## wpiea2024017-print-pdf - Annex Table 1. Variable Names, Definition and Sources

## Source details

**Canonical URL:** [wpiea2024017-print-pdf - Annex Table 1. Variable Names, Definition and Sources](https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024017-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2024/english/wpiea2024017-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2024/english/wpiea2024017-print-pdf.pdf.json)

---

### I. Study scope and main baseline finding
- FinTech current share in major markets: "around 2 percent of total credit".
- Average growth rate of FinTech volumes over 2012-2020, excluding China: "70 percent" (CCAF, 2021).
- Empirical scope: cross-country database of "over 10,000 traditional FIs" across "57 countries" over the "2012-2020" period.
- Main baseline result: "greater FinTech presence is associated with heightened risk taking" (supports competition-fragility); robust to alternative FinTech measures, extra controls, and estimation methods including Two-Stage Least Squares (2SLS) and Generalized Methods of Moments (GMM).

### II. Econometric approach and identification
- Dependent variable:
  - z-score: Log of ((Return on Assets + Equity-to-Assets)/s.d. Return on Assets).
- Core regressors:
  - Main regressor: natural logarithm of country-level FinTech transactions (FinTech_c,t); alternatives: FinTech scaled by GDP and FinBigTech.
- Baseline panel specification (simplified):
  - Z_b,c,t = β * FinTech_c,t + γ * X_b,c,t−1 + δ * W_c,t + Other_b,c,t
- Channels analyzed:
  - Decomposition of log(z-score) into risk-adjusted profitability (ROA/σ(ROA)) and risk-adjusted capitalization ((E/A)/σ(ROA)).
- Identification and endogeneity strategies:
  - Instruments in 2SLS: (1) mobile cellular subscriptions; (2) total volume of FinTech transactions in all countries except the one under consideration (FinTech−c,t).
  - Additional robustness: GMM estimation (Arellano-Bover / Blundell-Bond).
- Interaction specification for heterogeneity:
  - Z_b,c,t = β1 * FinTech_c,t + β2 * FinTech_c,t * CHAR_b,c,t−1 + γ * X_b,c,t−1 + δ * W_c,t + Other_b,c,t
  - CHAR_b,c,t−1 equals unity if indicator above sample median.

### III. Data sources and sample moments
- FinTech transactions: CCAF Global Alternative Finance repository, annual data 2012-2020 for "57 countries".
- FI balance sheet and income data: Bureau van Dijk Orbis database; "over 10,000" financial institutions.
- Country-level controls: IMF WEO, World Bank Governance Indicators, Haver, Global Financial Development Database, Romelli (2022), Anginer and others (2019).
- Sample composition and means (Table 1 highlights):
  - Observations: total FIs "10,167" (Banks "9,198", Non-Banks "969").
  - Average Total Assets (in billions, US$): All "2.8"; Banks "2.0"; Commercial Bank "10.2"; Cooperative Bank "0.4"; Non-Banks "10.7".
  - Average Equity-to-Asset ratio (percent): All "16.2"; Banks "15.3"; Commercial Bank "17.7"; Cooperative Bank "14.9"; Non-Banks "25.4".
  - Average ROA (percent): All "4.9"; Banks "4.7"; Commercial Bank "7.1"; Cooperative Bank "4.2"; Non-Banks "7.5".
  - Average z-score (log units): All "0.8"; Banks "0.7"; Commercial Bank "1.0"; Cooperative Bank "0.7"; Non-Banks "1.7".
- Summary statistics for main regression variables (Table 2):
  - z-score (log units): Observations "74,052"; Mean "4.2"; Median "4.2"; Standard Deviation "1.3"; Minimum "0.9"; Maximum "7.8".
  - FinTech (in billions, US$): Observations "87,384"; Mean "22.4"; Median "4.4"; Standard Deviation "33.3"; Minimum "0.0"; Maximum "356.8".
  - FinTech (log units): Observations "87,384"; Mean "20.9"; Median "22.2"; Standard Deviation "4.1"; Minimum "7.2"; Maximum "26.6".
  - Bank-controls examples: Total assets (log units) Observations "84,897"; Mean "11.6"; Equity-to-Asset ratio (percent) Observations "84,831"; Mean "16.2".
  - Macro-controls examples: GDP growth (percent) Observations "91,503"; Mean "1.5"; Inflation (percent) Observations "91,335"; Mean "2.4"; Rule of Law (index) Observations "91,503"; Mean "79.4".

### IV. Key empirical estimates (selected results preserved exactly)
- Baseline aggregate effect (Table 1 / Table 5):
  - FinTech (log units): -0.0139** (0.00639) — dependent variable: z-score (log units).
  - P2P Lending (log units): -0.0105** (0.00524).
  - Balance Sheet Lending (log units): -0.00121 (0.00578).
  - Other FinTech (log units): -0.0281*** (0.00739).
  - Size (log assets): 0.246*** (0.0315).
  - Equity-to-assets: 0.00980*** (0.00149).
  - GDP growth: -0.0110** (0.00539).
  - Inflation: -0.00735** (0.00357).
  - GDP per capita: 1.311*** (0.277).
  - N: 70,578; rho: 0.64.
- IV and GMM robustness (reported in main text/Table 4):
  - 2SLS estimate (instrument: FinTech−c,t) coefficient "-0.480" (Table 4, column 9) — interpretation: "a one percentage point increase in (log) FinTech transactions is associated with a decrease in the z-score of 0.5".
  - GMM estimate (Table 4 column 10): FinTech coefficient "-0.0752***" (0.0118).
- Decomposition into channels (Table 6 standardized results):
  - FinTech (standardized) on risk-adjusted ROA: -0.0604*** (0.0212) in column (1).
  - Balance Sheet Lending (standardized) on risk-adjusted ROA: -0.0866*** (–0.0132) in column (5).
  - P2P Lending (standardized): -0.0590*** (–0.0156) in column (3).
- Heterogeneity by FI type (Table 1–5 highlights):
  - Across Banks (Table 2): FinTech (log units): -0.0108 (0.00704); Other FinTech: -0.0305*** (0.00828); N: 64,957; rho: 0.66.
  - Commercial Banks (Table 3): Balance Sheet Lending (log units): -0.0301*** (0.0101); Other FinTech: -0.0233** (0.0101); N: 8,370; rho: 0.77.
  - Cooperative Banks (Table 4): P2P Lending (log units): -0.0436*** (0.0147); N: 52,593; rho: 0.69.
  - Non-Banks (Table 5): FinTech (log units): -0.0420** (0.0165); N: 5,621; rho: 0.65.

