## INTRODUCTION

## Source details

**Canonical URL:** [INTRODUCTION](https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021114-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2021/english/wpiea2021114-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2021/english/wpiea2021114-print-pdf.pdf.json)

---

### Structural changes and fintech emergence
- Financial liberalization in the 1990s was followed by a wave of foreign acquisition and consolidation in LAC; foreign presence waxed and waned through financial and sovereign debt crises.
- The use of financial technology (“fintech”) in banking services has emerged and grown rapidly in recent years (Berkmen et al., 2019).
- According to the Inter-American Development Bank, the number of fintech companies in Latin America grew 66 percent (from 703 to 1166) between 2017 and 2018.
- Fintech firms have been adept at unbundling traditional banking services, offering targeted solutions including reaching out to unbanked segments; they both complement and challenge traditional banks.

### How fintech may affect banking competition
- Different fintech entry types imply different effects:
  - Parallel-market entrants may not threaten incumbents.
  - Direct competitors such as peer-to-peer lending platforms can bypass banks entirely.
- Mansilla-Fernandez (2017) finds fintech investments higher in countries with less competitive banking sectors (greater concentration and higher net interest margins), suggesting fintech could foster more competition.
- Incumbent banks have increased IT investment and direct investment in fintech startups, indicating they see fintech as a potential threat and a way to maintain market relevance.
- Regulatory environment is an important force:
  - A level playing field between fintech firms and traditional banks could increase competition and lower concentration.
  - If incumbents acquire fintech or out-innovate newcomers, fintech may have little impact on banks’ concentration even if inclusion and efficiency improve.

### Purpose and empirical approach
- Revisits the competitive environment of the banking system in LAC and investigates the early impact of fintech in the region through empirical and case-study approaches.
- Empirical strategy:
  - Document stylized facts on the competitive landscape (notably high NIMs and limited financial depth).
  - Re-examine associations between changes in financial sector structure (concentration, foreign bank entries, M&A) and banking sector competition using panel regressions on bank-level data.
  - Extend panel regressions into a difference-in-differences (diff-in-diff) setting to illustrate the association of fintech emergence with changes in the competitive environment.
- Case studies focus on Brazil and Mexico, the two most developed fintech markets in the region.
- Analytical focus is for pre-pandemic periods; the paper does not address COVID-19 implications due to lack of data.

### Key stylized facts on LAC banking competition
- LAC is characterized by limited financial depth and high interest spreads; countries in LAC are among those with the highest net interest margins (NIMs) in the world.
- Cross-country and within-country heterogeneity:
  - In Argentina, Brazil, Chile, and Peru, private domestic banks have higher NIMs than private foreign-owned banks.
  - In Bolivia, Colombia, and Mexico, foreign banks have slightly higher NIMs than domestic banks.
  - Public banks typically have substantively lower NIMs than private domestic banks, with Uruguay an exception where public banks have higher NIMs.
- Lerner index (markups) evidence:
  - Brazilian and Uruguayan banks have among the highest NIMs yet appear to have lower markups, suggesting high NIMs may be driven by high costs rather than market power.
  - Colombian banks have high markups despite moderate NIMs.
  - Peruvian banks have both high markups and high NIMs, consistent with low competition.
- Concentration:
  - Regional 3-bank concentration: Emerging and Developing Asia (EDA) 51.6 percent; Advanced Economies (AEs) 68.4 percent; LAC 63.5 percent.
  - Across regions, assets of the top three banks on average make up more than half of total banking sector assets; concentration fell compared to the previous decade, with LAC showing the smallest improvement.
- Policy relevance: high NIMs imply higher costs of financial intermediation and limited financial inclusion; improving competition to lower NIMs is linked to financial deepening and inclusion.

### Data and empirical framework
- Bank-level data: historical balance sheet data for 692 banks from 1988 to 2018 from the Fitch Connect Database for: Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Mexico, Peru, and Uruguay.
- Ownership and M&A: Orbis, Zephyr, and banking institutions’ websites.
- Macroeconomic/macrofinancial: IMF World Economic Outlook and Financial Development Index databases.
- Definitions and measurements:
  - Foreign-owned banks: parent institutions with more than 50 percent foreign ownership.
  - Public banks: more than 50 percent government or subnational ownership.
  - NIM measured as (gross interest income + dividend income – total interest expense) as a percent of average assets.
  - Concentration measured mainly by the Herfindahl-Hirschman Index (HHI) and 3-bank concentration.

