How this report was built, what the available data supports and what it does not.
The primary source is the Statistical Bulletin for Sociedades Financieras Populares published by Mexico's banking regulator, the CNBV, in its Information Portfolio. All 64 available quarterly bulletins were downloaded, from September 2009 to June 2026, and extracted with a purpose-built parser that handles the five layout variants the document has had.
As a control comparison, the 67 equivalent bulletins for savings and credit cooperatives (Socaps) covering the same period were also processed.
From each bulletin only the current quarter column is taken, not the comparative columns. Chaining the bulletins yields the series without reconciling overlaps, and every value is the one the CNBV officially published for that cut.
Official: published by the CNBV or Banco de Mexico in a regulatory report.
Academic: study with published methodology, such as the boxes in the Financial Stability Report.
Own calculation: derived from official data with an explicit formula, documented on this page.
Reported by media: a figure circulated by the press that could not be verified against the primary data. It is labelled and kept in a separate dataset.
Scenario: a forward-looking assumption. Not a measurement.
The delinquency ratio (IMOR) is stage 3 loans over total loans: it measures what remains on the balance sheet and is not being paid. The adjusted ratio (IMORA) also adds loans written off over the previous twelve months.
The CNBV publishes IMOR by society in its quarterly bulletin. It does not publish write-offs by society, so IMORA cannot be computed entity by entity from public information. Every comparison between societies in this report is IMOR.
Sector write-offs were estimated by solving for the unknown in IMORA = (stage 3 + write-offs) / (total portfolio + write-offs), using the consumer IMORA circulated in the press. It is an approximation: the regulatory definition of IMORA uses twelve-month averages rather than point-in-time balances, and one input is not official. The result is therefore labelled as a scenario and expressed as an order of magnitude.
It is not assigned by brand or reputation. It is derived from the customers-per-branch figures the CNBV publishes: no branches, or a hundred thousand customers or more per branch, is digital; fewer than twenty thousand is traditional; in between is mixed. Without sufficient data, unclassified.
Each group's IMOR is a portfolio-weighted average, not an arithmetic one: averaging percentages unweighted would give a society with forty thousand pesos of loans the same weight as one with thirty-three billion.
Computed as implicit lending rate minus implicit funding cost minus delinquency ratio. It is deliberately conservative and is not an accounting measure of profitability: it ignores operating expenses and treats the stage 3 balance as if it were entirely lost within the period. It serves to rank societies by headroom, not to estimate earnings.
The CNBV does not publish Sofipos bulletins for 2019. That year appears as a gap in the series and is not interpolated.
The portfolio breakdown by credit type, which allows the consumer IMOR to be computed, is available from December 2012 onward and not before.
When a society does not file in a given quarter, the CNBV publishes zeros. Those quarters are treated as missing rather than as zero, because a licensed, operating society cannot have zero customers and zero assets. For branches and loans the zero is kept, because there it is a real value.
Some societies report zero interest expense despite holding significant deposits. The figure is kept exactly as the CNBV publishes it and flagged as a source anomaly, not corrected.
The CASFIM key identifies the legal entity and is stable even when the commercial name changes. Several fintechs entered by acquiring an existing Sofipo licence, so the legal series is continuous but the brand is not. Each society shows the quarter from which the CNBV publishes its current name.
Every series in the dashboard can be downloaded as CSV. The download and extraction scripts are part of the report's repository.