Midsize banks may have inherited their judgment, and they decide with less margin for error
Mexico's 16 midsize banks operate with less scale than the large ones, and a coordination error among the areas that decide their balance sheet would weigh more heavily on their results [H]. We start from a premise that public figures do not prove: midsize banks have the committees that decide the balance sheet—assets and liabilities, pricing, risk— [H] (section 7). Over the past year, 12-month return on equity (ROE) fell at 13 of the 16 [D]; the average liquidity coverage ratio (LCR) for April–June 2026 ranges from 127% to 1,048% [V]; and the delinquency ratio rose at 4 of the 5 consumer banks [D]. Each figure points to a different decision—the term at which liabilities are contracted, the size of the liquidity buffer, the duration taken on the asset side, and the capital each loan consumes—and at a midsize bank those decisions share the same balance sheet. Our hypothesis is that resolving those decisions together, with shared assumptions, internal charges, and limits, improves their evaluation [H]. Section 10 states what would confirm it and what would weaken it. This Perspective measures where each constraint binds, by business model and without naming the banks in the sample. It uses figures from Mexico's National Banking and Securities Commission (Comisión Nacional Bancaria y de Valores, CNBV), the banking supervisor, and leaves the reader 6 questions for their own bank.
0 · Structure of this note
Section 1 defines the sample and its models. Sections 2 and 3 examine profitability and funding. Sections 4 to 6 review liquidity, credit, and capital. Sections 7 and 8 connect those constraints to Decision Systems. Section 9 presents cases from other markets. Sections 10 and 11 gather testable hypotheses and questions for the committee. The note distinguishes verified facts, derived calculations, assumptions, and hypotheses about how decisions are made.
1 · The sample brings together 16 banks with 18.8% of assets, in 4 business models
The universe. The sample follows the selection criteria of the author of the original note. It includes Mexican-owned banks outside the 7 largest by assets. Each must hold at least 0.25% of commercial banking assets, equivalent to MXN 40.1 billion as of June 2026 [D]. Its business must correspond to one of the 4 models. Wholesale subsidiaries of foreign banks and neobanks are excluded. Under those criteria, 2 further domestic banks with more than 0.25% of assets are also excluded [D]. The author classifies 1 as a neobank. The balance sheet of the other is almost entirely repo, with no deposits from the public [D]. The 16 add up to MXN 3.02 trillion [D]. They represent 18.8% of the MXN 16.1 trillion in commercial banking assets as of June 2026 [D]. The 7 largest by assets hold 70.8% [D].
Model | Banks | Assets per bank, June 2026 | Share of system assets | Business profile | The constraint that binds today |
Regional commercial | 2 | From MXN 279 billion to MXN 411 billion | 4.3% | SME franchise and regional deposits | The margin, if assets reprice faster than liabilities when rates fall [H] |
Concentrated or specialized commercial | 6 | From MXN 62.5 billion to MXN 280 billion | 5.7% | Relationships with large borrowers, government, and midsize companies | Borrower concentration and funding with more market funding and repo [H] |
Mass-market or niche consumer | 5 | From MXN 44.7 billion to MXN 344 billion | 5.3% | Own sales network, high lending rates, first-time borrowers | The credit cycle; liquidity is ample at some banks and depends on the market at others [H] |
Treasury, foreign exchange, and wealth management | 3 | From MXN 63.5 billion to MXN 290 billion | 3.5% | Foreign exchange, markets, or funds business | Income depends on trading; funding depends on repo [H] |
Source: CNBV, Boletín Estadístico de Banca Múltiple (Commercial Banking Statistical Bulletin), June 2026, sheet EF Resumen [V]; shares calculated with full amounts, and grouping by the author of the original note [D]. 4.3% + 5.7% + 5.3% + 3.5% = 18.8%.
2 · Profitability fell at 13 of the 16, driven by rates at the regional banks and by credit in consumer lending
System 12-month ROE went from 17.7% in June 2025 to 16.4% in June 2026 [V]. Among the midsize banks it fell at 13 of 16 [D]. The cause varies with the model.
At the regional commercial banks, rate cuts coincide with lower earnings. Banco de México (Banxico), the central bank, set its target rate at 8% on June 26, 2025 [V]. A year later it stood at 6.5% [V]. Banxico held it at that level on June 25, August 6, and September 24, 2026 [V]. That is 150 basis points less between the 2 cutoff dates, with the rate now on hold [D]. The 2 regional banks grew their loan books by between 4.72% and 11.8% [D]. They earned 10.7% less in January–June 2026 than a year earlier [D]. Their 12-month ROE ended at 18% and 18.1%, down from 21.8% and 21.4% [V]. Higher volume with lower earnings is consistent with a margin that compresses when assets reprice before liabilities [H]. Testing this requires yields and costs on average balances for each bank. Section 10 sets that comparison as the test that discriminates H1.