### Baseline regression findings (key quantitative results)
- Overall finding: more competitive environment is associated with lower NIMs, but foreign bank entry and M&A activities may be associated with higher NIMs.
- Market concentration:
  - Lower NIMs are associated with less concentrated markets.
  - One standard deviation increase in the top three banks’ share is associated with over 110 basis points higher NIM.
- Foreign banks:
  - Foreign banks have lower NIMs on average: all else equal, foreign banks’ NIMs are around 100 basis points lower than those of domestic banks.
  - However, NIMs increase with the foreign bank’s market share.
- M&A:
  - M&A activities involving foreign banks are associated with higher NIMs; total M&As (local and foreign) are not.
  - The foreign-M&A association may reflect takeovers of less efficient domestic banks with higher NIMs; taking 1- or 2-year lags can eliminate the association.
- Bank-specific characteristics:
  - Liquidity: each percentage point increase in liquid to total assets is associated with a 2 basis points decrease in the NIM.
  - Non-performing loans (NPLs): higher NPLs associated with lower NIMs.
  - Equity ratios: banks with higher equity ratios see higher NIMs.
  - Non-interest (fee) income: banks with higher fee incomes have lower NIMs.
  - Administrative costs: a one-standard deviation increase in administrative costs is associated with 440 basis points increase in NIM.
  - Quantitatively, one standard deviation changes in bank-specific characteristics are associated with 40-70 basis points changes in NIMs (except administrative costs noted above).
- Macroeconomic influences:
  - Higher NIMs are associated with stronger real GDP growth and higher real interest rates.
  - NIMs decrease with total credit in the economy and EMBI spreads.
- Structural change post-GFC:
  - Margin differentials between domestic and foreign banks have more than halved since 2008.
  - Margin differentials between private and public banks have also diminished.
- Robustness with markups (Lerner index):
  - Higher markups associated with higher concentration (with lower statistical significance than NIM regressions).
  - Foreign banks on average have lower markups.
  - Public banks have higher markups on average in Lerner regressions even if NIM differences are not statistically different.

### Regressions incorporating fintech (diff-in-diff exploratory exercise)
- Fintech lending volumes remain very small: fintech lending in countries with data makes up only about 0.02 to 0.05 percent of total loans.
- Diff-in-diff specification:
  - D_{t>2012} = 1 for time periods after 2012 (when fintech activity started to take place in the region).
  - D_{Fintech} = 1 for countries with fintech lending activity.
  - Alternate specification replaces D_{Fintech} with fintech credit share (fintech “intensity”) — volume of fintech credit to total credit in the economy.
- Key exploratory result:
  - Difference-in-differences coefficient has the expected negative sign, suggesting banks in fintech countries have seen greater reductions in NIMs than those in non-fintech countries in periods after fintech lending emerged.
  - Unobserved time-invariant differences between fintech and non-fintech countries are associated with higher baseline NIMs in fintech countries (positive and significant β2), consistent with fintech developing where intermediation prices are higher.
- Caveats:
  - Results are exploratory given sparse fintech quantitative data and the very small overall fintech volumes; findings should be interpreted with caution.

### Case studies: Brazil and Mexico
Common threads across both countries:
- Fintech and bigtech are progressing more rapidly in payments than in lending.
- Many fintech startups focus on niche markets and the unbanked, enhancing financial inclusion.
- Incumbent banks respond by investing in innovation, improving efficiency and customer service, and partnering with fintechs.
- Partnerships can improve efficiency but effects on competition are ambiguous—incumbents may increase markups rather than lowering fees or rates.
- Encouraging signs that fees are declining in the payments segment in both countries.

Brazil:
- Banking sector assets around 105 percent of GDP; the five largest banks account for over 80 percent of total assets.
- Share of adult population with financial accounts: 86.5 percent (Central Bank of Brazil, 2018).
- Brazil had more than 650 fintech companies operating in the country at end-2019.
- Payments impact:
  - Merchant discount rates (MDRs) fell from nearly 3 to around 2½ percent after entrance of acquirers including fintechs and bigtechs.
- Fintech lending:
  - Fintech lenders’ credit growth outpaced largest banks but remain small in size compared with incumbents.
- Regulatory approach:
  - Regulators have accommodated fintech within existing frameworks using segmentation and proportionality.
  - Notable regulations and initiatives:
    - Payments: BCB Circular 3682 (November 2013), BCB Circular 3885 (March 2018), and BCB Resolution 80 (March 2021) revoking Circular 3885; four types of services allowed for payment institutions (electronic money issuer; post-paid payment instrument issuer; acquirer; payment transaction initiator (PISP)).
    - Securities: Regulation 588 (July 2017, CVM) crowdfunding registration exception.
    - Credit: CMN Resolution 4656 (April 2018) created SCD (Sociedade de Crédito Direto) and SEP (Sociedade Entre Pessoas) for digital credit origination; since 2019, 24 SCDs and 6 SEPs licensed.
    - Open banking: announced May 2020; implementation started February 2021 and will be fully operational by end-2021.
    - Regulatory sandboxes: joint statement June 2019; Susep, CVM, BCB have sandbox rules and have selected or expect to select projects (Susep authorized 11 projects in first cycle; CVM 7; BCB between 10 and 15).
  - Central Bank initiatives: Agenda BC+ (2016) and Agenda BC# (updated agenda promoting inclusion, competitiveness, transparency, financial education, and sustainability); Financial and Technology Innovation Laboratory (LIFT) launched 2018.
- Incumbent responses:
  - Digitization and branch reduction; incumbents launching digital banks (e.g., Banco Bradesco’s Next), partnering on digital products, or converting to digital banks (Banco Original, Banco Inter).
  - Banks funding fintech startups and creating innovation hubs (Cubo Itaú, InovaBra, Banco do Brasil’s fintech lab).

