Is it true that one in four digital loans in Mexico is in default?

The figure is the adjusted delinquency ratio (IMORA) for Sofipos' consumer credit, which adds loans written off over the previous twelve months. The delinquency ratio the CNBV publishes for that segment in June 2026 is 11.24%. Both figures are correct and they measure different things.

A Sociedad Financiera Popular is an entity regulated by the CNBV that takes public deposits and lends, under Mexico's Popular Savings and Credit Law. Its deposits are covered by a Protection Fund up to 25,000 UDIS per person.

Do fintechs have higher delinquency than traditional Sofipos?

On CNBV data to June 2026, no. On the total loan book, the weighted IMOR of digital Sofipos is 6.31% against 24.82% for traditional Sofipos. On consumer credit, the segment the article was about, 7.96% against 24.99%. The caveat is that the CNBV does not publish write-offs by society, so the adjusted indicator cannot be compared entity by entity.

Does this pose a systemic risk to Mexico's financial system?

On the available data, it does not behave like one: the sector kept growing through six licence revocations since 2022. The relevant risk is concentration in specific institutions and loss for savers above the Protection Fund's limit, which is 25,000 UDIS.

Where does this report's data come from?

From the 64 quarterly Sofipos statistical bulletins the CNBV published between 2009 and 2026, plus the 67 equivalent Socaps bulletins and Banco de Mexico's financial stability reports. Every series can be downloaded as CSV from the dashboard.

AI generated illustration. No data lives inside the images: every figure is in the text, with its source.

Scroll to read ↓

Pro-Latam data report

The price of risk

Digital credit, Sofipos and fintechs in Mexico, told with the numbers the CNBV publishes

On 24 August 2026 Mexican media reported that one in four digital loans in Mexico is in default, and the figure was repeated for days. It is real, but it does not measure what almost everyone understood. Here is the CNBV's official data, society by society, since 2009.

Source

Roberto A. Arrucha · Director, Pro-Latam

LinkedIn Instagram

Data as of 2026-06

Go to the summary and findings

The scale of it

39,272,917

customers in June 2026

689

branches, 471 fewer than in 2020

More customers, fewer branches = digital banking.

11.24%

delinquency on consumer credit, on the figure the CNBV publishes.

The figure the media reported, 25.41%, measures something else: the Imora, which adds loans already written off.

6.31%

24.82%

digital Sofipos against traditional ones

Each dot is a society authorised and operating as of the end of June 2026.

The risk in Mexican digital credit is not where it was said to be.

On their total loan book, digital Sofipos close June 2026 at 6.31% delinquency. Traditional Sofipos, at 24.82%. Same sector, same regulator, same customers, a near fourfold gap.

Source: CNBV, Sofipos Statistical Bulletin, June 2026.

Summary and key findings

The report's seven findings and the state of the sector at the CNBV's latest published cut, June 2026.

Findings

Each finding states whether it comes from official data or from an own calculation on official data.

The headline indicator is not what the CNBV measures by society
Sofipos' consumer credit IMOR is 11.24% in June 2026. The 25.41% figure the media reported from 24 August 2026 is the write-off adjusted index, which the CNBV does not break down by society. In that same segment, traditional Sofipos stand at 24.99% and digital ones at 7.96%.
Own calculation
Deterioration sits with traditional Sofipos, not digital ones
On the total loan book, weighted IMOR of 6.31% at digital Sofipos against 24.82% at traditional Sofipos. On consumer credit, 7.96% against 24.99%. Same customer base, same regulatory framework.
Own calculation
Nine times more customers with no break in delinquency
From 4,370,261 customers in December 2020 to 39,272,917 in June 2026, with sector IMOR moving from 9.3% to 9.9%.
Official
Banco de México reaches the same conclusion
Its June 2026 Financial Stability Report separates digital from non-digital Sofipos and notes that delinquency fell at the former while it rose at the latter. A different cut-off, a different methodology, the same reading.
Official
The spread explains more than the delinquency level
The sector charges 45.47% and pays 4.63%. Societies whose rate does not cover their own delinquency are the ones consuming themselves, regardless of size.
Own calculation
Fintechs entered by buying licences, not obtaining them
The CNBV itself documents the name change on the same society key: Klar over SEFIA, Nu Mexico over Akala, Stori over Mascaja and then Savvi, Fondeadora over Apoyo Multiple.
Official
The sector expels participants regularly
Six licence revocations and one merger recorded since 2022, the latest in September 2025 on a society with more than a million customers.
Official