In consumer lending, the pressure comes from loan-loss provisions. At 3 of the 5 consumer banks, January–June 2026 earnings were between 21% and 86% lower [D]. The comparison is with the same half of 2025 [D]. In the extreme case, 1 bank lost MXN 1.37 billion in the second quarter of 2026 [D]. Its loan-loss provisions were MXN 6.27 billion in the quarter, MXN 2.68 billion more than a year earlier [D].
In treasury, ROE rests on trading income. At 2 of the 3 banks in the model, January–June 2026 trading income exceeded net interest income [V]. The amounts were MXN 3 billion to MXN 4.3 billion, against MXN 764 million to MXN 2.06 billion [V]. Their 12-month ROE ended between 20.5% and 23% [V]. The third fell to 4.82% [V]. The recurrence of that income is what a bank should measure before projecting it [H].
Model | 12-month ROE, June 2026 | Banks with lower ROE than in June 2025 | January–June 2026 earnings vs. 2025 |
Regional commercial | From 18% to 18.1% | 2 of 2 | −10.7% at both |
Concentrated commercial | From 7.2% to 21.2% | 4 of 6 | From −20% to +36.7% |
Consumer | From 8.79% to 31% | 4 of 5 | From −85.8% to +66.9% |
Treasury and foreign exchange | From 4.82% to 23% | 3 of 3 | From −64% to −2.08% |
System | 16.4% | — | — |
Source: CNBV, Boletín Estadístico de Banca Múltiple, June 2026, sheets Indicadores and Pm2 [V]; changes calculated by InsightIA [D]. Banxico, monetary policy announcements of June 26, 2025, and of June 25, August 6, and September 24, 2026 [V].
3 · Midsize bank liabilities follow 3 distinct structures, and the mix sets how much they cost and how long they last
As of June 2026, the system funds itself 52% with demand deposits and 17% with repo [D]. The measure is taken over interest-bearing liabilities, which comprise deposits, interbank loans, and repo [D]. The midsize banks fall into 3 structures:
Demand deposit funding, at 2 consumer banks. Demand deposits make up between 72% and 92% of their interest-bearing liabilities [D]. This funding is cheap and dispersed. Its stability under stress depends on depositor concentration, which the public figure does not show [H]
Market and repo funding, at the 3 treasury banks and at 1 consumer bank: between 56% and 84% of interest-bearing liabilities [D]. In treasury, repo mainly finances the securities position [D]
Term market funding, at 1 consumer bank that takes almost no demand deposits. Demand deposits represent 9.2% of its interest-bearing liabilities [D]. The bank funds itself with loans from banks and other entities and with certificados bursátiles (local debt securities) [D]. These represent 85% of its interest-bearing liabilities [D]. That is the maximum in the consumer column of the table
Model | Demand deposits over traditional deposits | Market and repo funding over interest-bearing liabilities |
Regional commercial | From 45% to 51% | From 20% to 32% |
Concentrated commercial | From 18% to 58% | From 17% to 59% |
Consumer | From 3.7% to 95% | From 3.5% to 85% |
Treasury and foreign exchange | From 32% to 63% | From 56% to 83% |
System | 66% | 28% |
Source: CNBV, Boletín Estadístico de Banca Múltiple, June 2026, sheet EF Resumen [V]; ratios calculated by InsightIA [D]. Market and repo funding: money market term deposits, debt securities issued, interbank loans, and repo payables, divided by traditional deposits plus interbank loans plus repo.
The liability mix is the first balance sheet decision. The price and term of liabilities set the cost of funding each loan and the buffer needed to cover outflows. They also set the sensitivity of the margin to rates [H]. A bank may decide the mix in treasury, the buffer in risk, and the loan price in the pricing committee. It then resolves 3 parts of the same decision with 3 different figures [H].
