How to find which companies a bank lends to

The public routes to a bank's borrowers: rating annexures, exchange and annual-report disclosures, and the MCA charge index — and where each one stops.

Guide Last reviewed 7 min read

On this page — 7 sections

The short answer: no public source publishes a bank’s borrower list, and the supervisory data that comes closest — the RBI’s CRILC — is closed to non-lenders. What the public record does hold is partial: lender-wise annexures in published rating press releases, company-side disclosures in exchange filings and annual reports, and the MCA’s charge index. Each names real lending relationships; none is complete.

This page walks those three routes in depth. It is written for the reader behind the “[bank] exposure list” search — an arranger, an adviser, an NBFC credit team, a fintech — who needs to know what can actually be assembled from public record and on what terms.

One boundary first, so the rest reads honestly. A bank’s actual exposure data sits in two closed places: the bank’s own book, and the RBI’s CRILC, a supervisory return whose access runs with RBI-regulated lending status. Everything below is about the public record, and every figure it yields is a floor — the amount evidenced by published documents, never the true position.


Route 1 — Rating annexures: the richest public source

When a SEBI-registered rating agency rates a company’s bank facilities, the published press release frequently ends with a lender-wise annexure: a table mapping each rated facility to the named bank providing it, with the sanctioned amount. The reason is prudential — a bank loan rating is used for the lender’s capital computation, so the rating has to attach to a specific facility of a specific lender — and the consequence is that public documents, published free on the agencies’ own websites, name banks against borrowers at facility level.

What one annexure row actually carries

A worked example, with every value invented for illustration — this is the shape of a row, not a rendering of any real document:

Lender Facility Sanctioned amount Rating
Bank of AlphaCash credit₹18.00 croreBBB− / Stable
Bank of AlphaTerm loan₹12.50 croreBBB− / Stable
Beta BankBank guarantee₹4.00 croreA3
Proposed / unallocated₹5.50 croreBBB− / Stable

Read against the rating action date printed on the release, that row set tells you: which banks are named on the rated facilities, the type of each facility, and the sanctioned limit on each line. It does not tell you the outstanding balance, whether the limit is drawn at all, or anything about the borrower’s unrated facilities — and the proposed/unallocated line, which routinely appears with no lender named, must be excluded from any total or the position is overstated.

Where this route stops

  • Consent gates it. Disclosure of lender details depends on the issuer’s consent. Across our corpus, 46.2% of rating actions carry at least one lender row (32,498 of 70,337, as at 30 August 2026), and the share varies sharply by agency — India Ratings publishes no lender dimension at all.
  • Rated borrowers only. Most Indian companies have never been rated, and the rated book is a minority of any bank’s borrowers.
  • Dated, not current. An annexure reflects the sanctioned position as at the rating action date, on a broadly annual review cycle. A typical record is a year or more old when you read it.

Route 2 — Exchange disclosures and annual reports: the company-side record

The relationship is sometimes disclosed from the borrower’s side.

Exchange filings. A listed company, or an unlisted company with listed debt, files material events with NSE and BSE, and both publish corporate announcements free. Rating actions, significant facility agreements and disclosure-triggering defaults surface here, sometimes naming the lender. Listed issuers only, and event-driven rather than a standing statement.

Annual report borrowings schedules. Audited financial statements carry a borrowings note, and it frequently names the banks behind term loans and working capital facilities, with security and repayment terms. Once a year, one company at a time, and naming practice varies — many schedules give amounts by category with no lender named. For a private company, the statements come via MCA filings rather than a website.

This route runs company-first: it can confirm a suspected relationship, but you cannot start it from the bank.

Route 3 — The MCA charge index: public, broad, shallow

The Ministry of Corporate Affairs publishes an index of charges registered against companies, naming the charge-holder — typically the lending bank — and the amount secured. It is open to public users and covers companies whether rated or not, which makes it the broadest of the three routes.

It is also the shallowest. A charge evidences a secured lending relationship; it carries no facility type, no limit structure, nothing on unsecured lending, and satisfied charges can linger on the record. Use it to establish that a relationship exists, not to characterise it.


Where scale becomes the problem

For one borrower, the routes above are an afternoon’s work. But the question this page is titled with runs the other way — from the bank — and the public record is not organised that way. The richest source, the annexure, is published one borrower per PDF, across seven agencies with seven archives, seven layouts and no cross-agency index, with the bank’s name rendered differently document to document: full legal name, initialism, branch suffix, or the legacy name of a merged institution.

Assembling “which rated borrowers name this bank” by hand means reading every release, finding every annexure, and reconciling every lender-name variant. That is not a lookup; it is a document-processing project.

What TatvaRatings does with exactly that problem

TatvaRatings parses publicly published rating press releases from seven SEBI-registered agencies into one schema, parses each annexure row by row, resolves borrower and lender identities across the corpus, and indexes the annexure from the lender’s side — so the query starts with the bank. That is the whole product, and lender exposure analysis describes its direct use.

179,857
Lender-exposure rows
1,501
Distinct lenders
70,337
Rating actions
45,528
Borrowers
194,471
Rated facilities

Measured as at 30 August 2026. Earliest rating action held is from May 2019. Every count is a floor, not a total.

The limits of Route 1 carry through unchanged, because they are properties of the source, not of the parse: rated borrowers only, 46.2% annexure availability, sanctioned amounts as at each action date. The full specification is on Data limits. What the index changes is not what the public record contains — only that a question which was unaskable at scale becomes a query.


Common questions

Can I get a bank's complete borrower or exposure list?

No. No public source publishes one. A bank's exposure data sits in its own book and in the RBI's CRILC, a supervisory return whose access is confined to RBI-regulated reporting institutions — there is no public tier and no application route for a non-lender. What the public record holds is partial: rating annexures, company-side disclosures, and the MCA charge index, each described above, each a floor rather than a total.

Which public source names the most bank–borrower relationships?

At facility level, the lender-wise annexure in rating press releases — it names the bank, the facility type and the sanctioned amount. But it exists only for rated borrowers, and only where the issuer consented: 46.2% of rating actions in our corpus carry one, as at 30 August 2026. By breadth of companies covered, the MCA charge index reaches further — rated or not — but carries no facility detail.

Is the data current?

No, and no public route is. An annexure reflects the sanctioned position as at its rating action date, on a broadly annual review cycle, so a typical record is a year or more old. Annual reports are yearly; charge filings persist until satisfied on the record. A facility repaid the day after a rating action still reads as sanctioned until the next review publishes. Anything built on these sources is market intelligence, not exposure monitoring — the distinction is stated in full on Data limits.


Tell us the borrower, bank or sector you would start from. We will tell you what the current corpus can and cannot answer for it.

Talk to us