Mexico:
- Banking sector assets around 40 percent of GDP; bank credit as share of GDP is about one-quarter of levels in other emerging markets.
- Financial inclusion: 47 percent of adult population with financial accounts; about 70 percent of individual borrowers rely on informal credit.
- Fintech counts and growth:
  - Number of fintech startups reached 441 by March 2020 (Finnovista); average annual growth rate since 2016 of 23 percent.
  - By 2018 fintech credit share of total formal credit was 0.06 percent.
  - Most fintechs engage in payments and transfers (88 firms) and consumer and business lending (83 firms).
- Bigtechs and partnerships: Amazon introduced Amazon Cash (2017) and Amazon Rechargeable (2018) in partnership with Banorte; Konfio partnered with Facebook and PayPal.
- Regulatory response:
  - Fintech Law enacted March 2018 — first comprehensive fintech legislative framework in Latin America.
  - Law’s guiding six principles: (i) promote financial inclusion and innovation, (ii) foster competition, (iii) protect consumers, (iv) strengthen AML/CFT, (v) maintain financial stability, (vi) observe technological neutrality.
  - Recognizes and regulates: crowdfunding (including P2P), e-money platforms (FTIs); addresses virtual assets, APIs (open banking), and regulatory sandboxes.
  - Regulators: Banxico, CNBV, and SHCP (with other agencies such as CONDUSEF, CONSAR, CNSF involved in related roles).
  - Secondary regulations (September 2018) required FTIs to apply for licenses by September 2019; requirements included:
    - Minimum capital stock of 500,000 UDIs (approx. US$146,000) or 700,000 UDIs (approx. US$204,000) depending on activity.
    - Maximum limits on cash transfers per client and per day for electronic payment and crowdfunding platforms.
    - Maximum limits on lending activities for crowdfunding and P2P platforms.
    - Investor diversification obligations for crowdfunding platforms.
    - Reporting requirements for acquisitions of 10 percent or more of FTI shareholding.
  - Implementation challenges:
    - Only 85 out of 124 fintech entities that require a license filed applications by the September 2019 deadline.
    - First license issued January 2020 to NVIO Pagos Mexico (Bitso); authorization process interrupted by the Covid-19 pandemic.
    - Survey evidence: 52 percent of fintech startups view the Fintech Law as presenting high barriers to entry; an Endeavor Mexico study suggests 63 percent of startups could not absorb estimated compliance costs.
  - Supportive measures:
    - Banxico’s CoDi (Cobro Digital Systems) QR-code payment platform: five million users as of September 2020; participating banks include all traditional banks.
    - Regulatory sandbox and open banking promoted; CNBV’s “Sandbox Challenge” to encourage applicants.
- Incumbent responses and market dynamics:
  - Traditional banks invest in digital factories and collaborations (e.g., Santander’s Spotlight, BBVA’s Fast Track).
  - Collaboration and funding: Klar received US$57.5 million from Santander in 2019; BBVA Mexico acquired OpenPay (2016); Santander announced plan to acquire Elavon México (February 2020).
  - Finnovista projects fintech startups could take up to 30 percent of Mexico’s banking market over the next decade (forecast cited).

### Conclusions and policy implications
- Early evidence indicates fintech is associated with lower NIMs and is influencing incumbent banks’ behavior (fintech adoption, fee reductions in payments, digitalization).
- Whether fintech will reduce banking sector concentration remains unclear and depends on entry forms, incumbents’ responses (acquisitions, partnerships), and regulatory frameworks.
- Open banking could be a gamechanger by granting equal access to customer data across the financial system, potentially levelling the playing field for fintech firms.
- New challenges: data privacy, cybersecurity, AML/CFT, fraud, and cross-border transaction supervision.
- Regulators face a balancing act: promote innovation and competition, protect consumers, mitigate risks, and collaborate internationally to share best practices and prevent regulatory arbitrage.

