This page explains where VIGIL's credit rating data comes from, how each rating is linked to a listed company, and which rating actions VIGIL marks as red flags. It describes the rules in plain words. See also how the Signals table uses these flags and how the data is validated.
Where the data comes from
Rating agencies file every rating action on rated debt with NSE. NSE publishes these filings in its centralised rating database: NSE Corporate Filings - Credit Rating. VIGIL takes its credit rating data from there.
Seven agencies file there: CRISIL, CARE, ICRA, India Ratings, Brickwork, Infomerics and Acuité. NSE spells Acuité in two ways in its own data. VIGIL shows both spellings as NSE publishes them.
VIGIL checks NSE for new filings several times a day. The history starts with filings published in January 2024. Some rows carry an earlier rating date because the agency filed an older rating late.
What one row is
One row is one rating action, on one debt instrument, by one agency. The instrument can be a bond, a debenture, commercial paper or a bank loan facility.
A company with many bonds has many rows. A single rating review usually produces several rows on the same day, one for each instrument the agency rated.
Each row has the new rating, the outlook, the action (for example upgrade, downgrade or reaffirm) and the previous rating where the agency gave one.
How a rating is linked to a company
Many rated issuers have no listed shares. These include unlisted companies, finance companies, state bodies and trusts. Their ratings appear on the Ratings page. They have no NSE symbol, so they do not appear on a company page or on Signals.
For an issuer with listed shares, VIGIL links each rating to its NSE symbol in this order:
- Issuer code. Every bond has an ISIN, and its first seven characters identify the company that issued it. VIGIL matches that code to the company's listed shares and checks that the two names agree.
- Company name. Some bonds were issued by a company that later merged into its parent. Their code still points at the old company, which no longer trades. If the filing names a company that already has a symbol on its other ratings, VIGIL uses that symbol. It does this only when the bond is registered in that same company's name.
- A short checked list. A few companies cannot be placed by the first two steps, for example a recent listing. These are linked by hand.
- No match. Otherwise the rating stays without a symbol.
VIGIL does not move a rating to a different company. Two cases come up in the filings:
- A filing names one company but gives the ISIN of a different company. VIGIL leaves that rating unlinked.
- A subsidiary has its own rating. That rating stays with the subsidiary. It is not shown under the listed group company.
What counts as a red flag
VIGIL reads three fields on each row: the rating action, the outlook and the rating itself. It checks them in that order, and the first one that points to trouble sets the flag.
| Flag | What it means |
|---|---|
| Downgrade | The agency lowered the rating. |
| Default | The agency says the issuer is in default. |
| Default grade | The rating itself is D, the default grade. |
| Negative outlook | The agency expects the rating may come down. |
| Watch, negative | The rating is under review and may come down. |
| Watch, developing | The rating is under review and could move either way. |
| On watch | The rating is under review and the agency gave no direction. |
| Watch inside the rating | The agency wrote the watch status into the rating itself, for example CARE BBB (RWD). |
| Speculative grade | The rating is in the BB or B range, or A4+ on the short-term scale. These grades sit below investment grade. |
Not a red flag: upgrades, reaffirmations, new ratings at investment grade, a stable or positive outlook, a positive watch, and withdrawals. A withdrawn rating is judged by its grade. "AAA withdrawn" is not a flag. A speculative grade at withdrawal is.
How the wording is read. Agencies write the same thing in many ways, for example "Negative", "IND A/Negative" or "Rating watch with negative implications". VIGIL keeps a list of every value seen so far, and each one is assigned by hand to one of the flags above or to "not a flag". The grade decides, whatever words surround it.
Text VIGIL cannot read. Sometimes an agency puts other text in the outlook field, such as the name of a loan facility. That row is marked as a data quality case and is not counted as a red flag. Each new value of this kind goes on a review list and is assigned by hand.
Where you see it
- Ratings page: every rating action, with its flag.
- Company page: the ratings linked to that company.
- Signals: the Rating red flag option in the Flags menu picks out companies with at least one flagged rating action. The table's CSV download has a Red Flags column with the count for each company.
- Data & API: the full table, with the flag in the red_flag_reason column. See the VIGIL Data API.