### V. Heterogeneity by bank-, industry-, and country-specific characteristics (interaction results)
- Interactions indicating mitigation or reversal of adverse FinTech effect:
  - Fin*aboveMediancapital coefficient "0.0127*".
  - Fin*lessMedianloantodep coefficient "0.0101**".
  - Fin*aboveMediannonic coefficient "0.00861*".
  - Institutions: Fin*aboveMedianruleoflaw "0.0557***"; Fin*aboveMediancentralbankindependence "0.0322***".
- Policy-framework splits (Table 10 summary — FinTech effect by Below/Above median):
  - Supervisory Index:
    - Below median: -0.0227** (0.00930).
    - Above Median: 0.00313 (0.0101).
  - Regulatory Capital:
    - Below median: -0.0226** (0.00961).
    - Above Median: -0.00865 (0.0167).
  - Activity Restriction:
    - Below median: -0.0287** (0.0139).
    - Above Median: -0.0209*** (0.00785).
  - Multiple Supervisory Agencies:
    - Below median: -0.0227*** (0.00747).
    - Above Median: 0.0781*** (0.0231).
- Interpretation: stronger institutions, higher capital and liquidity ratios, and greater income diversification can attenuate or in some cases reverse the adverse association between FinTech presence and FI risk taking; regulatory and supervisory context matters for the sign and magnitude.

### VI. Variable definitions and data sources (Annex Table 1)
- Dependent variables:
  - z-score: Log of ((Return on Assets + Equity-to-Assets)/s.d. Return on Assets) — Authors calculations using Bureau van Dijk Orbis database.
  - Risk-adjusted ROA: Return on Assets/s.d. Return on Assets — Authors calculations using Bureau van Dijk Orbis database.
  - Risk-adjusted E/A: Equity to Assets/s.d. Return on Assets — Authors calculations using Bureau van Dijk Orbis database.
- Key explanatory variables:
  - FinTech: Log of (Total volume of digital lending and capital raising activities in US$) — Cambridge Center for Alternative Finance (2021).
  - P2P Lending: Log of (Total volume of P2P lending activities in US$) — Cambridge Center for Alternative Finance (2021).
  - Balance Sheet Lending: Log of (Total volume of Balance Sheet lending activities in US$) — Cambridge Center for Alternative Finance (2021).
- Bank-controls: Total assets (Log), Equity-to-Asset ratio (%), Non-Interest Income-to Average Assets (%), Total Capital Ratio (%), Net Loans to Total Deposits and Borrowing (%) — Bureau van Dijk Orbis.
- Macro-controls: GDP growth (%), GDP per capita, Inflation (%), Policy rate (%) — IMF WEO Database (2023), Haver Database, Global Financial Development Database (2022).
- Other robustness controls: Government Effectiveness (index), Financial Development (index), Investment Freedom (index), Financial Freedom (index), Macroprudential Policies — respective sources cited in Annex.
- Instrumental variables:
  - Mobile Subscriptions: Log of (Mobile cellular subscriptions (per 100 people)) — World Development Indicators (2023).
  - Sum of all FinTech transactions leaving out the Country under consideration: Log of (Total volume of FinTech transactions in all countries except the country under consideration) — Authors calculations using CCAF (2021).
- Institutions and policy framework indices: Central Bank Independence (Romelli, 2022); Rule of Law (Kaufmann and Kraay, 2023); Supervision, Regulatory capital requirements, Activity Restriction Index, Multiple Supervisors — Authors calculations using Anginer and others (2019) and other cited sources.

### VII. Policy implications (drawn directly from findings)
- Strengthen institutions (rule of law, contract enforcement) to limit translation of FinTech growth into higher FI risk taking.
- Ensure supervisory autonomy (central bank independence where relevant) and rigorous supervision to monitor FI risk-taking amid expanding FinTech activities.
- Calibrate regulatory frameworks to country-specific conditions and update legal foundations to encompass emerging FinTech models.
- Consider capital and liquidity buffers, and encourage income diversification, as tools to mitigate competitive pressures that may increase risk taking.
- Leverage multiple supervisory approaches where appropriate to provide complementary oversight and reduce potential regulatory arbitrage.

*Source: Annex Table 1. Variable Names, Definition and Sources — wpiea2024017-print-pdf (IMF Working Paper).*

### Annex Table 1. Variable Names, Definition and Sources ..................................................................

### wpiea2024017-print-pdf - Annex Table 1. Variable Names, Definition and Sources

### Annex listings and page references
- Annex Table 1. Variable Names, Definition and Sources ............................................................................ 27
- Annex Table 2. Descriptive Statistics ........................................................................................................... 28
- Annex Table 3. List of Countries included in the Sample. ............................................................................ 29
- Annex Table 4. Digital Lending and Capital Raising Activities Models and Definitions ................................ 30
- Annex Table 5. Detailed Regression Output Tables .................................................................................... 31

*Source: wpiea2024017-print-pdf - Annex Table 1. Variable Names, Definition and Sources*

### References .............................................................................................................