*Source: INTRODUCTION, wpiea2021114-print-pdf*

### INTRODUCTION

### INTRODUCTION

### Structural changes and fintech emergence
- Financial liberalization in the 1990s was followed by a wave of foreign acquisition and consolidation in LAC; foreign presence waxed and waned through financial and sovereign debt crises.
- The use of financial technology (“fintech”) in banking services has emerged and grown rapidly in recent years (Berkmen et al., 2019).
- According to the Inter-American Development Bank, the number of fintech companies in Latin America grew 66 percent (from 703 to 1166) between 2017 and 2018.
- Fintech firms have been adept at unbundling traditional banking services, offering targeted solutions including reaching out to unbanked segments; they both complement and challenge traditional banks.

### How fintech may affect banking competition
- Different fintech entry types imply different effects:
  - Parallel-market entrants may not threaten incumbents.
  - Direct competitors such as peer-to-peer lending platforms can bypass banks entirely.
- Mansilla-Fernandez (2017) finds fintech investments higher in countries with less competitive banking sectors (greater concentration and higher net interest margins), suggesting fintech could foster more competition.
- Incumbent banks have increased IT investment and direct investment in fintech startups, indicating they see fintech as a potential threat and a way to maintain market relevance.
- Regulatory environment is an important force:
  - A level playing field between fintech firms and traditional banks could increase competition and lower concentration.
  - If incumbents acquire fintech or out-innovate newcomers, fintech may have little impact on banks’ concentration even if inclusion and efficiency improve.

### Purpose and approach of the paper
- Revisits the competitive environment of the banking system in LAC and investigates the early impact of fintech in the region through empirical and case-study approaches.
- Empirical strategy:
  - Document stylized facts on the competitive landscape (notably high NIMs and limited financial depth).
  - Re-examine associations between changes in financial sector structure (concentration, foreign bank entries, M&A) and banking sector competition using panel regressions on bank-level data.
  - Extend panel regressions into a difference-in-differences (diff-in-diff) setting to illustrate the association of fintech emergence with changes in the competitive environment.
- Case studies focus on Brazil and Mexico, the two most developed fintech markets in the region.
- Analytical focus is for pre-pandemic periods; the paper does not address COVID-19 implications due to lack of data.

### Key stylized facts on LAC banking competition (section I highlights)
- LAC is characterized by limited financial depth and high interest spreads; countries in LAC are among those with the highest net interest margins (NIMs) in the world.
- Cross-country and within-country heterogeneity:
  - In Argentina, Brazil, Chile, and Peru, private domestic banks have higher NIMs than private foreign-owned banks.
  - In Bolivia, Colombia, and Mexico, foreign banks have slightly higher NIMs than domestic banks.
  - Public banks typically have substantively lower NIMs than private domestic banks, with Uruguay an exception where public banks have higher NIMs.
- Lerner index (markups) evidence:
  - Brazilian and Uruguayan banks have among the highest NIMs yet appear to have lower markups, suggesting high NIMs may be driven by high costs rather than market power.
  - Colombian banks have high markups despite moderate NIMs.
  - Peruvian banks have both high markups and high NIMs, consistent with low competition.
- Concentration:
  - Regional 3-bank concentration: Emerging and Developing Asia (EDA) 51.6 percent; Advanced Economies (AEs) 68.4 percent; LAC 63.5 percent.
  - Across regions, assets of the top three banks on average make up more than half of total banking sector assets; concentration fell compared to the previous decade, with LAC showing the smallest improvement.
- Policy relevance: high NIMs imply higher costs of financial intermediation and limited financial inclusion; improving competition to lower NIMs is linked to financial deepening and inclusion.

### Data and empirical framework (section II highlights)
- Bank-level data: historical balance sheet data for 692 banks from 1988 to 2018 from the Fitch Connect Database for: Argentina, Bolivia, Brazil, Chile, Colombia, Costa Rica, Mexico, Peru, and Uruguay.
- Ownership and M&A: Orbis, Zephyr, and banking institutions’ websites.
- Macroeconomic/macrofinancial: IMF World Economic Outlook and Financial Development Index databases.
- Definitions and measurements:
  - Foreign-owned banks: parent institutions with more than 50 percent foreign ownership.
  - Public banks: more than 50 percent government or subnational ownership.
  - NIM measured as (gross interest income + dividend income – total interest expense) as a percent of average assets.
  - Concentration measured mainly by the Herfindahl-Hirschman Index (HHI) and 3-bank concentration.

