PT

Dados B3 › Reapresentações

Restated financial statements on B3: 8,654 on record

When a company restates a published statement, the regulator overwrites the old version — the file shows only the new figure, with nothing saying it changed. We have kept both versions since 2010: 8,654 restatements across 402 of the 456 companies we track.

8,158
value revisions — the figure changed size
489
the account appeared or vanished between versions

These add up to 8,654. The remaining 7 are sign flips — a presentation convention, not a revision. We separate them because treating all three alike inflates the headline number.

When they restate, the figure GOES DOWN

One in three fiscal years in the base has some line restated — 1,745 of 5,666. And the change has a direction: revenue shrinks in 74.5% of cases, equity in 67.1%. We can only measure this because we keep both versions; the regulator overwrites the old one.

accountpairs shrank inmedian change
Receita DRE 3.0136474.5%5.1%
Custo (em magnitude) DRE 3.0243765.4%5.0%
Resultado bruto DRE 3.0350465.9%6.8%
EBIT DRE 3.0546055.7%7.3%
Lucro líquido DRE 3.1124158.9%9.4%
Ativo total BPA 139663.9%2.1%
Patrimônio líquido BPP 2.0324667.1%4.3%
Caixa operacional DFC_MI 6.0166750.5%10.3%

Clean pairs only: 1,244 rows were excluded because one side is zero or the filing carries a placeholder value — there the account appeared or vanished, it did not change size.

And the honest reading is not window dressing. Two things point to operations leaving the comparative — perimeter, not a revision of results: cost shrinks alongside revenue in similar proportion, and operating cash flow stays nearly symmetric (50.5%), when it would be the first to move if this were an artefact of our method. We describe what the two filings say; we do not assert which one is right.

The largest revisions, in reais

Ranked by ABSOLUTE difference, not percentage, and only on accounts that mean something, with both sides above R$ 100 million. Ranking by percentage crowns accounts that emptied out — true and misleading.

companyfiscal yearaccount first filing (R$ bn)restated (R$ bn)difference
ITSA4
ITAUSA S.A.
2012Ativo total
BPA 1
364.0239.05324.97
BBAS3
BANCO DO BRASIL S.A.
2012Ativo total
BPA 1
1,136.011,014.08121.93
ITUB4
ITAÚ UNIBANCO HOLDING S.A.
2018Receita
DRE 3.01
131.32174.8143.49
PETR4
PETRÓLEO BRASILEIRO S.A. - PETROBR
2018Receita
DRE 3.01
349.84310.2539.58
PCAR3
COMPANHIA BRASILEIRA DE DISTRIBUIÇ
2021Receita
DRE 3.01
51.2916.3034.99
PCAR3
COMPANHIA BRASILEIRA DE DISTRIBUIÇ
2015Receita
DRE 3.01
69.2237.2032.02
AXIA3
AXIA ENERGIA S.A.
2012Ativo total
BPA 1
172.20142.4329.76
AMER3
AMERICANAS S.A. - EM RECUPERAÇÃO J
2021Patrimônio líquido
BPP 2.03
15.91-12.6228.53
PCAR3
COMPANHIA BRASILEIRA DE DISTRIBUIÇ
2019Receita
DRE 3.01
56.6328.8427.80
ITUB4
ITAÚ UNIBANCO HOLDING S.A.
2022Receita
DRE 3.01
283.37259.0324.35
—
LITEL PARTICIPACOES SA
2010Ativo total
BPA 1
44.6922.2722.42
OIBR3
OI S.A. - EM RECUPERAÇÃO JUDICIAL
2016Patrimônio líquido
BPP 2.03
12.46-6.6719.12
ITSA4
ITAUSA S.A.
2012EBIT
DRE 3.05
-12.835.2318.06
OIBR3
OI S.A. - EM RECUPERAÇÃO JUDICIAL
2016Ativo total
BPA 1
82.1765.9716.20
PETR4
PETRÓLEO BRASILEIRO S.A. - PETROBR
2012Patrimônio líquido
BPP 2.03
345.43330.7814.66

Who restated most

companyrows
CEBR3 COMPANHIA ENERGÉTICA DE BRASÍLIA - CEB102
PCAR3 COMPANHIA BRASILEIRA DE DISTRIBUIÇÃO99
AXIA3 AXIA ENERGIA S.A.96
CEPE5 CIA ENERGÉTICA DE PERNAMBUCO - CELPE90
— YOU INC INCORPORADORA E PARTICIPAÇÃO S90
PDTC3 PADTEC HOLDING S.A.85
LUPA3 LUPATECH S.A.82
TPIS3 TPI - TRIUNFO PARTICIPACOES E INVESTIM77
EEEL3 CPFL TRANSMISSÃO S.A.74
OIBR3 OI S.A. - EM RECUPERAÇÃO JUDICIAL74
GFSA3 GAFISA SA70
PEAB3 CIA DE PARTICIPAÇÕES ALIANÇA DA BAHIA70

By fiscal year

yearrows
20254
2024482
2023410
2022574
2021539
2020438
2019561
2018557
2017500
2016490
2015443
2014386

Where revisions happen

accountrestatements
EBIT 3.05515
Ativo total 1425
Receita 3.01403
Lucro líquido 3.11278
Patrimônio líquido 2.03278

Why this matters

For backtesting: using today's figure for a 2015 fiscal year tests a strategy on information nobody had in 2015. It is the quietest form of look-ahead — you remember to lag the price, not the balance sheet.

For analysis: a company that restates often deserves a more skeptical read. The per-company count above is a signal, not a verdict.

What this is NOT: an accusation. Restating is foreseen and often mandatory — a change in standards, a reclassification, a merger. We record the FACT that it changed, not the intent.

Check it yourself

Download the year's file from dados.cvm.gov.br and compare the ORDEM_EXERC column: ÚLTIMO is the year as first reported, PENÚLTIMO is the same year reported one year later. A gap above 0.5% is what we record.

Methodology · Open audit · https://dadosb3.com/reapresentacoes (JSON, free key)

Numbers on this page are live. Data version 2026-09-27-reapresentacao-agreg · page generated 2026-09-27 05:43 UTC. If this does not match /saude, you are reading a cached copy.

Home · Methodology · Transparency · What a restatement is

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 2026-09-27-reapresentacao-agreg · page generated 2026-09-27 05:43 UTC. If this does not match /saude, you are reading a cached copy.