The sector at the latest cut

The six indicators the CNBV publishes for Sofipos as a whole, each against its value a year earlier.

Customers
39,272,917
30,233,479 a year earlier
As of: 2026-06
Source:Comisión Nacional Bancaria y de Valores (CNBV)Official
Delinquency ratio (IMOR)
9.9%
10.27% a year earlier
As of: 2026-06
Source:Comisión Nacional Bancaria y de Valores (CNBV)Official
Consumer credit IMOR
11.24%
11.02% a year earlier
As of: 2026-06
Source:Comisión Nacional Bancaria y de Valores (CNBV)Own calculation
Implicit lending rate
45.47%
46.41% a year earlier
As of: 2026-06
Source:Comisión Nacional Bancaria y de Valores (CNBV)Official
Implicit funding cost
4.63%
11.86% a year earlier
As of: 2026-06
Source:Comisión Nacional Bancaria y de Valores (CNBV)Official
Return on equity (ROE)
2.74%
-32.74% a year earlier
As of: 2026-06
Source:Comisión Nacional Bancaria y de Valores (CNBV)Official

Three readings the bulletin does not carry

Comparisons computed from the official data. The formula for each is in the methodology.

24.82%
delinquency at traditional Sofipos, total loan book
Against 6.31% at digital ones.
Context: Classification derived from the customers-per-branch data the CNBV publishes.
Source:Comisión Nacional Bancaria y de Valores (CNBV)Own calculation
85.9%
of customers sit in four societies
Nu México, Stori, Klar and Sofiexpress.
Context: None of the four grew with a branch network.
Source:Comisión Nacional Bancaria y de Valores (CNBV)Own calculation
4.35%
delinquency at Socaps
Same regulated sector, same customers, without mass digital origination.
Context: It is the cleanest control comparison available.
Source:Comisión Nacional Bancaria y de Valores (CNBV)Official

Are Sofipos in crisis? The difference between delinquency and write-offs

The figure the media have been reporting with alarm since 24 August 2026 is the IMORA for consumer credit: 25.41%. The delinquency ratio the CNBV publishes for that same segment and that same month is 11.24%. They do not contradict each other: they measure different things.

IMORA adds loans written off over the last twelve months; IMOR only measures what remains on the balance sheet. They are not interchangeable, and the CNBV does not publish IMORA by society.

A lender that writes off fast shows a low IMOR and a high IMORA. One that carries bad loans shows the opposite. The distance between them is not a technicality: it is the sector's entire write-off policy, and it amounts to roughly 13.6 billion pesos over the previous twelve months, about 1.7 times the portfolio still recognised as non-performing.

That is why this report never mixes the two series. The CNBV's official figures stand on one side; the ones the media reported are labelled, sourced, and always shown next to their official counterpart.

One hundred consumer loans, two ways of counting them

11/ 100

are not being paid

This is IMOR: non-performing loans still recognised on the society's balance sheet. It is the indicator the CNBV publishes by society.

Performing
Non-performing, on the books
Already written off, off the books

Proportions rounded to whole numbers out of one hundred loans, as a teaching device. The exact figures are 11.24% and 25.41%.

