–>

FIDC Credit Monitor

Asset side first (aging), then the liability side that absorbs it (tranches); CVM, monthly, 1 to 2 months lag. Carry three things: the sector rate says nothing about distribution; promised-versus-realised figures are medians (CVM's raw percentages carry outliers up to 1.6e8); subordination ratios use quota counts, not value. Screens: Suspicious Deal Screens. Comparable receivables: Securitization.

Sector Delinquency: 24 Months

Overdue receivables as a share of net assets, across FIDCs that filed both an aging table and a monthly report (CVM, monthly, 1 to 2 months lag). FIAGRO is on Industry Structure.

Most Delinquent FIDCs: Latest Period

The twenty worst rates among funds above R$1mm net assets, at the newest period the aging table covers (CVM, monthly, 1 to 2 months lag). Ranked by rate, so read net assets alongside it. Persistent cases: Suspicious Deal Screens.

Where Delinquency Worsened: and Why

The rate is overdue (R$) ÷ net assets (R$), so it also moves when net assets move. Each fund is scored on both over the last twelve complete months (CVM, monthly, 1 to 2 months lag), first observation against last. "Value moved" is |Δ R$| ≥ R$1mm, "rate moved" is |Δ| ≥ 1 p.p.; a fund needs ≥ 6 observations, and tables keep funds with ≥ R$10mm of latest net assets.

Consistent worsening: overdue value and rate both rose

Ranked by the R$ change, the more robust of the two. A rate above 100 % means overdue receivables exceed the fund's own net assets.

Real deterioration the rate hides

Overdue value rose by at least R$1mm while the rate barely moved or fell, because net assets grew (new funding dilutes a rate).

False positives: the rate rose only because net assets shrank

Overdue value did not rise and the rate still climbed: contraction or amortisation, not new delinquency.

Stopped reporting

A month not filed is absent, never zero, so a fund that quit drops out of every rate above rather than reading as clean. These filed at least six months of the window, then went two or more months quiet while the family kept filing.

Every scored fund

What this cannot tell you. CVM's monthly FIDC file carries no portfolio composition (no sector, no debtor, no guarantee), so nothing here says whose receivables went overdue or whether a subordinated tranche absorbs the loss first. vl_inadimpl is the overdue value, not a realised loss: no provisions or recovery, and a figure can sit unchanged for months then jump on a revaluation. Δ R$ is nominal. Filed only from 2025-01.

Delinquency by Aging Bucket: 12 Months

The overdue band (vl_inad_*) by days past due (CVM, monthly, 1 to 2 months lag). Weight moving toward 360d and beyond is deterioration a flat headline rate hides.

Performing vs Delinquent: Both Bands, Latest Period

Tab VI on one axis: vl_prazo_* (performing, days to maturity) against vl_inad_* (delinquent, days past due) (CVM, monthly, 1 to 2 months lag). The ten buckets are fixed by the form, so an empty bar is an empty bucket, not a missing one.

Performing Receivables by Remaining Term: 12 Months

Receivables that are not overdue, by days remaining (CVM, monthly, 1 to 2 months lag). Weight drifting into 721-1080d and >1080d is duration not yet collected.

What the Receivables Are: Sector Mix, Latest Period

Tab II files receivables by sector, a hierarchy of eleven lettered sectors with numbered members; this chart sums the lettered level only, so nothing is counted twice (CVM, monthly, 1 to 2 months lag). The share is of the summed sector lines, not of each fund's filed TOTAL, which can differ and would show as a phantom sector. Numbered detail: api.fidc_portfolio.

How the Receivables Are Graded: SCR Ladder, Latest Period

Tab X files the same receivables under BACEN SCR grades AA..H twice, by debtor and by operation: two views of one book, each summing to the graded total (CVM, monthly, 1 to 2 months lag). H is what a provisioning rule treats as near-total loss. Tab X exists from 2023-10 only.

Debtor Concentration: Books Most Exposed to One Sacado

Tab VIII publishes each fund's 25 largest debtors as anonymized ranks, a value per rank and nothing else, so this says how concentrated a book is, never in whom (CVM, monthly, 1 to 2 months lag). The ratio is the rank-1 value (and the sum of filed ranks) over the tab II receivables total of the same filing. The two tabs do not share a base for every fund: in 2026-07 rank-1 alone exceeded the total for 0.5% of funds, the top-25 sum for 1.9%. Shown as filed, not capped. Ranks Filed = 1 means a single debtor. Floor: receivables ≥ R$10mm.

Named Originators: Who Sells Receivables to the Most Funds

Tab I is the one concentration table with identities: each fund names the nine largest cedentes of each block by CPF/CNPJ, checksum-verified at ingest (CVM, monthly, 1 to 2 months lag). Block A is receivables acquired with retention of risks and benefits by the originator, block B without. This counts funds per originator and never adds shares across funds, since each share is a percent of one fund's block. Names appear only for listed companies, keyed by CNPJ. The share field is dirty as filed: 9% of slots exceed 100 (one reads 19,771); Max Share reads only 0 to 100, Outlier Slots counts the rest.

Tranche Performance: Promised vs Realised

pr_desemp_esperado against pr_desemp_real at the latest period (CVM, monthly, 1 to 2 months lag), folded into four tranche classes (classe_serie is free text). Medians, not means: these raw percentages carry outliers up to 1.6e8. Comparable counts tranches that filed both figures.

The same medians over 24 months, and the share of comparable tranches falling short.

Tranches Missing Their Target: Largest Funds

Series where realised fell short of promised at the latest period (CVM, monthly, 1 to 2 months lag), ordered by fund size, not gap, because the worst gap ranks the dirtiest numbers first. Rows Excluded counts filings outside the display band (|value| ≤ 1000%); nothing was rescaled.

Subordination Structure: Largest FIDCs

Senior versus subordinated quotas for the twelve largest FIDCs with tranche filings (CVM, monthly, 1 to 2 months lag). A quota-count ratio, not value-weighted: it divides qt_cota, so it matches a true subordination level only where senior and subordinated quotas share a unit price. Read it as capital-structure shape, not loss absorption in reais.

One fund tracked for 24 months; the ratio only means something inside a single capital structure, so it is never averaged across funds. A falling ratio with rising delinquency is the combination worth investigating.

Tranche Flows: Captação vs Resgate

Money into and out of FIDC tranches (tab X_4; CVM, monthly, 1 to 2 months lag). Both legs are positive as filed; Net Flow carries the sign; a month with no filing is blank, never zero. Zombie-growth screen: Suspicious Deal Screens.

The raw tp_oper values behind that split, matched on the CAPT / RESG substrings; CVM's vocabulary has drifted. Anything in (não classificado) means the rule stopped catching everything.