### References

### I. Introduction — scope and motivation
- FinTech finance current share in major markets: "around 2 percent of total credit".
- Average growth rate of FinTech volumes over 2012-2020, excluding China: "70 percent" (CCAF, 2021).
- Study objective: assess whether growing FinTech presence contributes to more risk taking among traditional financial institutions (competition-fragility) or less (competition-stability).
- Empirical scope: cross-country database of "over 10,000 traditional FIs" and FinTech activities across "57 countries" over the "2012-2020" period.
- Main baseline result: "greater FinTech presence is associated with heightened risk taking", supporting the competition-fragility hypothesis; result robust to alternative FinTech measures, extra controls, and estimation methods including Two-Stage Least Squares (2SLS) and Generalized Methods of Moments (GMM).

### II. Literature — competing hypotheses and prior findings
- Competition-fragility view: increased competition erodes franchise value and encourages excessive risk taking (e.g., Allen and Gale, 2004; Marcus, 1984; Keeley, 1990).
- Competition-stability view: competition lowers lending rates, reduces borrower credit risk, and can enhance stability (Boyd and De Nicoló, 2005).
- Mixed FinTech-related evidence:
  - Findings aligned with competition-fragility: Fang and others (2023); Ben Naceur and others (2023); Bakker and others (2023).
  - Findings aligned with competition-stability: Grennan and Michaely (2021); Deng et al. (2021); FSB (2017); Fung and others (2020).
- Contribution: global cross-country analysis including NBFIs and banks, differentiating FinTech business models and FI types, and investigating profitability and capitalization channels.

### III. Econometric approach — dependent variable, core regressors, and identification
- Dependent variable: z-score (log of z-score), where
  - z-score = ROA + (E/A) / σ(ROA); log(z-score) used because raw z-score is highly skewed.
- Main regressor: natural logarithm of country-level FinTech transactions (FinTech_c,t); alternative measures include FinTech scaled by GDP and combined FinBigTech.
- Baseline panel specification (simplified representation):  
  Z_b,c,t = β * FinTech_c,t + γ * X_b,c,t−1 + δ * W_c,t + Other_b,c,t
- Channels: decomposition of log(z-score) into risk-adjusted profitability (ROA/σ(ROA)) and risk-adjusted capitalization ((E/A)/σ(ROA)).
- Identification and endogeneity:
  - Instruments used in 2SLS: (1) mobile cellular subscriptions; (2) total volume of FinTech transactions in all countries except the one under consideration (FinTech−c,t).
  - Additional robustness: GMM estimation (Arellano-Bover / Blundell-Bond) to address unobserved heterogeneity and limited time periods.
- Interaction specification to assess heterogeneity:
  Z_b,c,t = β1 * FinTech_c,t + β2 * FinTech_c,t * CHAR_b,c,t−1 + γ * X_b,c,t−1 + δ * W_c,t + Other_b,c,t
  - CHAR_b,c,t−1 equals unity if indicator above sample median.

### IV. Data and summary statistics — datasets and sample moments
- Data sources:
  - FinTech transactions: CCAF Global Alternative Finance repository, annual data 2012-2020 for "57 countries".
  - FI balance sheet and income data: Bureau van Dijk Orbis database; "over 10,000" financial institutions.
  - Country-level macro and structural controls: IMF WEO, World Bank Governance Indicators, Haver, Global Financial Development Database, Romelli (2022), Anginer and others (2019).
- Sample composition and means (Table 1 highlights):
  - Observations: total FIs "10,167" (Banks "9,198", Non-Banks "969").
  - Average Total Assets (in billions, US$): All "2.8"; Banks "2.0"; Commercial Bank "10.2"; Cooperative Bank "0.4"; Non-Banks "10.7".
  - Average Equity-to-Asset ratio (percent): All "16.2"; Banks "15.3"; Commercial Bank "17.7"; Cooperative Bank "14.9"; Non-Banks "25.4".
  - Average ROA (percent): All "4.9"; Banks "4.7"; Commercial Bank "7.1"; Cooperative Bank "4.2"; Non-Banks "7.5".
  - Average z-score (log units): All "0.8"; Banks "0.7"; Commercial Bank "1.0"; Cooperative Bank "0.7"; Non-Banks "1.7".
- Summary statistics for main regression variables (Table 2):
  - z-score (log units): Number of Observations "74,052"; Mean "4.2"; Median "4.2"; Standard Deviation "1.3"; Minimum "0.9"; Maximum "7.8".
  - FinTech (in billions, US$): Observations "87,384"; Mean "22.4"; Median "4.4"; Standard Deviation "33.3"; Minimum "0.0"; Maximum "356.8".
  - FinTech (log units): Observations "87,384"; Mean "20.9"; Median "22.2"; Standard Deviation "4.1"; Minimum "7.2"; Maximum "26.6".
  - Bank-controls examples: Total assets (log units) Observations "84,897"; Mean "11.6"; Equity-to-Asset ratio (percent) Observations "84,831"; Mean "16.2".
  - Macro-controls examples: GDP growth (percent) Observations "91,503"; Mean "1.5"; Inflation (percent) Observations "91,335"; Mean "2.4"; Rule of Law (index) Observations "91,503"; Mean "79.4".