	A year earlier	Figure reported by media	Figure published by the CNBV
Sofipos, consumer creditConsumer credit IMOR	22.4%	25.41%	11.24%
StoriDelinquency ratio (IMOR)	23.21%	35.74%	12.63%
KlarDelinquency ratio (IMOR)	26.97%	31.53%	8.2%
Ku-BoDelinquency ratio (IMOR)	34.28%	52.73%	12.24%
Nu MéxicoDelinquency ratio (IMOR)	·	19.89%	6.59%
Commercial banks, total portfolio	·	4.71%	No official counterpart by society
Commercial banks, consumer credit	·	11.31%	No official counterpart by society

IMORA adds loans written off over the last twelve months; IMOR only measures what remains on the balance sheet. They are not interchangeable, and the CNBV does not publish IMORA by society.

Source:Agencia ReformaReported by media

The article the figure came from

Reportan impago 1 de cada 4 en Sofipos
Charlene Domínguez / Agencia Reforma, 24 August 2026

The figure that spread in August 2026 comes from an article by Charlene Domínguez, distributed by Agencia Reforma on 24 August. This report consulted it in the syndicated Diario de Juárez version.

See the article's full context

Precision is owed here, because the article is precise: its body names the indicator, Índice de Morosidad Ajustada (Imora), attributes it to the CNBV and gives the full series, from 22.40 percent in June 2025 to 25.41 percent in June 2026. It also compares with commercial banking, 4.71 percent on the total book and 11.31 percent on consumer credit, and quotes Vicente Gómez, an analyst at Moody's México, explaining that Sofipos are the entry point to the financial system for historically excluded segments and that this carries more risk.

The ambiguity is not in the reporting, it is in the headline and in what travels from it. Saying that one in four loans is in default suggests non-performing loans on the balance sheet, and Imora does not measure that: it includes loans already written off. When the figure is repeated without the indicator's name, which is how it usually gets repeated, it stops meaning what the article said.

This report does not correct the article. It corrects what was understood from it, and it does so by placing the figure the CNBV does publish by society right next to it.

There is an element of trial and error, of trying to find and separate who can actually be banked.
Vicente Gómez, analyst at Moody's México, quoted in the article.

Read the original articleSource:Agencia ReformaReported by media

Who is who: Sofipos, Socaps and commercial banks

Before comparing groups it helps to know what they are. Sofipos and Socaps are not banks, and the difference is not one of size: it is the law, the supervisor and, above all, how much of a depositor's savings is protected.

Sociedades Financieras Populares and savings and credit cooperatives share a sector, a supervisor and a customer base. They differ in nature: a Sofipo is a for-profit company and has customers; a Socap is a cooperative and has members, who own it.

The real distance from commercial banking is deposit insurance. A saver at a bank is covered up to 400,000 UDIS by the IPAB. At a Sofipo or a Socap, up to 25,000 UDIS, sixteen times less. Anyone depositing above that threshold carries the institution's risk.

And how big is all this? Banco de México sizes it in its June 2026 Financial Stability Report: Sofipos add up to 0.6% of GDP and Socaps to 0.9%. Small in assets, yet the group of non-bank intermediaries they belong to extends 28.5% of the country's household credit. Little balance sheet, many people.

A note on scope: this report analyses Sofipos. Socaps appear as a control comparison, because they isolate the effect of the digital model while holding everything else constant, and only sector aggregates are shown for them. Commercial banking appears only as a reference for scale.

And a warning about this report's vocabulary: digital, mixed and traditional are not legal figures. They are an own classification, derived from the customers-per-branch data the CNBV publishes, and it applies only to the 35 Sofipos. There is no such thing as a digital Socap or a traditional bank in this sense.

Popular savings and credit sector

Both take public deposits, are supervised by the CNBV and share the same depositor protection limit. This is the sector this report analyses.

Sofipos

Sociedades Financieras Populares. For-profit companies.