4 · Midsize bank liquidity sits at the ends of the corridor, and the term of liabilities may explain more than the amount
The LCR is the ratio of a bank's eligible liquid assets to its net cash outflows over 30 days [V]. If it is at least 100% every day of the previous month, the bank falls in scenario I of the liquidity provisions [V]. In that scenario it faces no corrective measures [V]. The average April–June 2026 LCR ranges from 127% to 1,048% across the 16 [V]. The extremes do not line up with any one model:
Model | Average LCR, April–June 2026 |
Regional commercial | From 127% to 194% |
Concentrated commercial | From 127% to 302% |
Consumer | From 136% to 1,048% |
Treasury and foreign exchange | From 212% to 362% |
Source: CNBV, Coeficiente de Cobertura de Liquidez y Coeficiente de Financiamiento Estable Neto (Liquidity Coverage Ratio and Net Stable Funding Ratio), April–June 2026, individual LCR [V]. Banxico and CNBV, Disposiciones sobre los requerimientos de liquidez (Provisions on Liquidity Requirements), arts. 9, 11, and 12, and art. 13 for the scenarios [V]. The 1,048% is written as the CNBV publishes it.
Ample slack at one end, little slack at the other. The 2 consumer banks funded with demand deposits operate with LCRs of 565% and 682% [V]. If they keep that slack in liquid assets that yield less than loans, it carries a margin cost [H]. At the other end, 5 banks operate between 127% and 149% [V]. They are 1 regional, 3 concentrated, and 1 consumer bank with market funding [V]. They have less room for an outflow the rule does not anticipate [H].
The LCR measures 30 days, and the term decides what falls inside that window. The provisions count as an outflow any liability that matures in 30 days or less [V]. They allow cash inflows to offset up to 75% of those outflows [V]. A market liability contracted for several years barely enters the window. A loan book that rolls over in a few months generates inflows the rule recognizes [H]. A high LCR can therefore result from the term of liabilities rather than from an idle buffer. A low LCR can be comfortable if 30-day outflows are predictable [H]. The number that matters is how much margin each point of buffer costs and how much outflow risk it avoids [H].
The liquidity corridor has 3 bands. In the practice of asset-liability committees, banks operate above the regulatory threshold of 100% [H]. There is an appetite limit, typically around 115%, and a management limit, around 130% [H]. These are typical values of the trade, not of a regulation. Above the management limit, slack stops protecting the bank and starts costing it [H]. That is why the slack should be measured in pesos, together with its opportunity cost, in addition to the ratio. It is calculated by subtracting the management limit multiplied by net outflows from eligible liquid assets [H].
The liquidity corridor is a pricing decision. A liquidity Decision System translates the buffer into a charge per product. The charge is allocated according to how much buffer each product forces the bank to hold. That charge enters the price of loans and deposits [H]. Without that charge, treasury defends the ratio and the commercial area defends volume. The asset-liability committee then arbitrates figures instead of deciding [H].
5 · In consumer lending, delinquency rose at 4 of 5 banks, and the cost already shows in results
Between June 2025 and June 2026, delinquency is measured with the CNBV definition [V]. It is the stage 3 loan book divided by the loan book in stages 1, 2, and 3 [V]. It rose at 4 of the 5 consumer banks [D]. Among the concentrated commercial banks it was split. It fell at 3 of the 6 and rose at the other 3 [D]. During that period, the loan book of that model grew by up to 48.9% [D].
Model | Delinquency ratio, June 2026 | Loan-loss provisions over net interest income, January–June 2026 | Loan book growth, June 2025 to June 2026 |
Regional commercial | From 1.34% to 1.81% | From 7.65% to 10% | From 4.72% to 11.8% |
Concentrated commercial | From 2.21% to 4.57% | From 7.54% to 26.4% | From −0.74% to 48.9% |
Consumer | From 1.46% to 6.21% | From 10.9% to 74.2% | From 7.99% to 21.5% |
Treasury and foreign exchange | From 0.93% to 3.41% | From −0.28% to 15.6% | From −0.56% to 14.2% |
System | 2.3% | 25.4% | 7.84% |
Source: CNBV, Boletín Estadístico de Banca Múltiple, June 2026, sheets CCT, EF Resumen, and Portadilla II, note 5/ [V]; ratios calculated by InsightIA [D].
At 1 consumer bank, loan-loss provisions equaled 74.2% of net interest income for the half-year, against 25.4% for the system [D]. At 2 others, they equaled 28.6% and 34.4% [D]. Provisions can grow faster than the margin. If that happens, earlier pricing and origination may have been set with a cost of risk that is no longer current [H]. In consumer lending, loan recovery is the decision that sets that cost. It covers which account is pursued, with what intensity, and with which provider. A bank should check whether that cost reaches the price or is managed only as an operation [H].