### Baseline regression findings (key quantitative results preserved)
- Overall finding: more competitive environment is associated with lower NIMs, but foreign bank entry and M&A activities may be associated with higher NIMs.
- Market concentration:
  - Lower NIMs are associated with less concentrated markets.
  - One standard deviation increase in the top three banks’ share is associated with over 110 basis points higher NIM.
- Foreign banks:
  - Foreign banks have lower NIMs on average: all else equal, foreign banks’ NIMs are around 100 basis points lower than those of domestic banks.
  - However, NIMs increase with the foreign bank’s market share.
- M&A:
  - M&A activities involving foreign banks are associated with higher NIMs; total M&As (local and foreign) are not.
  - The foreign-M&A association may reflect takeovers of less efficient domestic banks with higher NIMs; taking 1- or 2-year lags can eliminate the association.
- Bank-specific characteristics:
  - Liquidity: each percentage point increase in liquid to total assets is associated with a 2 basis points decrease in the NIM.
  - Non-performing loans (NPLs): higher NPLs associated with lower NIMs.
  - Equity ratios: banks with higher equity ratios see higher NIMs.
  - Non-interest (fee) income: banks with higher fee incomes have lower NIMs.
  - Administrative costs: a one-standard deviation increase in administrative costs is associated with 440 basis points increase in NIM.
  - Quantitatively, one standard deviation changes in bank-specific characteristics are associated with 40-70 basis points changes in NIMs (except administrative costs noted above).
- Macroeconomic influences:
  - Higher NIMs are associated with stronger real GDP growth and higher real interest rates.
  - NIMs decrease with total credit in the economy and EMBI spreads.
- Structural change post-GFC:
  - Margin differentials between domestic and foreign banks have more than halved since 2008.
  - Margin differentials between private and public banks have also diminished.
- Robustness with markups (Lerner index):
  - Higher markups associated with higher concentration (with lower statistical significance than NIM regressions).
  - Foreign banks on average have lower markups.
  - Public banks have higher markups on average in Lerner regressions even if NIM differences are not statistically different.

### Regressions incorporating fintech (diff-in-diff exploratory exercise)
- Fintech lending volumes remain very small: fintech lending in countries with data makes up only about 0.02 to 0.05 percent of total loans.
- Fintech country sample for lending (per Cambridge Center for Alternative Finance) used to define treated units.
- Diff-in-diff specification uses:
  - D_{t>2012} = 1 for time periods after 2012 (when fintech activity started to take place in the region);
  - D_{Fintech} = 1 for countries with fintech lending activity (Figure 7); interaction term is the coefficient of interest.
  - Alternate specification replaces D_{Fintech} with fintech credit share (fintech “intensity”) — volume of fintech credit to total credit in the economy.
- Key exploratory result:
  - Difference-in-differences coefficient has the expected negative sign, suggesting banks in fintech countries have seen greater reductions in NIMs than those in non-fintech countries in periods after fintech lending emerged.
  - Unobserved time-invariant differences between fintech and non-fintech countries are associated with higher baseline NIMs in fintech countries (positive and significant β2), consistent with fintech developing where intermediation prices are higher.
- Caveats:
  - Results are exploratory given sparse fintech quantitative data and the very small overall fintech volumes; findings should be interpreted with caution.

### Case studies: Brazil and Mexico (section III highlights)
Common threads across both countries:
- Fintech and bigtech are progressing more rapidly in payments than in lending.
- Many fintech startups focus on niche markets and the unbanked, enhancing financial inclusion.
- Incumbent banks respond by investing in innovation, improving efficiency and customer service, and partnering with fintechs.
- Partnerships can improve efficiency but effects on competition are ambiguous—incumbents may increase markups rather than lowering fees or rates.
- Encouraging signs that fees are declining in the payments segment in both countries.

Brazil:
- Banking sector assets around 105 percent of GDP; the five largest banks account for over 80 percent of total assets.
- Share of adult population with financial accounts: 86.5 percent (Central Bank of Brazil, 2018).
- Brazil had more than 650 fintech companies operating in the country at end-2019.
- Payments impact:
  - Merchant discount rates (MDRs) fell from nearly 3 to around 2½ percent after entrance of acquirers including fintechs and bigtechs.
- Fintech lending:
  - Fintech lenders’ credit growth outpaced largest banks but remain small in size compared with incumbents.
- Regulatory approach:
  - Regulators have accommodated fintech within existing frameworks using segmentation and proportionality.
  - Notable regulations and initiatives:
    - Payments: BCB Circular 3682 (November 2013), BCB Circular 3885 (March 2018), and BCB Resolution 80 (March 2021) revoking Circular 3885; four types of services allowed for payment institutions (electronic money issuer; post-paid payment instrument issuer; acquirer; payment transaction initiator (PISP)).
    - Securities: Regulation 588 (July 2017, CVM) crowdfunding registration exception.
    - Credit: CMN Resolution 4656 (April 2018) created SCD (Sociedade de Crédito Direto) and SEP (Sociedade Entre Pessoas) for digital credit origination; since 2019, 24 SCDs and 6 SEPs licensed.
    - Open banking: announced May 2020; implementation started February 2021 and will be fully operational by end-2021.
    - Regulatory sandboxes: joint statement June 2019; Susep, CVM, BCB have sandbox rules and have selected or expect to select projects (Susep authorized 11 projects in first cycle; CVM 7; BCB between 10 and 15).
  - Central Bank initiatives: Agenda BC+ (2016) and Agenda BC# (updated agenda promoting inclusion, competitiveness, transparency, financial education, and sustainability); Financial and Technology Innovation Laboratory (LIFT) launched 2018.
- Incumbent responses:
  - Digitization and branch reduction; incumbents launching digital banks (e.g., Banco Bradesco’s Next), partnering on digital products, or converting to digital banks (Banco Original, Banco Inter).
  - Banks funding fintech startups and creating innovation hubs (Cubo Itaú, InovaBra, Banco do Brasil’s fintech lab).