### V. Empirical results — main findings and heterogeneity
- Baseline finding:
  - FinTech (log units) coefficient negative and statistically significant in baseline (Table 3, column 1): coefficient "-0.0139" with standard error "(0.00639)" and significance "**".
  - Interpretation example from IV: 2SLS estimate (instrument: FinTech−c,t) coefficient "-0.480" (Table 4, column 9) — "a one percentage point increase in (log) FinTech transactions is associated with a decrease in the z-score of 0.5", with median and standard deviation of z-score "4.2" and "1.3", respectively.
- Robustness:
  - Negative relationship persists when using FinTech-to-GDP ("-0.0882***", Table 3 column 2), FinBigTech ("-0.0150***", Table 3 column 3), and when adding controls for revenue mix, provisioning, loan-to-assets.
  - 2SLS and GMM specifications (Table 4) reaffirm negative FinTech effect; GMM estimate in Table 4 column 10: FinTech coefficient "-0.0752***" (0.0118).
- Decomposition of z-score (Table 5):
  - FinTech (log units) negative correlation with risk-adjusted ROA and risk-adjusted E/A; standardized FinTech coefficient on risk-adjusted ROA: "-0.0604***".
- FinTech business models and FI types (Table 6 summary):
  - Aggregate effect: FinTech (log units) "-0.0139**"; P2P "-0.0105**"; Balance Sheet (B/S) "-0.00121"; Other "-0.0281***".
  - Commercial banks: B/S model adverse effect significant (commercial banks: B/S coefficient "-0.0301***" in detailed tables); cooperative banks: P2P model adverse effect significant (cooperative banks: P2P coefficient "-0.0436***").
  - Decomposition: Balance sheet models negatively associated with risk-adjusted profits across FIs (aggregate ROA coefficient "-0.0866***"); cooperative banks show particularly large negative effects in ROA for B/S ("-0.251***" standardized).
- Heterogeneity by bank-, industry-, and country-specific characteristics (Table 7 highlights):
  - Interaction results indicate that higher capital and liquidity ratios and greater income diversification attenuate the impact of FinTech on risk taking:
    - Fin*aboveMediancapital coefficient "0.0127*" (reduces negative FinTech impact).
    - Fin*lessMedianloantodep coefficient "0.0101**".
    - Fin*aboveMediannonic coefficient "0.00861*".
  - Institutions: Fin*aboveMedianruleoflaw "0.0557***"; Fin*aboveMediancentralbankindependence "0.0322***" — stronger institutions can reverse or reduce the adverse effect.
- Policy framework heterogeneity (Table 8 summary):
  - Splitting sample by policy indicators yields mixed outcomes:
    - Supervisory Index: βFinTech (Below median) "-0.0227**"; (Above median) "0.00313" (not significant).
    - Regulatory Capital: βFinTech (Below median) "-0.0226**"; (Above median) "-0.00865" (not significant).
    - Activity Restriction: βFinTech (Below median) "-0.0287**"; (Above median) "-0.0209***".
    - Multiple Supervisory Agencies: βFinTech (Below median) "-0.0227***"; (Above median) "0.0781***".
  - Interpretation: in some regulatory/supervisory environments, greater FinTech is associated with less risk taking; in others, especially where activity restrictions are stringent, FinTech expansion associates with greater risk taking.

### VI. Conclusions and policy implications
- Core conclusion: robust evidence that "greater FinTech presence is associated with heightened risk taking by FIs", supporting the competition-fragility hypothesis in the aggregate sample.
- Business-model-specific conclusion: commercial banks are most adversely affected by FinTech balance sheet lending; cooperative banks are disproportionately influenced by P2P FinTechs. Deterioration in profitability is a primary channel.
- Conditional findings: higher capitalization, higher liquidity, and greater non-interest income diversification mitigate FinTech’s adverse effect on risk taking; stronger institutions (rule of law, central bank independence) can in some cases flip the sign.
- Policy implications (bullet recommendations drawn from findings):
  - Strengthen institutions (rule of law, contract enforcement) to help ensure FinTech growth does not translate into higher FI risk taking.
  - Ensure supervisory autonomy (central bank independence where relevant) and rigorous supervision to monitor FI risk-taking behavior amid expanding FinTech activities.
  - Calibrate regulatory frameworks to country-specific conditions and update legal foundations to encompass emerging FinTech models.
  - Consider capital and liquidity buffers, and encourage income diversification, as tools to mitigate competitive pressures that may increase risk taking.
  - Leverage multiple supervisory approaches where appropriate to provide complementary oversight and reduce potential regulatory arbitrage.

*Source: IMF Working Paper — “Does FinTech Increase Bank Risk Taking?” (content unit: wpiea2024017-print-pdf - References).*

### Annex Table 1. Variable Names, Definition and Sources

### Annex Table 1. Variable Names, Definition and Sources

### Dependent variables
- z-score: Log of ((Return on Assets + Equity-to-Assets)/s.d. Return on Assets) — Authors calculations using Bureau van Dijk Orbis database
- Risk-adjusted ROA: Return on Assets/s.d. Return on Assets — Authors calculations using Bureau van Dijk Orbis database
- Risk-adjusted E/A: Equity to Assets/s.d. Return on Assets — Authors calculations using Bureau van Dijk Orbis database

### Explanatory variables
- FinTech: Log of (Total volume of digital lending and capital raising activities in US$) — Cambridge Center for Alternative Finance (2021)
- P2P Lending: Log of (Total volume of P2P lending activities in US$) — Cambridge Center for Alternative Finance (2021)
- Balance Sheet Lending: Log of (Total volume of Balance Sheet lending activities in US$) — Cambridge Center for Alternative Finance (2021)

### Bank-controls
- Total assets: Log of (Total Assets) — Bureau van Dijk Orbis
- Equity-to-Asset ratio: Equity to Total Assets (%) — Bureau van Dijk Orbis
- Non-Interest Income-to Average Assets: Non-Interest Income-to Average Assets (%) — Bureau van Dijk Orbis
- Total Capital Ratio: Total Capital Ratio (%) — Bureau van Dijk Orbis
- Net Loans to Total Deposits and Borrowing: Net Loans to Total Deposits and Borrowing (%) — Bureau van Dijk Orbis

### Macro-controls
- GDP growth: GDP, at constant prices, percent change (%) — IMF WEO Database (2023)
- GDP per capita: Gross domestic product per capita, constant prices — IMF WEO Database (2023)
- Inflation: Annual percentage of average consumer prices (%) — IMF WEO Database (2023)
- Policy rate: Central Bank Policy rate (%) — Haver Database
- Bank concentration: Assets of five largest banks to total bank assets (%) — Global Financial Development Database (2022)

