Seven agencies' rating disclosures, parsed into one schema: borrowers, facilities, financials, grade movements, and the lender-wise annexure that names the bank behind each rated facility. Query it from the borrower, the bank, the sector or the downgrade.
Seven agencies Source-linked Floors, not totals Split ratings kept split
From the team behind TatvaMoney — loan origination for Indian NBFCs & DSA networks.
Lender profile
IllustrativeBorrower profile
IllustrativeRating movements
IllustrativeKey financials
IllustrativeThe problem
Every existing route answers “what is this borrower rated”. None of them answers the question a lender actually asks first.
The regulator holds the real picture of large exposures — and restricts it to regulated lenders. An advisory firm, a consultant, an NBFC outside the perimeter, or a bank studying a competitor's book gets nothing.
A deep single-borrower file, priced per report, and the published material describes the lookup running one way: borrower → lenders. Starting from the bank is not what these reports are sold to answer.
Agencies do publish the lender list — inside a press-release PDF, one issuer at a time, one agency at a time. It is public, and it is unusable at scale.
The product
Borrowers, rated facilities, key financials, grade movements and the lender-wise annexure — the same rows the agencies published, parsed into one shape. The difference is that they can be read from either end.
How the market reads it
Pick a borrower, get the banks behind it. One issuer, one agency, one date.
4 LENDERS · ONE ANNEXURE
How we read it
Pick a bank, get every rated borrower whose annexure names it — with rating, band and date.
5,173 BORROWERS NAMING SBI · AS AT 22 AUG 2026
CORPUS AS AT 30 AUGUST 2026 · EARLIEST ACTION MAY 2019 · EVERY ROW TRACEABLE TO A PUBLISHED PRESS RELEASE
Why us
Source agency, source document, publication date. Nothing is inferred from a model we can't point back to a press release.
CARE, CRISIL, ICRA, India Ratings, Acuité, Brickwork and Infomerics — parsed into one schema, so a borrower rated twice resolves to one entity.
Former and erstwhile names are resolved across agencies, so a company that changed its name doesn't become two borrowers.
Coverage bias and staleness are on this page, in the contract, and in the product — not buried in a footnote after you've quoted a number.
Who it's for
Built around the question each desk asks first.
See where a prospective borrower already banks, and how your named exposure compares with the peer set in the same sector and band.
Find which lenders are active in a sector, a band or a ticket size before you take a mandate to market.
Cross-sectional views of the rated book that no single-borrower report can produce.
Early sight of rating migration across a named set of borrowers, across all seven agencies we cover at once.
Coverage
Publicly published credit rating press releases from seven of the credit rating agencies registered with SEBI, parsed field by field.
Entity identity with former names resolved across agencies. Sector and state.
Agency, rating, outlook, action type and date — plus the transition from the prior action.
Instrument type, facility type, amount, and long- or short-term scale.
The lender-wise annexure, parsed row by row: which bank, which facility, how much. Everything here rests on this table.
The figures the agency itself printed — not restated, not modelled.
Strengths, weaknesses and rating sensitivities, as published.
Limits
We would rather lose a deal than have a credit officer misread a figure. Two limits apply to every exposure number here, and neither goes away with better engineering.
Caveat 01 — Coverage bias
When we say “SBI lends to 5,173 borrowers”, what is literally true is: SBI appears as a named lender in 5,173 rating annexures we have ingested, as at 22 August 2026.
Two filters sit between that number and reality. Not every rating action publishes an annexure — measured across the corpus, 43.1% do, and the share ranges from 93.1% at one agency to 0% at another. And the rated book is a minority of any bank's borrowers; most lending never involves a public rating at all.
WHAT WE CAN EVIDENCE TRUE EXPOSURE
The gap is real and varies by bank and sector. We cannot estimate it for you — the bar above is schematic, not a measurement.
Caveat 02 — Staleness
A lender-wise annexure reflects the sanctioned position as at the rating action date — not today, and not the outstanding balance.
SEBI requires each agency to set its own review cadence in its operations manual rather than fixing one interval, so a record is commonly a year or more old — older still where the issuer stopped cooperating. A facility repaid the day after the action shows as sanctioned until the next review.
One record's lifecycle. A record you read today sits somewhere in the shaded span.
TatvaRatings is not exposure monitoring, not a system of record, and not a control for any process requiring a current or complete exposure position on a named borrower. Use it to see the shape of the market, find counterparties, size a sector and generate questions — then verify through your own channels before acting on a name.
FAQ
Credit rating press releases published by seven of the credit rating agencies registered with SEBI. SEBI mandates a standardised format and a public archive, so the source is a regulatorily-required public disclosure. Every row keeps the agency, the document and the publication date.
No, and it must not be used as one. CRILC is the regulator's own near-complete picture of large exposures, restricted to regulated lenders. This is the published, rated slice of the market — a floor built from public documents, available to institutions CRILC excludes.
It reflects the sanctioned position at the rating action date, not today and not an outstanding balance. SEBI requires each agency to set its own review cadence in its operations manual, so a record is commonly a year or more old, and older where the issuer stopped cooperating. Every figure carries its date.
Because they are floors. 46.2% of rating actions publish a lender-wise annexure — 32,498 of 70,337, measured as at 30 August 2026 — and the rated book is a minority of any bank's borrowers. A figure means “at least this much, from public record” — never “this much”.
No — deliberately. Nothing you hold enters the platform. That keeps the service outside the material-outsourcing perimeter and means there is no customer lending data here to lose.
Through a conversation. There is no self-service signup and no public sandbox. Access is scoped to your organization, and your own administrator provisions users inside it. Sign-in is a one-time code to your work email — no password, no shared credential.
Get access
Tell us who you are and what you'd want to look up first. We'll come back with what our coverage can and can't answer for that question — including where it can't.
Sold and provisioned directly. No self-service signup, no public sandbox.
Your own administrator provisions users and controls what each role sees.
Sign-in by one-time code to your work email. No password to rotate.