39,272,917

customers

9.9%

IMOR

Governing law: Popular Savings and Credit Law
Supervisor: CNBV and Condusef
Deposit protection: Protection Fund, 25,000 UDIS
Entities in operation: 35
Total assets: 218.5 billion pesos
Share of GDP (Banxico): 0.6%

Within Sofipos: distribution model

Digital · 10 societies6.31%
Mixed · 1 society5.72%
Traditional · 19 societies24.82%

Own classification derived from the customers-per-branch data the CNBV publishes. It is not a legal figure and does not apply to Socaps or banks.

Source:Comisión Nacional Bancaria y de Valores (CNBV)Official

Socaps

Savings and credit cooperatives. Members are the owners.

10,072,072

members

4.35%

IMOR

Governing law: Law Regulating Savings and Credit Cooperatives
Supervisor: CNBV
Deposit protection: Deposit insurance, 25,000 UDIS
Entities in operation: 152
Total assets: 336.3 billion pesos
Share of GDP (Banxico): 0.9%

This report does not analyse Socaps. It uses them as a control comparison, because they share sector, supervisor, customer base and protection limit with Sofipos but do not run mass digital origination models. Only sector aggregates are shown for them, never society-level figures.

Source:Comisión Nacional Bancaria y de Valores (CNBV)Official

Commercial banking

Commercial banks. A different law and a different deposit insurance.

11.31%

reported consumer IMORA

Governing law: Credit Institutions Law
Supervisor: CNBV and Banco de México
Deposit protection: IPAB, 400,000 UDIS

The CNBV published its commercial banking statistical bulletin through December 2021 and then moved that information to interactive dashboards; the historical series file its own site links to returns an error. That is why this report builds no official commercial banking series and its delinquency figure is labelled as reported. What can be said with a source: Banco de México puts Sofipos at 0.6% of GDP and Socaps at 0.9%, and describes non-bank intermediaries as small relative to the financial system, even though they extend 28.5% of household credit.

Source:Agencia ReformaReported by media

The curve that did not break

The sector multiplied its customer base ninefold and its delinquency moved 0.6 percentage points: from 9.3% in December 2020 to 9.9% in June 2026.

It is one of Latin America's fastest credit expansions, and it happened almost entirely inside a phone. Four societies now hold 85.9% of the sector's customers, and none of them grew by opening offices.

The question that matters to an investor is not whether digital credit grew, that much is known. It is whether the price charged covers what goes unpaid. This report answers that with the regulatory filings the societies themselves submit to Mexico's banking regulator, the CNBV.

This is the opposite of what is usually assumed when a lender grows fast among customers with no credit history. The total loan book went from 22.5 billion pesos to 100.4 billion pesos, and sector assets reached 218.5 billion pesos.

The stability of the average, however, hides a shift in composition. The risk did not disappear: it changed hands.

Sector historical series

Metric:

Source:Comisión Nacional Bancaria y de Valores (CNBV)OfficialDownload CSV

View data as a table

Period	Value (count)
2009-09	1,395,084
2010-03	1,756,492
2010-06	1,787,013
2010-09	2,031,112
2010-12	2,116,711
2011-03	2,169,633
2011-06	2,245,529
2011-09	2,356,128
2011-12	2,348,043
2012-03	2,404,423
2012-06	2,445,211
2012-09	2,517,498
2012-12	2,621,591
2013-03	2,645,849
2013-06	2,700,282
2013-09	2,758,867
2013-12	2,743,054
2014-03	2,890,535
2014-06	2,932,803
2014-09	2,996,926
2014-12	2,900,641
2015-03	3,056,752
2015-06	3,175,594
2015-09	3,307,173
2015-12	3,395,249
2016-03	3,423,465
2016-06	3,475,437
2016-09	3,493,998
2016-12	3,558,971
2017-03	3,695,515
2017-09	3,842,849
2017-12	3,983,072
2018-03	3,706,143
2018-06	3,681,372
2018-09	3,861,866
2018-12	3,899,662
2020-03	4,946,365
2020-06	4,829,057
2020-09	4,655,616
2020-12	4,370,261
2021-03	4,780,345
2021-06	4,651,098
2021-09	5,056,616
2021-12	5,192,351
2022-03	4,847,982
2022-06	4,914,685
2022-09	5,345,857
2022-12	8,299,264
2023-03	9,247,961
2023-06	10,570,989
2023-09	11,377,175
2023-12	11,576,903
2024-03	15,462,335
2024-06	17,598,804
2024-09	22,145,688
2024-12	24,957,226
2025-03	25,269,283
2025-06	30,233,479
2025-09	32,091,886
2025-12	35,895,132
2026-03	32,256,996
2026-06	39,272,917