### Other Robustness Controls
- Government Effectiveness (index): Perceptions of the quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of policy formulation and implementation, and the credibility of the government's commitment to such policies. — Kaufmann and Kraay (2023)
- Financial Development (index): A relative ranking of countries on the depth, access, and efficiency of their financial institutions and financial markets. — Sahay and others (2023)
- Investment Freedom (index): National treatment of foreign investment, Foreign investment code, Restrictions on land ownership, Sectoral investment restrictions, Expropriation of investments without fair compensation, Foreign exchange controls, Capital controls. — Heritage Foundation (2023)
- Financial Freedom (index): The extent of government regulation of financial services, The degree of state intervention in banks and other financial firms through direct and indirect ownership, Government influence on the allocation of credit, The extent of financial and capital market development, and Openness to foreign competition. — Heritage Foundation (2023)
- Macroprudential Policies: The sum of Macorprudential policy action indicators - Each tightening event is coded as +1, each loosening event is coded as -1 and no or neutral action is coded as a zero. — Alam and others (2019)

### Instrumental Variables
- Mobile Subscriptions: Log of (Mobile cellular subscriptions (per 100 people)) — World Development Indicators (2023)
- Sum of all FinTech transactions leaving out the Country under consideration: Log of (Total volume of FinTech transactions in all countries except the country under consideration). — Authors calculations using CCAF (2021)

### Institutions
- Central Bank Independence (index): A comprehensive index that captures a number central bank characteristics: governor and central bank board; monetary policy and conflicts resolution; objectives; limitations on lending to the government; financial independence; reporting and disclosure. — Romelli (2022)
- Rule of Law (index): Perceptions on the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence. — Kaufmann and Kraay (2023)

### Policy frameworks
- Supervision (index): Quality of supervision based on a number of questions: 12.1, 12.5, 12.10, 12.11, 12.12.2, 12.13, 12.14, 12.20, 12.27. — Authors calculations using Anginer and others (2019)
- Regulatory capital requirements: Minimum required risk-based regulatory capital ratio (%) — Anginer and others (2019)
- Activity Restriction Index (index): Conditions under which banks can engage in securities activities, insurance and real estate activities, along with engaging in nonfinancial businesses. — Authors calculations using Anginer and others (2019)
- Multiple Supervisors (index): Single or multiple body/agencies supervising banks for prudential purposes — Authors calculations using Anginer and others (2019)

*Source: Annex Table 1. Variable Names, Definition and Sources (IMF working paper).*

### Annex Table 5. Detailed Regression Output Tables

### Annex Table 5. Detailed Regression Output Tables

### Table 1: Effect of FinTech Business Models and Risk Taking Across All FIs
- Dependent variable: z-score (log units)
- FinTech (log units): -0.0139** (0.00639)
- P2P Lending (log units): -0.0105** (0.00524)
- Balance Sheet Lending (log units): -0.00121 (0.00578)
- Other FinTech (log units): -0.0281*** (0.00739)
- Size (log assets): 0.246*** (0.0315)
- Equity-to-assets: 0.00980*** (0.00149)
- GDP growth: -0.0110** (0.00539)
- Inflation: -0.00735** (0.00357)
- GDP per capita: 1.311*** (0.277)
- Policy rate: -0.00365 (0.00384)
- Concentration: -0.000829 (0.00122)
- N: 70,578
- rho: 0.64
- Bank Fixed Effects: Yes; Time Fixed Effects: Yes

### Table 2: Effect of FinTech Business Models and Risk Taking Across Banks
- Dependent variable: z-score (log units)
- FinTech (log units): -0.0108 (0.00704)
- P2P Lending (log units): -0.00741 (0.00565)
- Balance Sheet Lending (log units): 0.00325 (0.00659)
- Other FinTech (log units): -0.0305*** (0.00828)
- Size (log assets): 0.240*** (0.0353)
- Equity-to-assets: 0.0121*** (0.00185)
- GDP growth: -0.00854 (0.00599)
- Inflation: -0.00757* (0.00399)
- GDP per capita: 1.678*** (0.302)
- Policy rate: -0.00443 (0.00458)
- Concentration: -0.00113 (0.00129)
- N: 64,957
- rho: 0.66
- Bank Fixed Effects: Yes; Time Fixed Effects: Yes

### Table 3: Effect of FinTech Business Models and Risk Taking Across Commercial Banks
- Dependent variable: z-score (log units)
- FinTech (log units): -0.00459 (0.00936)
- P2P Lending (log units): 0.000537 (0.00731)
- Balance Sheet Lending (log units): -0.0301*** (0.0101)
- Other FinTech (log units): -0.0233** (0.0101)
- Size (log assets): 0.117** (0.0493)
- Equity-to-assets: 0.00261 (0.00251)
- GDP growth: 0.0206** (0.0101)
- Inflation: 0.00320 (0.00508)
- GDP per capita: 1.692*** (0.444)
- Policy rate: -0.00343 (0.00601)
- Concentration: 0.00238 (0.00262)
- N: 8,370
- rho: 0.77
- Bank Fixed Effects: Yes; Time Fixed Effects: Yes

### Table 4: Effect of FinTech Business Models and Risk Taking Across Cooperative Banks
- Dependent variable: z-score (log units)
- FinTech (log units): -0.00766 (0.0202)
- P2P Lending (log units): -0.0436*** (0.0147)
- Balance Sheet Lending (log units): 0.0143 (0.0164)
- Other FinTech (log units): -0.0284 (0.0195)
- Size (log assets): 0.269*** (0.0475)
- Equity-to-assets: 0.0217*** (0.00336)
- GDP growth: -0.0383*** (0.0107)
- Inflation: -0.0128 (0.0140)
- GDP per capita: 2.418*** (0.784)
- Policy rate: -0.0215** (0.00982)
- Concentration: -0.00265 (0.00170)
- N: 52,593
- rho: 0.69
- Bank Fixed Effects: Yes; Time Fixed Effects: Yes