6 · Capital is ample in aggregate and concentrated in the detail
As of July 31, 2026, all 16 were in early-warning category I [V]. Their capitalization ratio (índice de capitalización, ICAP) ranged from 13.1% to 32.2%, against 20.8% for the system [V]. At 6 of them, the fundamental capital ratio (coeficiente de capital fundamental, CCF) is below their ICAP [V]. The CCF is the highest-quality capital over risk-weighted assets [V]. The difference reaches up to 5.22 percentage points [D].
The binding constraint may be borrower concentration, ahead of regulatory capital [H]. HR Ratings, a credit rating agency, reports the balances of the 10 largest borrowers at 2 of the 6 concentrated commercial banks [V]. Those balances added up to 3 times stockholders' equity at one bank, at year-end 2025 [V]. They added up to 1.6 times at the other, as of June 2026 [V]. At that second bank, the ratio was 1.1 times 12 months earlier [V]. The 2 figures have different cutoff dates and are neither added nor averaged. With that concentration, the default of 1 borrower is enough to move a year's results [H].
Model | ICAP, July 2026 | CCF where it differs from ICAP |
Regional commercial | From 15% to 15.9% | — |
Concentrated commercial | From 13.1% to 26.5% | From 10.1% to 16.5%, at 4 of 6 |
Consumer | From 13.7% to 32.2% | From 10.1% to 13.6%, at 2 of 5 |
Treasury and foreign exchange | From 17.3% to 22.4% | — |
System | 20.8% | 17.2% |
Source: CNBV, Índice de Capitalización de la Banca Múltiple al cierre de julio de 2026 (Commercial Banking Capitalization Ratio as of the End of July 2026), press release of September 18, 2026 [V]; HR Ratings, reports of May 25 and October 12, 2026, published on the Mexican Stock Exchange (Bolsa Mexicana de Valores, BMV) website [V].
The price of a large loan has to include what the loan consumes. A loan that concentrates risk consumes capital, liquidity, and loss-absorbing capacity that the rest of the book can no longer use. If the transfer rate covers only the cost of funding and interest rate risk, that consumption does not reach the price. The bank charges its largest borrower as it would charge a small one [H].
7 · Midsize banks have the committees, and what they may lack is resolving together the decisions that share the balance sheet
The original note holds that midsize banks inherited judgment. It refers to the asset-liability committee, the pricing committee, and the risk committee [H]. That is the premise of the title, and it is taken as the author's expert judgment. Public figures do not prove it. Taking part in the committee is different from deciding in it. A committee that exists, with every area attending, does not guarantee that liabilities, liquidity, duration, and capital are resolved jointly. When each area brings its own figure, the committee debates the data and not the decision [H].
InsightIA reads a bank's balance sheet as 9 coupled Decision Systems, its own framework developed in the Perspective SF-3 [H]:
# | Decision System | What is decided | How it couples with the others |
1 | Balance sheet allocation and structure, the core | Liability mix and price, liquidity corridor, duration, and allocation of balance sheet capacity | Issues the transfer curve, the liquidity buffer charge, and the capital charge |
2 | Credit pricing, limits, and supply | Whom to lend to, with what amount, price, and limit | Receives the prices set by the core |
3 | Commercial, service, operational, and technology capacity | What capacity to build and in what order | By limit and by price |
4 | Markets: inventory, pricing, and hedging | What inventory to take, at what price, and with what hedge | By price and by limit |
5 | Recovery of impaired loans | What intervention to make on each account and with what intensity | Produces the recovery cost that pricing should use |
6 | Liquidity, collateral, and intraday payments | How the position moves during the day | By limit, price, and state |
7 | Allocation of service capacity and offers | Through which channel and with which offer to reach each customer | By limit, price, and state |
8 | Fraud, authorization, and intervention | Where thresholds are set | By limit and state |
9 | Targets, budget, and capitalization framework | Appetite, allocation of ambition, and capitalization | Sets the objective function of the core |
The regime changes the binding constraint. In a rate-cutting cycle, like the one from June 2025 to June 2026, margin sensitivity binds. The core decision becomes one of duration. Near the internal liquidity limit, liquidity binds. With ample slack, profitability binds [H]. The same liability price is right in one regime and wrong in the others. A bank that does not know which regime it is in sets prices with the previous quarter's constraint [H].
A capacity bought at a known cost is expressed as a price, but its limits remain. External collections and market funding are contracted at a cost. At a midsize bank that cost weighs more on results, and availability has a ceiling [H].