What is charged and what is paid

In June 2026 Sofipos charged an implicit rate of 45.47% on their loan book and paid 4.63% on deposits. The spread is 40.8 percentage points.

A spread like that only makes sense if delinquency is high: the price of credit is absorbing expected losses, not capturing rent. The right question is not whether 45.47% is an expensive rate, but whether it covers what that book fails to repay.

Funding costs collapsed in a year, from 11.86% to 4.63%, while lending rates barely moved. That shift explains much of the sector's return to profit: ROE went from -32.74% to 2.74%.

Lending rate against funding cost

Implicit lending rateImplicit funding costSpread

Source:Comisión Nacional Bancaria y de Valores (CNBV)OfficialSource:Comisión Nacional Bancaria y de Valores (CNBV)Official

Download CSV

View data as a table

Period	Implicit lending rate (%)	Implicit funding cost (%)	Spread (%)
2023-12	43.37	9.92	33.45
2024-03	41.66	12.3	29.36
2024-06	43.16	13.16	30
2024-09	46.16	13.27	32.89
2024-12	44.02	12.33	31.69
2025-03	45.12	12.25	32.87
2025-06	46.41	11.86	34.56
2025-09	46.88	11.17	35.7
2025-12	46.42	6.1	40.31
2026-03	45.96	9.42	36.54
2026-06	45.47	4.63	40.84

Who holds up and who does not

We already saw the gap in consumer credit. The same holds for the total loan book: classifying each society by its physical network, digital Sofipos show a weighted IMOR of 6.31% and traditional Sofipos 24.82%. The gap is nearly fourfold, and it runs opposite to the dominant narrative.

These comparisons are IMOR, not IMORA. The CNBV does not publish write-offs by society, so the adjusted indicator cannot be computed entity by entity. Part of the digital lenders' better showing is better calibrated origination and part of it is writing off sooner; public information cannot separate how much is which.

The extreme case is not a fintech. Nu Mexico, the sector's largest, holds 49.8% of all consumer credit with an IMOR of 6.59%. The sector's two extremes read better side by side:

Libertad

Consuming itself

Traditional Sofipo. Branch network, two million customers.

42.23%
implicit rate charged: 46.8%
goes unpaid (IMOR): 2,430,756
customers: -11.2
points of headroom

Source:Comisión Nacional Bancaria y de Valores (CNBV)Official

Stori

Sustained

Digital Sofipo. No physical network, one of the sector's largest consumer books.

75.3%
implicit rate charged: 12.63%
goes unpaid (IMOR): 7,553,345
customers: 52.4
points of headroom

Source:Comisión Nacional Bancaria y de Valores (CNBV)Official

Digital Sofipos against traditional ones

June 2026 delinquency ratio, weighted by each society's portfolio within its group. The classification is not assigned by brand: it is derived from the customers-per-branch figures the CNBV publishes, and the rule is in the methodology.

Source:Comisión Nacional Bancaria y de Valores (CNBV)Own calculationDownload CSV

View data as a table

Traditional (19 societies)	24.82
Digital (10 societies)	6.31
Mixed (1 society)	5.72

The gap above is on the total loan book. But the headline was about consumer credit, so it is worth looking at exactly that segment.