### Table 5: Effect of FinTech Business Models and Risk Taking Across Non-Banks
- Dependent variable: z-score (log units)
- FinTech (log units): -0.0420** (0.0165)
- P2P Lending (log units): -0.0304** (0.015)
- Balance Sheet Lending (log units): -0.0208 (0.0133)
- Other FinTech (log units): -0.0296* (0.0166)
- Size (log assets): 0.276*** (0.0595)
- Equity-to-assets: 0.00654** (0.00256)
- GDP growth: -0.00802 (0.0147)
- Inflation: -0.00025 (0.00752)
- GDP per capita: 0.266 (0.737)
- Policy rate: -0.0015 (0.00741)
- Concentration: -0.000909 (0.00367)
- N: 5,621
- rho: 0.65
- Bank Fixed Effects: Yes; Time Fixed Effects: Yes

### Table 6: Effect of FinTech Business Models and Risk Taking Across ALL FIs (Risk-adjusted ROA / E/A, standardized)
- Columns (1)–(8) present alternate specifications of Risk-adjusted ROA and Risk-adjusted E/A (standardized)
- FinTech (standardized): -0.0604*** (0.0212) in column (1); -0.0453** (–0.0195) in column (2)
- P2P Lending (standardized): -0.0590*** (–0.0156) in column (3); -0.0640*** (–0.0137) in column (4)
- Balance Sheet Lending (standardized): -0.0866*** (–0.0132) in column (5); -0.0153 in column (6)
- Other FinTech (standardized): -0.0746*** (0.019) in column (7); -0.0593*** (0.0171) in column (8)
- Size (log assets): coefficients range e.g., 0.0635*** (0.0235), 0.0733*** (0.0225), etc.
- GDP growth: negative and significant across specifications, e.g., -0.0163*** (0.00426), -0.0248*** (0.00461)
- Inflation: mixed signs; significant negative in some specifications, e.g., -0.00845*** (0.00167)
- GDP per capita: positive and significant across specifications, e.g., 0.788*** (0.223)
- Policy rate: coefficients vary; e.g., 0.00995*** (0.00262) in column (1)
- Concentration: e.g., 0.00328*** (0.00125)
- N: 71,672 (columns 1–2), 68,446 (3–4), 57,565 (5–6), 71,645 (7–8) as reported
- rho: ranges, e.g., 0.58, 0.55, 0.59, 0.59, 0.79, 0.71, 0.55, 0.53
- Bank Fixed Effects: Yes; Time Fixed Effects: Yes

### Table 7: Effect of FinTech Business Models and Risk Taking Across Banks (Risk-adjusted ROA / E/A, standardized)
- FinTech (standardized): -0.0535** (0.0243) in column (1); -0.0344 (0.0223) in column (2)
- P2P Lending (standardized): -0.0631*** (0.0174) in column (3); -0.0686*** (0.0153) in column (4)
- Balance Sheet Lending (standardized): -0.0986*** (0.0159) in column (5)
- Other FinTech (standardized): -0.0888*** (0.023) in column (7); -0.0675*** (0.0207) in column (8)
- Size (log assets): small positive coefficients, some significant, e.g., 0.0452* (0.0260)
- GDP growth: consistently negative and significant, e.g., -0.0205*** (0.00504)
- Inflation: mixed; significant negative in some specs, e.g., -0.0104*** (0.00189)
- GDP per capita: positive and significant in multiple columns, e.g., 0.768*** (0.249)
- Policy rate: mixed signs; positive and significant in some columns, e.g., 0.0103*** (0.00330)
- Concentration: e.g., 0.00239* (0.00140)
- N: varies by column, e.g., 65,880; 63,121; 54,457; 65,860
- rho: ranges from 0.52 to 0.78
- Bank Fixed Effects: Yes; Time Fixed Effects: Yes

### Table 8: Effect of FinTech Business Models and Risk Taking Across Commercial Banks (Risk-adjusted ROA / E/A, standardized)
- FinTech (standardized): -0.027 (0.0323) in column (1); -0.0672** (0.0274) in column (2)
- P2P Lending (standardized): 0.0145 (0.0196) in column (3); -0.0108 (0.0157) in column (4)
- Balance Sheet Lending (standardized): -0.0843*** (0.0241) in column (5)
- Other FinTech (standardized): -0.0616** (0.0288) in column (7); -0.0661*** (0.0234) in column (8)
- Size (log assets): coefficients reported with larger standard errors; examples include 0.040 (0.0411)
- GDP growth: varied; significant negative in some columns, e.g., -0.0279** (0.0114)
- Inflation: mixed; some positive significant coefficients in specific columns, e.g., 0.0133*** (0.00414)
- GDP per capita: e.g., 0.604* (0.351)
- Policy rate: mixed; e.g., 0.00224 (0.00338)
- Concentration: positive and significant in many columns, e.g., 0.00489*** (0.00182)
- N: varies by specification (e.g., 8,568; 7,559; 2,614; 5,488)
- rho: ranges e.g., 0.49 to 0.85
- Bank Fixed Effects: Yes; Time Fixed Effects: Yes

### Table 9: Effect of FinTech Business Models and Risk Taking Across Cooperative Banks (Risk-adjusted ROA / E/A, standardized)
- FinTech (standardized): -0.0944 (0.0660) in column (1); 0.107 in column (2)
- P2P Lending (standardized): -0.251*** (0.0373) in column (3); -0.233*** (0.035) in column (4)
- Balance Sheet Lending (standardized): -0.0970** (0.0433) in column (5)
- Other FinTech (standardized): -0.0726 (0.054) in column (7)
- Size (log assets): positive and often significant, e.g., 0.0668** (0.0295)
- GDP growth: strongly negative and significant across columns, e.g., -0.0588*** (0.0107)
- Inflation: negative and significant in several columns, e.g., -0.0417*** (0.0125)
- GDP per capita: positive and significant in several columns, e.g., 1.116* (0.656)
- Policy rate: mixed, with some significant negative coefficients in certain columns, e.g., -0.0423*** (0.0153)
- Concentration: mixed; examples include 0.00247 (0.00207)
- N: 53,237 in multiple specifications
- rho: varies across columns, e.g., 0.57 to 0.87
- Bank Fixed Effects: Yes; Time Fixed Effects: Yes

