rainfall realisation against the forecast, and sowing progress are all published or observable during the disbursement window. Tranches linked to those checkpoints cost little and convert a one-time sanction decision into a monitored exposure.

Treat rainfed exposure as a separate book. Reservoir data says nothing directly about rainfed districts, where Rabi outcomes rest on residual soil moisture left by a monsoon that left about 45% of the country with deficient rainfall. Rainfed Rabi lending needs its own moisture and sowing-progress signals, and should not borrow comfort from a canal-command storage number — or vice versa.

The limits of the signal deserve equal clarity. Reservoir storage is shared between irrigation, drinking water and other uses, and releases are an administrative decision, so a storage percentage is an availability indicator, not a delivery guarantee to any particular outlet. Groundwater depth, power supply for pumping and the borrower's own water management all mediate the final outcome. Storage data tells a lender where to look harder and earlier; it does not replace looking.

Where Farmioc fits

Farmioc DataStore serves the financial sector with agricultural data and analytics — supply and demand, yield and harvest outlooks, and satellite and ground-based data organised for portfolio use. Joined to a lender's own book, that is the layer that turns a national reservoir headline into the command-level and district-level view a Rabi sanction actually needs: Farmioc DataStore organises those signals by crop and region so risk teams can see which parts of the portfolio the water position supports, and which parts it does not, while there is still time to act on the difference.

Rabi credit will be disbursed this month either way — the season does not wait. The only choice is whether it is disbursed against last year's assumptions or against this year's water.

Farmioc is India's agriculture data analytics and agri trading company. Explore DataStore for the financial sector, or PriceHub for commodity market intelligence across 200+ crops.