RBI Project Finance Directions 2025 and Your Insurance Covenants: What Corporate Borrowers Must Now Carry

The RBI/2025-26/59 Reserve Bank of India (Project Finance) Directions, 2025 were issued on 19 June 2025 and took effect on 1 October 2025. They replace a patchwork of earlier circulars with a single framework covering how banks and NBFCs appraise, monitor, and provision for project loans across construction and operational phases. For a corporate borrower, the practical effect is not abstract. Your lenders have rewritten the insurance covenants in their common loan agreements to match the new regime, and those covenants now bite harder at disbursement, at each review date, and at any date the commercial operations date is pushed out.

Three features of the Directions flow straight into insurance. First, the date of commencement of commercial operations (DCCO) is now a monitored, documented milestone with defined extension limits, up to three years for infrastructure projects and up to two years for non-infrastructure projects where the delay is beyond the borrower's control. Second, disbursement is conditioned on stricter verification of project readiness, which lenders operationalise partly through insurance conditions precedent. Third, provisioning during construction is calibrated to project stage, so lenders police anything that could impair recovery, and an uninsured or under-insured asset is exactly that.

The borrower-side reading matters because the covenant is your obligation, not the bank's. If your builder's-risk policy lapses, if the sum insured drifts below reinstatement cost, or if the agreed bank clause names the wrong lender, you sit in breach even when no loss has occurred. This post walks the covenant stack a borrower must now carry, from material-damage cover through delay-in-startup protection to the mortgagee-interest and agreed-bank-clause obligations that make the whole structure enforceable for your lenders.

The single most common covenant failure under the new regime is a timing mismatch. Your construction-phase material-damage cover, whether written as Contractors' All Risks (CAR) or Erection All Risks (EAR), is bought to a policy period tied to the original construction schedule. Your Delay in Start-Up (DSU), also sold as Advance Loss of Profits (ALOP), cover is written with an indemnity period and a scheduled DCCO baked in. When the Directions permit a DCCO extension of up to three years for an infrastructure project, and the lender grants it, both policies must be extended in lockstep or the covenant is broken.

The mechanics deserve care. EAR material-damage cover must be extended by endorsement to the revised completion date, with the sum insured reviewed to reflect actual escalated project cost, not the day-one contract value. DSU indemnity depends on physical loss or damage under the underlying EAR section, so a DSU claim for a delayed DCCO only responds if the trigger event is an insured peril, not merely a commercial or financing slippage. A borrower who assumes DSU covers a regulatory extension will be disappointed.

Brokers should map the DSU indemnity period against the lender's revised repayment start and the debt-service reserve assumptions. If the indemnity period is shorter than the realistic ramp-up to revenue, the borrower carries the residual gap directly, and lenders increasingly ask to see that arithmetic before they approve the extension.

The Directions sharpen a lender's focus on recovery, and the two clauses that protect a financier's stake in an insured asset are the mortgagee interest clause and the agreed bank clause (also called the financier's interest or bank clause). A corporate borrower must understand both because it is the borrower who buys the policy and instructs the insurer to attach them.

The agreed bank clause directs that any claim payment be made to the lender, or jointly, and that the insurer notify the lender of non-payment of premium, material alteration, or cancellation before those events prejudice the cover. It preserves the lender's security in the policy proceeds. The mortgagee interest clause goes further in some wordings by protecting the lender's interest even where the borrower's own conduct, such as a misrepresentation or breach of warranty, would let the insurer avoid the policy against the borrower. The two are not interchangeable, and lenders under the 2025 regime increasingly specify which they require, on which policies, and naming which security trustee or facility agent.

For a borrower this creates a documentation discipline. Every material-damage and DSU policy on a financed project must carry the correct clause, name the current lender or security trustee exactly, and reflect any change in the lending consortium. A syndicated facility that adds or refinances a lender mid-construction requires the endorsement to be updated, or the new lender is unsecured in the insurance proceeds. Insurers price and word these clauses differently, so the borrower cannot assume a market-standard form will satisfy a specific loan agreement. Read the covenant, then read the wording, then close the gap by endorsement.

Under the tighter monitoring in the Directions, lenders convert insurance requirements into hard conditions precedent and conditions subsequent. A borrower will not draw a tranche until the lender's agent confirms that the required policies are in force, at the required sums insured, with the correct clauses attached and premium paid. This is a shift from a soft covenant policed at annual review to a gate on cash.