8 · Each model has the Decision System that binds it most
Decision System | Regional commercial | Concentrated commercial | Consumer | Treasury and foreign exchange |
1 · Core: liability mix and price | High: market funding from 20% to 32% | High: market and repo funding from 17% to 59% | Medium: 2 banks with demand deposits, 1 with 84% market funding | High: 56% to 83%, mainly repo |
1 · Core: liquidity corridor | High: LCR from 127% | High: LCR from 127% to 302% | High: LCR from 136% to 1,048% | Medium: LCR from 212% to 362% |
1 · Core: duration and margin | High: rate on hold after the cuts; earnings −10.7% | Medium | Low: high lending rates | Medium |
2 and 5 · Credit and recovery | Low: delinquency from 1.34% to 1.81% | High: borrower concentration; delinquency from 2.21% to 4.57% | High: provisions of up to 74.2% of the margin | Low |
1 and 9 · Capital charge and capitalization framework | Medium | High: CCF from 10.1% | Medium: ICAP from 13.7% | Low: ICAP from 17.3% to 22.4% |
4 · Markets | Low | Low | Low | High: trading income exceeds the margin at 2 of 3 |
The rating is InsightIA's judgment on sections 2 to 6 [H]: it indicates where tension shows in public figures, not whether the bank has a Decision System to resolve it or how it decides today. The 2 and 5 row measures credit and recovery with aggregate figures; account-by-account recovery cannot be read from outside. Capacity (3), intraday (6), service and offers (7), and fraud (8) have no row because no public figure measures them.
9 · In other markets, midsize banks broke when liquidity, rates, and concentration moved together
Case | What went wrong for them | What happened | Reading for Mexican midsize banks |
Mexico, 2025 · 2 midsize banks placed under intervention | Compliance and liquidity | After the measures the United States Department of the Treasury announced in June 2025, between May and June one bank lost 25.1% of its deposits, MXN 12.8 billion, and the other lost 8.27% [D]. The CNBV ordered the temporary management intervention of both on June 26, 2025 [V]; the first entered liquidation under the Instituto para la Protección al Ahorro Bancario (IPAB), Mexico's deposit insurer, on October 10 [V] | A reputational crisis can turn into a liquidity crisis within weeks [H] |
United States, 2023 · a regional bank in California | Rates, liquidity, and depositor concentration, all at once | At year-end 2022, about 94% of its holding company's deposits were uninsured. The bank lost more than USD 40 billion on March 9, 2023, and expected to lose more than USD 100 billion the next day: about 85% of its deposits. In April 2022, the holding company changed its deposit duration assumptions to avoid breaching a limit and, in the same period, removed hedges [V] | Liabilities, liquidity, and duration moved together, as the core would when acting as a single decision [H] |
Spain, 2010 to 2014 · savings banks | Sector concentration and scale | The sector went from 45 savings banks at the start of 2010 to 11 entities or groups in December 2014 [V] | Concentrated banks consolidated when capital stopped being enough [H] |
Sources: CNBV, Boletín Estadístico for May and June 2025 and press releases of June 26 and October 13, 2025; Federal Reserve, review of the supervision of that bank, April 2023; CECA, Proceso de Reestructuración del Sector de Cajas de Ahorros (Restructuring Process of the Savings Bank Sector), March 2015.
Two lessons for Mexico.
Liabilities, liquidity, and duration are a single decision. The first 2 cases show that a concentrated or reputation-sensitive deposit base can leave within days. The buffer therefore has to be decided with liability concentration built in, in addition to the ratio [H]
A midsize bank's risk lies in concentration. The Spanish savings banks show that a single sector can force consolidation. In Mexico, borrower concentration at the specialized commercial banks is the local version of that risk [H]
10 · The central hypothesis and 4 hypotheses by constraint, each with what would confirm or weaken it
The figures suggest 5 hypotheses. The CNBV quarterly series and each bank's internal data can confirm or weaken them. All 5 are alive and untested.
