Dados B3 › Covenants
The covenant EBITDA is not the income-statement EBITDA in 563 of 829 definitions
The clause that decides whether debt accelerates is almost always written in Net Debt/EBITDA. What nobody answers with a number is whether that EBITDA is the one in the financial statements. We counted, document by document, in the indentures the regulator itself publishes.
The distribution
| Documents read | 1,517 | |
| Companies | 236 | |
| Mention EBITDA | 855 | |
| Define EBITDA in the contract | 829 | |
| — only objective terms (computable) | 266 | 32% |
| — not the income-statement EBITDA | 563 | 68% |
| · with a discretionary adjustment | 463 | 56% |
| · only adding acquired companies | 100 | 12% |
Which adjustments make the figure irreproducible
| Adjustment | Definitions | Share |
|---|---|---|
| non-recurring or extraordinary expenses | 294 | 35% |
| non-operating expenses | 179 | 22% |
| non-cash items | 117 | 14% |
| pro forma adjustments | 105 | 13% |
| items at management's discretion | 17 | 2% |
What every definition shares
| Adjustment | Definitions | Share |
|---|---|---|
| depreciation, amortisation or depletion | 814 | 98% |
| financial result or interest | 722 | 87% |
| taxes on profit | 432 | 52% |
| equity-method result | 137 | 17% |
| leases (IFRS 16) | 113 | 14% |
Depreciation and amortisation is the line we corrected in 232 EBITDAs by removing debt transaction costs wrongly filed there. quality page
What we do not do, and why
We do not say whether any company complies with its covenant. When the definition adds back “non-recurring expenses”, the figure depends on a management judgement no outsider can reproduce — that is the case in 463 of the 829 definitions. When it adds the EBITDA of an acquired company (100 definitions), the figure is objective but not in the published income statement. We flag, we do not rewrite.
The threshold we publish is a FLOOR
We read a Net Debt/EBITDA threshold in 471 documents. That is a floor, not a count: thresholds usually sit in a TABLE, and text extraction scrambles tables. In one indenture we checked by hand, carrying five thresholds in a table, automated reading found none.
Covenants get renegotiated
923 of the documents are amendments, and 149 companies have an amendment later than the indenture — including a bondholders' meeting changing the very definition of “Net Debt” used in the ratio. Those rows are flagged. Publishing a superseded covenant would be a wrong figure wearing a right face.
Source and how to check
Brazilian securities regulator, periodic-filings channel, category “Debenture indentures and amendments”. Public documents, with a direct link to the original on every row. We cover 236 of the 456 companies in our base — those that issued debentures since 2021. The regulator listed 0 documents and we could read 1,517: -1,517 failed to download or to yield text and became no row. The loss is stated because coverage that looks complete misleads.
Numbers on this page are live. Data version 2026-10-18-pl-classe-versoes · page generated 2026-10-24 05:02 UTC. If this does not match /saude, you are reading a cached copy.
Sources: CVM (open data, ODbL) and B3 (COTAHIST). Not affiliated with B3 or the CVM. Not investment advice.
Numbers on this page are live. Data version 2026-10-18-pl-classe-versoes · page generated 2026-10-24 05:02 UTC. If this does not match /saude, you are reading a cached copy.