The practical checklist a borrower must satisfy at each drawdown typically includes: material-damage cover (CAR or EAR) for the full contract value plus escalation and debris removal; DSU or ALOP cover with an indemnity period matching the lender's debt-service assumptions; marine and marine-DSU cover where critical equipment is imported, because a delayed shipment can move the DCCO as surely as a site fire; third-party and public-liability cover to the contractually specified limit; and the agreed bank clause or mortgagee interest clause on each policy that secures the lender.

Borrowers should build an insurance conditions-precedent tracker that mirrors the loan agreement, held jointly by the treasury team, the broker, and the lender's agent. A missing certificate of insurance or an out-of-date endorsement stalls the tranche, and on a construction schedule a stalled tranche cascades into liquidated damages under the EPC contract. The insurance covenant, once an afterthought, is now on the critical path.

The reason lenders police these covenants more aggressively is that the Directions tie provisioning to project stage and to the credibility of the completion plan. During construction, banks carry a standard provision on project loans, stepping down as the project reaches DCCO and demonstrates cash flow. Anything that raises the probability of loss, an uninsured asset, an under-insured sum, a lapsed policy, feeds directly into the lender's risk assessment and, in a stressed case, into higher provisioning and classification pressure.

For the borrower this changes the negotiating dynamic. Insurance is no longer a box the relationship manager ticks. It is a variable the lender's credit and monitoring teams watch because it protects the recoverable value behind their provision. A borrower who lets the sum insured drift below reinstatement value, or who fails to update cover after a cost overrun, is not just technically in breach. The borrower is degrading the lender's collateral position at exactly the moment the Directions require the lender to scrutinise it.

Under-insurance is the quiet risk. Most material-damage policies carry an average clause, so a sum insured set at eighty percent of reinstatement cost reduces every claim payment proportionately, not just claims above the shortfall. On a large infrastructure asset that proportional cut can run to crores, and it lands on the lender's security. Reinstatement-value cover and periodic sum-insured reviews against escalated project cost are therefore not optional refinements. They are how a borrower keeps the covenant honest and keeps the lender's provisioning stable, which in turn preserves the borrower's own access to further disbursement.

Bringing the covenant stack together, a corporate borrower financing a project under the new regime should be able to answer yes to each of the following at every review and every drawdown.

1. Period alignment: Does the CAR/EAR and DSU/ALOP policy period run to the current lender-approved DCCO, including any extension granted under the Directions, with no gap in cover?
2. Sum insured integrity: Is the material-damage sum insured set to full reinstatement value including escalation, debris removal, and professional fees, so the average clause cannot bite on a claim?
3. DSU indemnity fit: Does the DSU indemnity period match the lender's debt-service and ramp-up assumptions, and is the trigger tied to insured physical damage rather than commercial delay?
4. Financier clauses: Does every financed policy carry the correct agreed bank clause or mortgagee interest clause, naming the current lender or security trustee and updated for any consortium change?
5. Premium and 64VB: Is premium paid and receipted so risk has validly attached before the lender disburses against the cover?
6. Imports and transit: Where critical equipment is imported, is marine and marine-DSU cover in place so a shipment delay does not silently move the DCCO?

This discipline is not busywork. It is the difference between a clean drawdown and a stalled one, and between a lender who treats your project as low-monitoring and one who tightens terms at the next review.

The recurring theme across the 2025 Directions is that generic assumptions about cover no longer survive contact with a lender's monitoring team. A loan agreement may require a mortgagee interest clause while your broker has placed only an agreed bank clause. A DSU wording may exclude the very delay cause your project is most exposed to. An average clause may sit quietly in a material-damage policy whose sum insured has not been revised since financial close. Each gap is invisible until a claim or a review exposes it, and by then the covenant is already breached.

Closing these gaps requires reading the actual insurer wording against the exact covenant text, clause by clause, rather than trusting that a market-standard policy meets a bespoke loan agreement. That comparison is slow when wordings sit in scattered PDFs and slower still across a portfolio of financed projects. Sarvada makes insurer policy wordings searchable, so a broker or a borrower's risk team can pull the exact agreed-bank-clause language, the DSU trigger definition, or the average-clause provision across insurers and check it against the covenant in minutes rather than days. If your team is aligning project-finance insurance covenants to the RBI Directions and wants faster, wording-level answers, Request Access to see how Sarvada supports that work.