### Table 10: Effect of FinTech on Risk Taking based on selected policy frameworks (z-score, below/above median splits)
- Dependent variable: z-score (log units); specifications split by Below median / Above Median across supervisory/regulatory frameworks
- FinTech (log units):
  - Below median: -0.0227** (0.00930)
  - Above Median: 0.00313 (0.0101)
  - Below median (other spec): -0.0226** (0.00961)
  - Above Median: -0.00865 (0.0167)
  - Below median: -0.0287** (0.0139)
  - Above Median: -0.0209*** (0.00785)
  - Below median: -0.0227*** (0.00747)
  - Above Median: 0.0781*** (0.0231)
- Size (log assets): positive and significant across all splits, e.g., 0.185*** (0.0454); 0.263*** (0.0434)
- Equity-to-assets: positive and often significant, e.g., 0.00427** (0.00207); 0.0136*** (0.00213)
- GDP growth: mixed signs across splits; examples include -0.0192* (0.00982) and 0.00818 (0.00912)
- Inflation: mostly insignificant in these splits
- GDP per capita: large positive effects in some below-median specifications, e.g., 1.408*** (0.354); some above-median specifications show smaller or negative coefficients
- Policy rate: mixed; examples include -0.00352 (0.00528) and 0.00517 (0.00945)
- Concentration: e.g., -0.00336* (0.00174) in one below-median spec
- N: reported per split, e.g., 14,051 (Below), 56,527 (Above) in first pair; other pairs vary (e.g., 61,371 / 8,921; 14,657 / 55,921; 21,010 / 49,568)
- rho: ranges reported, e.g., 0.72, 0.63, 0.63, 0.62, 0.68, 0.63, 0.67, 0.60
- Bank Fixed Effects: Yes; Time Fixed Effects: Yes

*Source: Authors calculations.*

### 2011. NBER Working Paper No. W18733. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2207268

### Does FinTech Increase Bank Risk Taking? — References and Thematic Summary

### Literature on FinTech, Digital Credit, and Bank Risk
- Ben Naceur, S., Candelon, B., Elekdag, S., & Emrullahu, D. (2023). Is FinTech Eating the Bank’s Lunch? IMF Working Paper No. 2023/239. https://www.imf.org/en/Publications/WP/Issues/2023/11/18/Is-FinTech-Eating-the-Bank-s-Lunch-540817
- Cevik, S. (2023). The Dark Side of the Moon? Fintech and Financial Stability. IMF Working Paper No. 2023/253. https://www.imf.org/en/Publications/WP/Issues/2023/12/08/The-Dark-Side-of-the-Moon-Fintech-and-Financial-Stability-542212
- Cornelli, G., Frost, J., Gambacorta, L., Rau, P. R., Wardrop, R., & Ziegler, T. (2023). Fintech and big tech credit: drivers of the growth of digital lending. Journal of Banking & Finance, 148, 106742. https://doi.org/10.1016/j.jbankfin.2022.106742
- Daud, S. N. M., Ahmad, A. H., Khalid, A., & Azman-Saini, W. N. W. (2022). FinTech and financial stability: Threat or opportunity? Finance Research Letters, 47, 102667. https://doi.org/10.1016/j.frl.2021.102667
- Deng, L., Lv, Y., Liu, Y., & Zhao, Y. (2021). Impact of Fintech on Bank Risk-Taking: Evidence from China. Risks, 9(5), 99. https://doi.org/10.3390/risks9050099
- Fang, Y., Wang, Q., Wang, F., & Zhao, Y. (2023). Bank fintech, liquidity creation, and risk-taking: Evidence from China. Economic Modelling, 127, 106445–106445. https://doi.org/10.1016/j.econmod.2023.106445
- Fung, D. W. H., Lee, W. Y., Yeh, J. J. H., & Yuen, F. L. (2020). Friend or foe: The divergent effects of FinTech on financial stability. Emerging Markets Review, 45, 100727. https://doi.org/10.1016/j.ememar.2020.100727
- Jakšič, M., & Marinč, M. (2018). Relationship banking and information technology: the role of artificial intelligence and FinTech. Risk Management, 21(1), 1–18. https://doi.org/10.1057/s41283-018-0039-y
- Murinde, V., Rizopoulos, E., & Zachariadis, M. (2022). The impact of the FinTech revolution on the future of banking: Opportunities and risks. International Review of Financial Analysis, 81(102103), 102103. https://doi.org/10.1016/j.irfa.2022.102103
- World Bank. (2022). Fintech and the Future of Finance. World Bank. https://www.worldbank.org/en/publication/fintech-and-the-future-of-finance

### Competition, Market Structure, and Bank Risk
- Beck, T., De Jonghe, O., & Schepens, G. (2013). Bank competition and stability: Cross-country heterogeneity. Journal of Financial Intermediation, 22(2), 218–244. https://doi.org/10.1016/j.jfi.2012.07.001
- Berger, A. N., Klapper, L. F., & Turk-Ariss, R. (2008). Bank Competition and Financial Stability. World Bank Policy Research Working Paper No.4696. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1243102
- Boyd, J. H., & De Nicoló, G. (2005). The Theory of Bank Risk Taking and Competition Revisited. The Journal of Finance, 60(3), 1329–1343. https://www.jstor.org/stable/3694928
- Martinez-Miera, D., & Repullo, R. (2010). Does Competition Reduce the Risk of Bank Failure? The Review of Financial Studies, 23(10), 3638–3664. https://www.jstor.org/stable/40865571
- Keeley, M. C. (1990). Deposit Insurance, Risk, and Market Power in Banking. The American Economic Review, 80(5), 1183–1200. https://www.jstor.org/stable/2006769
- Carvallo Valencia, O., & Ortiz Bolaños, A. (2018). Bank capital buffers around the world: Cyclical patterns and the effect of market power. Journal of Financial Stability, 38, 119–131. https://doi.org/10.1016/j.jfs.2018.02.004