One in four, but whose?

Consumer credit IMOR, by distribution model

Here is who the headline actually describes. Digital Sofipos lend four out of every five pesos of the sector's consumer credit, and their delinquency is three times lower.

Group	Consumer IMOR	of the consumer portfolio
Digital10 societies · 79% of the consumer portfolio	7.96%
Traditional19 societies · 19% of the consumer portfolio	24.99%
Mixed1 society · 1% of the consumer portfolio	5.72%

Source:Comisión Nacional Bancaria y de Valores (CNBV)Own calculation

Mind the coincidence: the 24.99% for traditional Sofipos and the 25.41% that was reported look alike, but they are not the same measure. One is a group's IMOR, the other the whole sector's IMORA. That they nearly match is revealing, not equivalent. And it still cannot be claimed that digital Sofipos have a lower IMORA: the CNBV does not publish write-offs by society.

The result settles the debate. In the very segment the headline is about, consumer credit, traditional Sofipos show an IMOR of 24.99% and digital ones 7.96%. The figure that travelled as a portrait of digital credit describes, on the balance sheet, almost exactly the opposite.

DigitalMixedTraditionalSource:Comisión Nacional Bancaria y de Valores (CNBV)Official

Each bubble is a society; size is its portfolio. Below the diagonal the rate covers recognised delinquency; above it, it does not.

The cushion

The sector closed June 2026 with booked reserves at 111.77% of those required and a coverage ratio of 103.66%.

Meeting the regulatory minimum is not the same as having room to spare. Several societies book exactly one hundred percent of what is required, with no surplus, while others reserve well above it.

Behind these figures sit 176.6 billion pesos of public deposits. The Protection Fund covers up to 25,000 UDIS per person, so small savers are protected and anyone above that threshold is not.

Reserve coverage

Source:Comisión Nacional Bancaria y de Valores (CNBV)OfficialDownload CSV

View data as a table

Caja Bienestar	103.98
Unagra	103.2
Libertad	100
Fincomún	100
Tamazula	100
Monte de Piedad	100
Klar	100
Credicapital	100
Capital Activo	100
Fondeadora	100
Crediclub	100
Premo	100

The systemic question

Socaps, which serve the same population segment and sit in the same popular savings and credit sector but without mass digital origination, closed June 2026 with an IMOR of 4.35%, against 9.9% for Sofipos.

It is the cleanest control comparison available, because it isolates the distribution model while holding customer type and regulatory framework constant.

Nor is the sector hypothetical about exits: since 2022 the CNBV has revoked the licences of 6 societies and recorded one merger. The latest revocation, in September 2025, hit a society that a year earlier reported more than a million customers.

This report cannot size the comparison with commercial banking on official data, for the reason already given: the CNBV stopped publishing that series through the available route. What the data does support is where the risk sits. With 218.5 billion pesos in assets spread across 35 societies, and six licence revocations since 2022 through which the sector never stopped growing, the relevant problem is not contagion to the system: it is concentration in specific institutions and loss for savers above the Protection Fund's limit.

Sofipos against Socaps

SofiposSocaps

Source:Comisión Nacional Bancaria y de Valores (CNBV)OfficialSource:Comisión Nacional Bancaria y de Valores (CNBV)Official