Mexico:
- Banking sector assets around 40 percent of GDP; bank credit as share of GDP is about one-quarter of levels in other emerging markets.
- Financial inclusion: 47 percent of adult population with financial accounts; about 70 percent of individual borrowers rely on informal credit.
- Fintech counts and growth:
  - Number of fintech startups reached 441 by March 2020 (Finnovista); average annual growth rate since 2016 of 23 percent.
  - By 2018 fintech credit share of total formal credit was 0.06 percent.
  - Most fintechs engage in payments and transfers (88 firms) and consumer and business lending (83 firms).
- Bigtechs and partnerships: Amazon introduced Amazon Cash (2017) and Amazon Rechargeable (2018) in partnership with Banorte; Konfio partnered with Facebook and PayPal.
- Regulatory response:
  - Fintech Law enacted March 2018 — first comprehensive fintech legislative framework in Latin America.
  - Law’s guiding six principles: (i) promote financial inclusion and innovation, (ii) foster competition, (iii) protect consumers, (iv) strengthen AML/CFT, (v) maintain financial stability, (vi) observe technological neutrality.
  - Recognizes and regulates: crowdfunding (including P2P), e-money platforms (FTIs); addresses virtual assets, APIs (open banking), and regulatory sandboxes.
  - Regulators: Banxico, CNBV, and SHCP (with other agencies such as CONDUSEF, CONSAR, CNSF involved in related roles).
  - Secondary regulations (September 2018) required FTIs to apply for licenses by September 2019; requirements included:
    - Minimum capital stock of 500,000 UDIs (approx. US$146,000) or 700,000 UDIs (approx. US$204,000) depending on activity.
    - Maximum limits on cash transfers per client and per day for electronic payment and crowdfunding platforms.
    - Maximum limits on lending activities for crowdfunding and P2P platforms.
    - Investor diversification obligations for crowdfunding platforms.
    - Reporting requirements for acquisitions of 10 percent or more of FTI shareholding.
  - Implementation challenges:
    - Only 85 out of 124 fintech entities that require a license filed applications by the September 2019 deadline.
    - First license issued January 2020 to NVIO Pagos Mexico (Bitso); authorization process interrupted by the Covid-19 pandemic.
    - Survey evidence: 52 percent of fintech startups view the Fintech Law as presenting high barriers to entry; an Endeavor Mexico study suggests 63 percent of startups could not absorb estimated compliance costs.
  - Supportive measures:
    - Banxico’s CoDi (Cobro Digital Systems) QR-code payment platform: five million users as of September 2020; participating banks include all traditional banks.
    - Regulatory sandbox and open banking promoted; CNBV’s “Sandbox Challenge” to encourage applicants.
- Incumbent responses and market dynamics:
  - Traditional banks invest in digital factories and collaborations (e.g., Santander’s Spotlight, BBVA’s Fast Track).
  - Collaboration and funding: Klar received US$57.5 million from Santander in 2019; BBVA Mexico acquired OpenPay (2016); Santander announced plan to acquire Elavon México (February 2020).
  - Finnovista projects fintech startups could take up to 30 percent of Mexico’s banking market over the next decade (forecast cited).

### Conclusions and policy implications (section IV highlights)
- Early evidence indicates fintech is associated with lower NIMs and is influencing incumbent banks’ behavior (fintech adoption, fee reductions in payments, digitalization).
- Whether fintech will reduce banking sector concentration remains unclear and depends on entry forms, incumbents’ responses (acquisitions, partnerships), and regulatory frameworks.
- Open banking could be a gamechanger by granting equal access to customer data across the financial system, potentially levelling the playing field for fintech firms.
- New challenges: data privacy, cybersecurity, AML/CFT, fraud, and cross-border transaction supervision.
- Regulators face a balancing act: promote innovation and competition, protect consumers, mitigate risks, and collaborate internationally to share best practices and prevent regulatory arbitrage.