### Regulation, Supervision, and Financial Stability
- Basel Committee on Banking Supervision (BCBS). (2016). Minimum capital requirements for market risk. Www.bis.org. https://www.bis.org/bcbs/publ/d352.htm
- Financial Stability Board (FSB). (2017). FinTech credit: Market structure, business models and financial stability implications. Www.bis.org. https://www.bis.org/publ/cgfs_fsb1.htm
- Federal Deposit Insurance Corporation (FDIC). (2015). Supervisory Insights. Vol.12, Issue 2. https://www.fdic.gov/regulations/examinations/supervisory/insights/siwin15/siwin15.pdf
- Coelho, R., Mazzillo, J., Svoronos, J.-P., and Yu, T. (2019). Regulation and supervision of financial cooperatives. https://www.bis.org/fsi/publ/insights15.pdf
- Chronopoulos, D. K., Wilson, J. O. S., & Yilmaz, M. H. (2023). Regulatory oversight and bank risk. Journal of Financial Stability, 64, 101105. https://doi.org/10.1016/j.jfs.2023.101105
- Kandrac, J., & Schlusche, B. (2020). The Effect of Bank Supervision and Examination on Risk Taking: Evidence from a Natural Experiment. The Review of Financial Studies, 34(6). https://doi.org/10.1093/rfs/hhaa090

### Bank Behavior, Risk Management, and Governance
- Berger, A. N., & Udell, G. F. (1995). Relationship Lending and Lines of Credit in Small Firm Finance. The Journal of Business, 68(3), 351–381. https://www.jstor.org/stable/2353332
- DeYoung, R., & Roland, K. P. (2001). Product Mix and Earnings Volatility at Commercial Banks: Evidence from a Degree of Total Leverage Model. Journal of Financial Intermediation, 10(1), 54–84. https://doi.org/10.1006/jfin.2000.0305
- Bülbül, D., Hakenes, H., & Lambert, C. (2019). What influences banks’ choice of credit risk management practices? Theory and evidence. Journal of Financial Stability, 40, 1–14. https://doi.org/10.1016/j.jfs.2018.11.002
- Laeven, L., & Levine, R. (2009). Bank governance, regulation and risk taking. Journal of Financial Economics, 93(2), 259–275. https://doi.org/10.1016/j.jfineco.2008.09.003
- Demsetz, R. S., Saidenberg, M. R., & Strahan, P. E. (1996). Banks with Something to Lose: The Disciplinary Role of Franchise Value. Economic Policy Review, 2(2). https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1028769
- Roy, A. D. (1952). Safety First and the Holding of Assets. Econometrica, 20(3), 431. https://doi.org/10.2307/1907413

### Empirical Methods and Data Sources Cited
- Blundell, R., and Bond, S. (1998). Initial conditions and moment restrictions in dynamic panel data models. Journal of Econometrics, 87(1), 115–143. https://doi.org/10.1016/s0304-4076(98)00009-8
- Sahay, R., Čihák, M., N’Diaye, P., Barajas, A., Ayala Pena, D., Bi, R., Gao, Y., & Kyobe, A. (2015). Rethinking Financial Deepening: Stability and Growth in Emerging Markets. IMF Staff Discussion Note SDN/15/08.
- World Bank. (2012). Global Financial Development Report 2013: Rethinking the Role of the State in Finance. Washington, DC: World Bank. doi:10.1596/978- 0-8213-9503-5
- Kaufmann, D., & Kraay, A. (2023). Worldwide Governance Indicators, 2023 Update. (www.govindicators.org)
- The Heritage Foundation. (2023). 2023 Index of Economic Freedom. https://www.heritage.org/index/pdf/2023/book/2023_IndexOfEconomicFreedom_FINAL.pdf

### Empirical Findings and Thematic Synthesis (as reflected in cited works)
- Evidence on whether FinTech increases bank risk-taking is mixed across studies and contexts:
  - Country-level and sectoral heterogeneity is documented (Beck, De Jonghe, & Schepens, 2013).
  - China-focused studies find impacts of FinTech on bank risk-taking, liquidity creation, and lending practices (Deng et al., 2021; Fang et al., 2023; Hu, Zhao, & Yang, 2022).
  - Some studies highlight divergent effects of FinTech on financial stability depending on market structure, business models, and regulatory responses (Fung et al., 2020; FSB, 2017).
- Competitive pressures from FinTech and big tech are identified as drivers of digital lending growth and potential shifts in banks’ business models (Cornelli et al., 2023; Ben Naceur et al., 2023).
- Regulation and supervisory frameworks, including capital requirements and oversight practice, are central to mitigating potential risks from FinTech-enabled credit expansion (BCBS, 2016; FDIC, 2015; Chronopoulos et al., 2023).

### Policy-Relevant Themes Highlighted by the References
- Strengthen supervisory capacity and adapt oversight to digital credit business models (FDIC, 2015; FSB, 2017; Chronopoulos et al., 2023).
- Reassess capital and market-risk frameworks in light of evolving digital lending and market structures (BCBS, 2016).
- Monitor heterogeneity across countries and banking systems when designing FinTech-related policy responses (Beck et al., 2013; World Bank, 2022).
- Incorporate governance, franchise value, and market-power considerations into risk assessments as FinTech shifts competitive dynamics (Demsetz et al., 1996; Laeven & Levine, 2009; Carvallo Valencia & Ortiz Bolaños, 2018).

*Does FinTech Increase Bank Risk Taking? Working Paper No. WP/2024/017 — References section (extracted from source content).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2024/english/wpiea2024017-print-pdf.pdf_