Download CSV

View data as a table

Period	Sofipos (%)	Socaps (%)
2009-09	11.52	9.62
2009-12	11.75	No data for this cut.
2010-03	11.45	7.31
2010-06	11.18	7.23
2010-09	10.79	7.31
2010-12	10.26	7.14
2011-03	10.66	7.29
2011-06	11.19	7.79
2011-09	10.42	8.22
2011-12	9.6	6.21
2012-03	9.29	7.18
2012-06	10.66	7.46
2012-09	10.82	6.86
2012-12	8.66	6
2013-03	10.04	6.14
2013-06	8.83	6.43
2013-09	10.62	6.44
2013-12	7.38	6.11
2014-03	8.39	6.24
2014-06	8.01	6.07
2014-09	8	6.02
2014-12	9.81	5.77
2015-03	9.63	5.89
2015-06	9.71	5.86
2015-09	8.68	5.68
2015-12	8.5	5.21
2016-03	8.89	5.26
2016-06	9.14	5.15
2016-09	9.7	4.92
2016-12	9.84	4.39
2017-03	10.8	4.43
2017-06	No data for this cut.	4.43
2017-09	10.24	4.44
2017-12	10.32	4.3
2018-03	11	4.36
2018-06	10.32	4.43
2018-09	9.67	4.42
2018-12	9.2	4.3
2019-03	No data for this cut.	4.39
2019-06	No data for this cut.	4.44
2019-09	No data for this cut.	4.51
2019-12	No data for this cut.	4.26
2020-03	13.36	4.52
2020-06	13.27	4.79
2020-09	12.56	4.65
2020-12	9.3	4.83
2021-03	9.45	5.55
2021-06	9.75	5.36
2021-09	9.41	5.22
2021-12	9.03	4.59
2022-03	10.21	4.73
2022-06	10.93	4.51
2022-09	9.43	4.3
2022-12	9.74	3.97
2023-03	9.68	3.97
2023-06	9.91	3.92
2023-09	9.85	4.25
2023-12	9.81	4.25
2024-03	9.12	4.39
2024-06	9.29	4.41
2024-09	9.4	4.3
2024-12	9.77	4.05
2025-03	10	4.24
2025-06	10.27	4.21
2025-09	9.91	4.23
2025-12	9.91	4.05
2026-03	9.18	4.28
2026-06	9.9	4.35

September 19, 2025Licence revoked
CAME: licence revoked
The CNBV revoked its licence to operate as a Sofipo.
Source:Comisión Nacional Bancaria y de Valores (CNBV)Official
December 16, 2024Licence revoked
Auxi: licence revoked
The CNBV revoked its licence to operate as a Sofipo.
Source:Comisión Nacional Bancaria y de Valores (CNBV)Official
July 7, 2023Licence revoked
Popular Nacional: licence revoked
The CNBV revoked its licence to operate as a Sofipo.
Source:Comisión Nacional Bancaria y de Valores (CNBV)Official
January 17, 2023Licence revoked
Sierra Gorda: licence revoked
The CNBV revoked its licence to operate as a Sofipo.
Source:Comisión Nacional Bancaria y de Valores (CNBV)Official
October 31, 2022Merger
Te Creemos: merged
Ceased operating as an independent society after merging.
Source:Comisión Nacional Bancaria y de Valores (CNBV)Official
September 14, 2022Licence revoked
Planfía: licence revoked
The CNBV revoked its licence to operate as a Sofipo.
Source:Comisión Nacional Bancaria y de Valores (CNBV)Official
June 24, 2022Licence revoked
COFIA: licence revoked
The CNBV revoked its licence to operate as a Sofipo.
Source:Comisión Nacional Bancaria y de Valores (CNBV)Official

What this means if you operate in or invest in the sector

The figures above describe a market. These are the four facts that change an actual decision: how you get in, what unit economics you need, how you fund the book and how far you can grow.

1. Getting in is not applying for a licence, it is buying one

Of the 35 Sofipos in operation, 26 hold licences authorised between 2005 and 2009. Over the past fifteen years the CNBV has authorised only two new societies that remain active. The window closed.

That is why none of the fintechs that now dominate the sector applied for a licence: they bought an existing one. Klar operates on one authorised on 17 June 2008, when it was called SEFIA. Stori on another from the same day, then Mascaja. Nu México on one from 29 September 2006, then Akala. Fondeadora on one from 29 May 2006, then Apoyo Múltiple. The CNBV documents each change in the society's own name.