*Source: INTRODUCTION, wpiea2021114-print-pdf*

### References

### References

### Key cited works
- Agur, I., S.m., Peria, and C. Rochon, 2020, “Digital Financial Services and the Pandemic: Opportunities and Risks for Emerging and Developing Economies,” IMF Special Series on COVID-19, July.
- Beck, T., Demirguc-Kunt, A., and Levine, R., 2003 "Bank Concentration and Crises," NBER Working Papers 9921, National Bureau of Economic Research.
- Berkmen, P., Beaton, K., Gershenson, D., Arze del Granado, J., Ishi, K., Kim, M., Kopp, E., and Rousset, M., 2019, “Fintech in Latin America and the Caribbean: Stocktaking,” IMF Working Paper 19/71.
- Claessens, S., Frost, J., Turner, G., and Zhu, F., 2018, “Fintech Credit Markets Around the World: Size, Drivers and Policy Issues,” BIS Quarterly Review September, 2018.
- Demirgüç-Kunt, Asli, Leora Klapper, Dorothe Singer, Saniya Ansar, and Jake Hess (2018), “The Global Findex Database 2017: Measuring Financial Inclusion and the Fintech Revolution,” World Bank.
- Frost, J., Gambacorta, L., Huang, Y., Shin, H. S., and Zbinden, P., 2019, “BigTech and the Changing Structure of Financial Intermediation” BIS Working Paper 779.
- Igan, D., Martinez Peria, M. S., Pierri, N., and Presbitero, A., 2020, “When They Go Low, We Go High? Bank Market Power and Interest Rates”, IMF Working Paper.
- World Bank, 2016, “Technical Note: SME Finance”, Mexico: Financial Sector Assessment Program.
- Yildirim, H.S. and Philippatos, G.C., “Restructuring, Consolidation and Competition in Latin American Banking Markets,” Journal of Banking & Finance 31(3), March 2007, pp. 629-39.
- (Additional citations appear in the source document; the above are extracted verbatim from the References list.)

### Table 1. Summary Statistics (variables with Source, Observations, Mean, S.D.)
- Net interest margin — Fitch Connect and authors' calculations — 811 — 86.55 — 5.71
- Net fees — Fitch Connect and authors' calculations — 769 — 01.86 — 2.17
- Liquid assets — Fitch Connect and authors' calculations — 886 — 724.93 — 19.37
- Operating cost — Fitch Connect and authors' calculations — 812 — 78.23 — 6.5
- Total equity — Fitch Connect and authors' calculations — 899 — 317.94 — 16.28
- Non-performing loans — Fitch Connect and authors' calculations — 611 — 07.09 — 10.05
- Market share — Fitch Connect and authors' calculations — 877 — 82.87 — 6.45
- Foreign share — Fitch Connect and authors' calculations — 9132 — 24.68 — 15.48
- HHI — Fitch Connect and authors' calculations — 9140 — 11.86 — 6.86
- 3-bank concentration — Fitch Connect and authors' calculations — 9140 — 48.71 — 1.49
- Gross loans to GDP — Fitch Connect and authors' calculations — 9140 — 32.14 — 17.66
- Real GDP growth — IMF's WEO database — 9140 — 2.99 — 3.64
- Inflation — IMF's WEO database — 8447 — 43.22 — 58.82
- Real interest rate — IMF's IFS database and authors' calculations — 7752 — 54.49 — 452.98
- EMBI spreads — Thomson Reuters Datastream — 775 — 96.02 — 0.84
- Financial markets access index — IMF's Financial Development Index — 8823 — 0.42 — 0.2
- Lerner index — Fitch Connect and Igan et al. (2020) — 6039 — 0.36 — 0.23

### Table 2. Baseline Regression Results — Selected coefficients for NIM (columns labeled (1)–(4))
- Liquid Assets: -0.0229*** (standard errors: (0.00728), (0.00721), (0.00729), (0.00728))
- Net Fees: -0.212** (0.0983), -0.209** (0.0981), -0.212** (0.0982), -0.213** (0.0980)
- Operating Costs: 0.680*** (0.0422), 0.680*** (0.0422), 0.681*** (0.0422), 0.682*** (0.0421)
- NPL: -0.0517*** (0.0150), -0.0523*** (0.0149), -0.0522*** (0.0150), -0.0532*** (0.0150)
- Equity: 0.0427*** (0.00949), 0.0420*** (0.00941), 0.0427*** (0.00948), 0.0427*** (0.00948)
- Foreign Dummy: -1.074*** (0.267), -1.116*** (0.263), -1.073*** (0.267), -1.078*** (0.266)
- Foreign Dummy x Market Share: 0.128*** (0.0338), 0.128*** (0.0340), 0.128*** (0.0338), 0.127*** (0.0336)
- Real Growth: 0.173*** (0.0378), 0.188*** (0.0376), 0.172*** (0.0376), 0.171*** (0.0377)
- Total Loans/GDP: -0.0269** (0.0120), -0.0419*** (0.0120), -0.0269** (0.0120), -0.0263** (0.0120)
- Real Interest Rate: 0.0750*** (0.0186), 0.0836*** (0.0183), 0.0746*** (0.0186), 0.0747*** (0.0186)
- EMBI Spread: -0.488** (0.209), -0.547*** (0.209), -0.486** (0.209), -0.473** (0.209)
- HHI: 0.149** (0.0577), 0.148** (0.0578), 0.147** (0.0578)
- 3-bank concentration (included in one specification): 0.101*** (0.0186)
- Foreign M&A Dummy (included in one specification): 1.522** (0.721)
- Constants reported across specifications: 3.859** (1.948), 1.561 (1.999), 3.851** (1.950), 3.827** (1.946)
- Observations: 4,085 (across columns)
- R-squared: 0.619, 0.623, 0.619, 0.620
- Country, Year FE: Yes
- N_clust: 507
- Note: Robust standard errors in parentheses. Significance: *** p<0.01, ** p<0.05, * p<0.1.