H0 · Resolving together the decisions that share the balance sheet—with shared assumptions, internal charges, and limits—improves their evaluation [H]. To confirm it, one would compare banks of the same model facing the same rate change. Those that publish a transfer price with a liquidity charge and a capital charge would be set apart. A less sensitive margin and more stable earnings than among banks that do not publish one would strengthen the hypothesis. The absence of that difference in the third and fourth quarters of 2026, controlling for model and size, would weaken it. So would a finding that business mix and size alone explain the dispersion of results. Data: each bank's risk management and transfer pricing disclosures, and the CNBV quarterly series
H1 · Banxico's cuts compress the margin of midsize commercial banks more, because their assets reprice before their liabilities [H]. What discriminates it: each bank's margin sensitivity to rates, with interest earned and paid on each quarter's average balances
H2 · Part of the high liquidity carries an avoidable cost, and part of the low liquidity is sensitive to outflows [H]. To discriminate it, one would observe whether the July–September 2026 LCR stays within the range of 127% to 1,048%. Internal data would show whether the buffer can be reduced without breaching limits under stress. They would also show whether it proves insufficient in plausible scenarios
H3 · Borrower and depositor concentration binds before regulatory capital [H]. What discriminates it: whether the price and limit of credit to large borrowers incorporate the capital and liquidity they consume
H4 · In consumer lending, pricing and origination incorporate credit deterioration with a lag [H]. To discriminate it, one would observe whether the consumer banks' loan-loss provisions over margin keep rising from July to September 2026. Internal data would show whether expected losses on new cohorts exceed those the price includes
11 · Six questions for the asset-liability committee of a midsize bank
Liquidity corridor and liability mix. At what point in the corridor should the bank operate, how much slack in pesos is there above the management limit, and what lending could that slack fund without breaching limits under stress?
Margin with the rate on hold. What duration should the bank take on the asset side, and with what hedge, with the target rate on hold at 6.5%, and with what sensitivity figure does the committee decide it?
Recovery cost as price. Does this quarter's consumer loan pricing use the current recovery cost or last year's?
The price of concentration. How much capital and how much liquidity does your largest borrower consume, and is that consumption in the rate you charge?
Capital and dividends. Are dividend payouts and loan book growth decided with the same capital constraint?
Scale. Is your technology capacity an investment decision of your own or a consolidation decision that has not yet been framed as such? [H]
Each question is a decision that today may be resolved with figures from different areas; a Decision System resolves it with a single one [H]. This Perspective will be reissued when the CNBV publishes the third quarter of 2026, at the end of November, and H0 to H4 will then have their first public data point to discriminate them.
Sources
CNBV · Boletín Estadístico de Banca Múltiple (Commercial Banking Statistical Bulletin), June 2026, and those of March 2026, June 2025, May 2025, and March 2025 · https://portafolioinfo.cnbv.gob.mx/Paginas/Contenidos.aspx?ID=40
CNBV · Coeficiente de Cobertura de Liquidez y Coeficiente de Financiamiento Estable Neto (Liquidity Coverage Ratio and Net Stable Funding Ratio), April–June 2026 · https://portafolioinfo.cnbv.gob.mx/PortafolioInformacion/CCL%20CFEN%202T2026%20(1%20hoja).pdf
CNBV · Índice de Capitalización de la Banca Múltiple al cierre de julio de 2026 (Commercial Banking Capitalization Ratio as of the End of July 2026) · https://portafolioinfo.cnbv.gob.mx/PortafolioInformacion/ICAP_BM_202607.pdf
Banxico · Monetary policy announcements of June 26, 2025, and of June 25, August 6, and September 24, 2026 · https://www.banxico.org.mx/publicaciones-y-prensa/anuncios-de-las-decisiones-de-politica-monetaria/anuncios-politica-monetaria-t.html
Banxico and CNBV · Disposiciones de carácter general sobre los requerimientos de liquidez para las instituciones de banca múltiple (General Provisions on Liquidity Requirements for Commercial Banks), compiled text, arts. 9, 11, 12, and 13
HR Ratings · rating reports of May 25 and October 12, 2026, published at https://www.bmv.com.mx
CNBV · Comunicado conjunto (Joint Press Release) of June 26, 2025, and Comunicado No. 021 (Press Release No. 021) of October 13, 2025 · https://www.gob.mx/cnbv/archivo/prensa
Federal Reserve · Review of the Supervision and Regulation of Silicon Valley Bank, April 2023 · https://www.federalreserve.gov/publications/files/svb-review-20230428.pdf
CECA · Proceso de Reestructuración del Sector de Cajas de Ahorros (Restructuring Process of the Savings Bank Sector), March 13, 2015 · https://www.ceca.es/wp-content/uploads/2021/07/Reestructuracion-13-3-2015.pdf
InsightIA · Perspective SF-3, on the 9 Decision Systems of the bank balance sheet, October 2026 · own framework, linked once it is on the site