For anyone assessing entry into Mexico, this reorders the plan: the scarce asset is neither capital nor technology, it is the licence. And its price appears in no bulletin.

2. The unit economics this market demands

The sector charges an implicit rate of 45.47% and pays 4.63% on deposits. At a delinquency of 9.9%, headroom is 30.9 points before operating costs.

That is the order of magnitude to match. A model arriving in Mexico with European or US rates in mind does not cover this segment's expected loss. The calculator below lets you test combinations against the real societies.

Headroom calculator

Move the three controls and check whether the model holds. It is the same calculation this report applies to the 35 Sofipos: lending rate minus funding cost minus delinquency.

Lending rate45.5%

Funding cost4.6%

Expected delinquency9.9%

Headroom before operating costs

+30.9points

There is room

There is space to absorb operating costs, write-offs and a worse than expected deterioration. This is the zone digital Sofipos operate in.

Calibrate against real societies

CNBV figures at the latest cut. Tap one to load its values.

This calculation is deliberately conservative and is not an accounting measure of profitability: it ignores operating costs and treats non-performing loans as if they were entirely lost within the period. A positive result is a necessary condition, not a sufficient one.

3. Funding repriced, and it has a ceiling

The implicit cost of deposits went from 11.86% to 4.63% in a year. It is the largest improvement in the sector's economics and explains much of the return to profit, with ROE jumping from -32.74% to 2.74%.

But retail deposits carry a structural limit: the Protection Fund covers 25,000 UDIS per person. Funding a large book with small deposits works up to a point; beyond it, savers above the threshold demand a premium or leave. That ceiling shapes how far a book funded on deposits alone can grow.

4. There is a ladder, and someone climbed it

The Sofipo licence is not a destination, it is a rung. The CNBV authorised Nu to become a credit institution in April 2025 and the conversion materialised in August 2026, according to Banco de México's Financial Stability Report and press coverage.

So a documented path exists: acquire a Sofipo licence, build a customer base under lighter regulation, and apply to convert into a bank when the balance sheet justifies it. It is the only completed case so far, which is exactly why it is the reference for anyone planning the same route.

What to watch at the next cut

Five signals with today's value. The CNBV publishes the next bulletin in three months, so anyone can check for themselves whether they moved.

40.8pts
Sector spread
It is the cushion everything comes out of. Today there are 40.8 points between what is charged and what is paid, against delinquency of 9.9%. If the spread narrows faster than delinquency, the margin stops absorbing a bad year.
24.82%
Traditional Sofipos IMOR
They already charge less than they fail to collect. What decides whether the problem stays bounded is not this level but whether their book grows or shrinks: a shrinking group stops mattering, a growing one stops being bounded.
111.77%
Reserves booked over required
One hundred percent is the regulatory minimum, not a target. Below it there is a shortfall; exactly at one hundred there is no cushion. Several societies book precisely what is required, and nobody over-reserves by accident.
4.63%
Implicit funding cost
It fell from 11.86% to 4.63% in a year and it is what returned the sector to profit, with ROE going from -32.74% to 2.74%. A reversal eats the bottom line before any other variable.
35
Societies in operation
Six revocations since 2022 and one merger. Each exit concentrates the sector in fewer hands, and it already has 85.9% of its customers in four societies.

Source:Comisión Nacional Bancaria y de Valores (CNBV)Own calculation

The values are official; the reading of why each matters is this report's.

How to read this report

Every figure on these pages points to a source and carries an evidence label. Official data comes from CNBV and Banco de Mexico bulletins. Calculated figures are identified as own calculations and their formula is on the methodology page. Anything that could only be attributed to the press is marked as such and kept apart from the official series.

The data is public and reproducible: the dashboard lets you download every series as CSV.

Open the dashboard Methodology

Author

Roberto A. Arrucha

Director, Pro-Latam

LinkedIn Instagram