### Table 3. Baseline Regression Results: Structural Shifts After the GFC — Selected coefficients (Before GFC / After GFC)
- Liquid Assets:
  - Before GFC: -0.0164* (0.00898), -0.0163* (0.00896), -0.0166* (0.00901), -0.0166* (0.00899)
  - After GFC: -0.0363*** (0.0101), -0.0364*** (0.0101), -0.0363*** (0.0101), -0.0364*** (0.0101)
- Net Fees:
  - Before GFC: -0.253** (0.112), -0.249** (0.112), -0.252** (0.112), -0.254** (0.113)
  - After GFC: -0.231* (0.139), -0.232* (0.139), -0.231* (0.140), -0.232* (0.139)
- Operating Costs:
  - Before GFC: 0.635*** (0.0451) ... 0.636*** (0.0451)
  - After GFC: 0.712*** (0.0562) ... 0.714*** (0.0563)
- NPL:
  - Before GFC: -0.0530*** (0.0172) ... -0.0538*** (0.0172)
  - After GFC: -0.0499** (0.0249) ... -0.0532** (0.0248)
- Equity:
  - Before GFC: 0.0627*** (0.0132) ... 0.0625*** (0.0132)
  - After GFC: 0.0296*** (0.0112) ... 0.0302*** (0.0112)
- Foreign Dummy:
  - Before GFC: -1.703*** (0.314) ... -1.702*** (0.314)
  - After GFC: -0.821** (0.345) ... -0.830** (0.344)
- Foreign Dummy x Market Share:
  - Before GFC: 0.211*** (0.0441) ... 0.208*** (0.0441)
  - After GFC: 0.0921** (0.0409) ... 0.0938** (0.0410)
- Foreign Market Share:
  - Before GFC: -0.0408** (0.0163) ... -0.0414** (0.0162)
  - After GFC: -0.00880 (0.0370) ... -0.00299 (0.0374)
- Observations:
  - Before GFC: 1,679 (R-squared: 0.625–0.627 across columns)
  - After GFC: 2,032 (R-squared: 0.652–0.653 across columns)
- Country, Year FE and clustering reported; significance levels as above.

### Table 5. Difference-in-Differences — Selected coefficients for NIM (columns (1) and (2))
- Liquid Assets: -0.0112 (0.00684) ; -0.0139** (0.00656)
- Net Fees: -0.244** (0.100) ; -0.253** (0.101)
- Operating Costs: 0.694*** (0.0429) ; 0.693*** (0.0434)
- NPL: -0.0471*** (0.0150) ; -0.0468*** (0.0151)
- Equity: 0.0405*** (0.0106) ; 0.0414*** (0.0105)
- Foreign Dummy: -0.737*** (0.251) ; -0.840*** (0.253)
- Real Growth: 0.128*** (0.0236) ; 0.113*** (0.0246)
- Inflation: -0.0482*** (0.0141) ; -0.0591*** (0.0148)
- Total Loans/GDP: -0.0217*** (0.00804) ; -0.0292*** (0.00804)
- Real Interest Rate: 0.0916*** (0.0180) ; 0.0920*** (0.0187)
- EMBI spread: -0.342** (0.143) ; -0.419*** (0.146)
- Foreign Market Share: -0.0406*** (0.0109) ; -0.0532*** (0.0107)
- Financial Market Access Index: -3.309*** (0.942) ; -1.399** (0.705)
- HHI: 0.208*** (0.0411) ; 0.220*** (0.0403)
- Fintech Period Dummy: 1.437** (0.592) ; 0.314* (0.184)
- Fintech Country Dummy: (included in column (2)) 2.343*** (0.741)
- Fintech Period Dummy x Fintech Country Dummy: -1.194* (0.619)
- Observations: 4,085 (both columns)
- R-squared: 0.579 ; 0.575
- N_clust: 507
- Note: Robust standard errors in parentheses. Significance: *** p<0.01, ** p<0.05, * p<0.1.

*Content extracted verbatim from "wpiea2021114-print-pdf - References."*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2021/english/wpiea2021114-print-pdf.pdf_