When a credit co-operative fails without formally failing

Brazil’s financial inclusion success story faces a hidden test

A credit co-operative can disappear as a legal entity while accounts, payment services and loan contracts are able to continue under a successor. Under Brazilian law, however, members may not only lose their quota capital: if statutory reserves are insufficient, a valid loss allocation can require them to cover verified losses from personal funds. The apparent contradiction comes from the co-operative model: the customer may also be an owner, and a distressed institution can be absorbed without formal liquidation.

Preserving deposits and continuity may amount to a successful resolution. However, if the transaction never enters a public register of assisted resolutions, its economic substance remains difficult to examine, compare or learn from.

The scale makes this more than a local disclosure issue. The World Council of Credit Unions reports 412.7m members and $3.83tn in assets across 101 countries at the end of 2024. Brazil offers an unusually large emerging-market test of whether resolution disclosure can keep pace as co-operative finance reaches bank-like scale.

A strong sector with a growing transparency burden

Brazil’s National Co-operative Credit System ended 2025 with R$1.04tn in assets, 21.2m members and 742 individual (‘singular’) credit co-operatives. Assets grew 17% in one year and the network reached 59% of municipalities. Banco Central do Brasil describes the sector as strengthening competition, efficiency and financial inclusion.

The aggregate numbers do not describe a sector on the brink. BCB reports that provisions remained above expected credit losses, the aggregate Basel ratio for singular co-operatives reached 21.42% and those entities recorded R$21.1bn of earnings before remunerating member capital in 2025. Credit risk rose, but system-wide capital remained comfortable. The sector is also regulated: credit co-operatives are financial institutions authorised and supervised by BCB, subject to prudential rules, audit and special resolution regimes.

The concern lies elsewhere: public reporting does not clearly show who absorbs losses or how distress is resolved. The number of individual co-operatives fell to 742 from 753 in 2025, mainly through mergers, which may reflect efficiency, ordinary succession or early intervention, but public data do not readily distinguish among these mergers.

A merger can therefore produce two outcomes at once: continuity for the depositor and loss absorption by the member-owner.

One person, two balance-sheet positions

An ordinary bank customer principally has a creditor claim. A member of a Brazilian credit co-operative may hold several legally distinct positions at once. Eligible deposits are protected by the private sector-funded Fundo Garantidor do Cooperativismo de Crédito (FGCoop) – up to R$250,000 per client at each member institution.

For credit co-operatives, the general assembly must set the formula based on each member’s operations during the financial year. BCB explains that affected members can be required to meet the allocation from personal funds, not merely accept a write-down of their quota. Since a 2022 reform, an assembly approving an incorporation may also assign member-specific loss claims to a guarantee fund, which then becomes the creditor after the original co-operative has been absorbed.

That is not a blanket liability for every customer. A non-member client does not assume a member’s obligations, and an individual member is not liable for the institution’s entire balance sheet. Losses must be verified, reserves insufficient, the allocation valid and the charge proportional to the member’s operations. Still, the distinction matters: deposit insurance can protect one pocket while co-operative law exposes another.

The missing resolution ledger

An FGCoop presentation using data through April 2025 counted 355 mergers between 2014 and 2024, of which 207 involved losses and 18 involved negative equity. The figure does not represent 207 failures; it does show why formal liquidations are an inadequate measure of co-operative distress.

FGCoop’s annual report shows that the mechanism has been used repeatedly. In 2024, it arranged three new financial-assistance operations, all involving the allocation of losses. They involved 88,448 members, around R$594.5m in deposits and R$130.8m in member capital. Since the assistance process began, the fund reports 16 operations, R$1.77bn in deposits for which standard payouts of covered deposits were avoided, R$634.32m in member capital preserved and R$355.11m contributed by the fund.

Assistance can be preventive, and maintaining access to deposits is exactly what a well-designed safety net should achieve. The reporting problem is that the fund’s assistance portfolio remains visible only in aggregate. Outsiders cannot distinguish ordinary consolidation from assisted resolution, identify how member capital was treated, compare fund support with recoveries or judge whether intervention came early enough.

The US National Credit Union Administration’s searchable public register distinguishes conservatorships, involuntary liquidations and ‘merger with NCUA assistance’, recording whether an institution was closed, merged or released. US-insured ‘share accounts’ are not equivalent to Brazilian quota capital; the comparison is institutional, not mechanical. The lesson is that a merger need not disappear from a public resolution history.

Apply the new global standard to co-operatives

The International Association of Deposit Insurers has now supplied an international benchmark with its revised 2025 core principles. A dedicated section on co-operative systems recognises that umbrella organisations can intervene early, but require the deposit insurer and other safety-net authorities to have the power and capacity to address weaknesses in both central organisations and individual co-operatives. The principles also call for operational independence, clear communication of what is and is not insured, crisis testing, timely information-sharing and resolution procedures in which ownership instruments absorb losses before insured deposits.

Brazil was one of three jurisdictions used to pilot the revised standard, but IADI’s published account says the Brazilian exercise was conducted with Fundo Garantidor de Créditos, the guarantee fund for conventional bank deposits, rather than with FGCoop. That should be the starting point for a public, co-operative-specific assessment, not evidence that one has already been completed.

Brazil can close much of this gap with three measures. First, BCB and FGCoop should develop – and, where necessary, seek legal authority for – a registry of assisted resolutions by legal entity, date, successor, form of intervention, deposits preserved, member-capital impairment or loss allocation, fund assistance and eventual recoveries. A confidentiality lag and carefully delimited fields could protect statutory secrecy and a live transaction without erasing its history.

Second, every member should receive a one-page ‘coverage map’ on joining and annually thereafter. It should separate insured deposits, uncovered balances, quota capital and possible loss-sharing obligations, while identifying the individual legal entity behind the national brand.

Third, FGCoop should publish an annual aggregate readiness assessment: its target-fund methodology, insured exposure, concentration, scenarios involving simultaneous failures, available liquidity, extraordinary funding arrangements and tested payout capacity. Such disclosure could remain entirely aggregate. Its purpose would be to prevent the fund’s current size from being read as proof of either sufficiency or inadequacy.

Brazil’s co-operative expansion is a financial inclusion success. That success is precisely why the safety net deserves a clearer public account. Resolution outcomes should be reported in two columns: what happened to the depositor and what happened to the member-owner. When a co-operative disappears without formally failing, preserving the first column should not make the second invisible.

Gustavo Pessoa is Professor of Economics at Fundação Getulio Vargas.

Join OMFIF on 23 September to examine Brazil’s elections in regional perspective: implications for the global economy.

Interested in this topic? Subscribe to OMFIF’s newsletter for more.

 

Join Today

Connect with our membership team

Scroll to Top