## _wp1381

## Source details

**Canonical URL:** [_wp1381](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp1381.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp1381.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp1381.pdf.json)

---

### I. Introduction — key findings and framing
- Positive, albeit non-linear, relationship between financial system depth, economic growth, and macroeconomic volatility.
- Rapid credit expansion associated with higher bank fragility and greater likelihood of systemic banking crises.
- Core mechanism: maturity and liquidity transformation from short-term savings into long-term investments supports growth but increases susceptibility to shocks; information asymmetries and agency problems create fragility.
- Financial possibility frontier introduced as a constrained optimum level of financial development to gauge relative performance across countries and over time.
- Policy taxonomy tied to position relative to the frontier:
  - Market-developing: push the frontier outwards (macroeconomic stability, long-term institutions, overcome small-size constraints).
  - Market-enabling: move the system toward the frontier (remove regulatory barriers, increase competition).
  - Market-harnessing: prevent movement beyond the frontier (regulatory oversight, macroeconomic management).
- Empirical approach: regression-based benchmarking to predict countries’ level of financial development from structural characteristics; relate gaps between predicted and actual levels to macroeconomic, regulatory, and institutional variables.
- Preliminary evidence: overshooting predicted financial development is associated with credit boom-bust episodes.

### II. The Financial Possibility Frontier — mechanisms, constraints, typology
- Market frictions constraining financial deepening: information, enforcement, and transactions costs; fixed transaction costs create economies of scale and higher intermediation costs in smaller systems.
- Risk constraints: contract-specific/idiosyncratic risks and systemic risks; macroeconomic uncertainty and weak contract enforcement worsen agency problems.
- State variables imposing an upper limit (invariant or slow-moving):
  - Structural: income, savings, market size, population density, age dependency ratios.
  - Macroeconomic management and credibility (fiscal discipline).
  - Legal and contractual frameworks (enforceability of contracts, credit registries, accounting and auditing standards, debtor/collateral information sharing).
  - Prudential oversight.
  - Technology and infrastructure (transportation and communications quality).
  - Socio-economic factors (conflict, financial illiteracy, informality).
- Definition: financial possibility frontier = maximum sustainable depth, outreach, or breadth of a financial system achievable at a given point in time, given state variables and market frictions.
- Possible positions relative to the frontier:
  - Frontier low relative to peers (structural deficits).
  - System below frontier due to demand- or supply-side constraints.
  - System beyond frontier (overshooting): unsustainable expansion, credit boom-bust cycles, often with weak regulation and governance.

### III. Taxonomy of financial sector policies — objectives and instruments
- Market-developing policies (long-term; push out frontier):
  - Legal upgrades, macroeconomic performance improvements (especially fiscal).
  - Regional integration and foreign bank entry to mitigate small market size constraints (with risk management).
  - Strengthen informational/contractual frameworks and market infrastructure.
- Market-enabling policies (short- to medium-term; move toward frontier):
  - Foster competition (e.g., expand micro- and consumer lending).
  - Remove regulatory impediments; reform tax policies.
  - Open infrastructures (payment systems, credit registries) to more institutions; platform sharing.
  - Use limited government interventions that incentivize private entry without shifting undue risks (partial credit guarantees, joint platforms).
- Market-harnessing policies (prevent overshoot; stabilize):
  - Risk oversight and management: regulatory framework, macroprudential management.
  - Upgrade regulation for non-bank providers and cross-border integration.
  - Calibrate liberalization pace to supervisory capacity.
  - User-side measures: financial literacy and consumer protection to reduce household over-indebtedness.
- Caveat: reform effects and priorities vary widely across countries; paths of financial deepening are not fully replicable.

### IV. Benchmarking financial systems — empirical specification and stylized results
- Empirical strategy: large cross-country panel over ~40 years to estimate time-varying benchmark levels and measure gaps.
- Benchmark regression: FD_{i,t} = β X_{i,t} + ε_{i,t}; FD is log of financial development indicator; X is array of structural factors.
- Structural factors included:
  - log of GDP per capita and its square; log of population; log of population density; log of age dependency ratio; dummies (off-shore center, transition country, oil-exporter); time dummies.
- Benchmark prediction: FD_{i,t}^B from regression; Gap = FD^B − FD (positive gap = under-performance; negative gap = over-performance).
- Stylized median changes (examples, 1990 to 2009):
  - Banking sector depth (Private Credit to GDP, extended):
    - In the median LIC: increased from 14 to 23 percent.
    - In high-income countries: increased from 41 to 98 percent of GDP.
    - In middle-income countries: increased from 22 to 37 percent.
  - Stock market capitalization:
    - LICs: increased from 5 percent of GDP in 1990 to 16 percent by 2009.
    - Middle- and high-income countries: increased from 20 to 40 percent of GDP over the same period.
- Benchmarking by income group (median gaps over time, selected highlights):
  - LICs:
    - Private Credit to GDP: gap just over 1 percent in 1990 and became negative over three decades; by 2009 median LIC outperformed benchmark by about 2 percent.
    - Stock Market Capitalization: positive gap of 4 percent in 1990 turned into negative gap of 7 percent by 2009.
    - Stock Market Turnover: by 2009 a positive gap of almost 3 percent persisted.
  - High-Income Countries:
    - Private Credit to GDP: eliminated a 25 percent gap in the run up to the crisis.
    - Stock Market Capitalization: gaps eliminated from 2000 onward.
    - Stock Market Turnover: gap reduced modestly (by 5 percent), remaining at a positive 6 percent in 2009.
  - Middle-Income Countries:
    - Private Credit to GDP: gaps remained virtually unchanged.
    - Stock Market Capitalization: gap turned negative in 2000 and dropped to -12 percent by 2009.
- Crisis impact: global crisis affected non-LICs more; median high-income countries’ Private Credit to GDP fell markedly below benchmark in 2008-09.
- Heterogeneity (1990–2007):
  - LICs’ Private Credit to GDP gap changes ranged between -40 and +30 percent; several countries lowered gaps by up to 20 percent.
  - Non-LICs: ranges much larger; some gaps closed or widened by over 100 percent.
  - By construction, sample-wide over-/under-performance shares should approach 50 percent but vary by income group and time.

### V. Explaining gaps — variables, univariate and multivariate findings
- Explanatory variable groups:
  - Macroeconomic: exchange rate regime flexibility (0 to 8), inverse of inflation, lagged growth, banking crisis dummy, remittances/GDP, gross capital inflows/GDP.
  - Market structure: foreign bank entry restrictions, five-bank concentration ratio, government ownership share, foreign-owned banks share, Lerner index.
  - Regulatory policy: geographic lending diversity requirements; Abiad et al. (2008) reform indicators (credit controls, privatization, quality of supervision, composite reform index).
  - Institutional: ICRG financial/economic/political risk indicators; World Bank creditor rights index.
- Key univariate correlates:
  - Lower inflation associated with over-performance (lower gaps).
  - Higher remittance inflows associated with over-performance (lower gaps).
  - Faster prior growth associated with over-performance (lower gaps).
  - Lower government-owned bank share associated with over-performance.
  - Fewer restrictions on foreign bank entry associated with faster gap reduction.
  - Higher quality banking supervision associated with lower gaps.
  - Geographic diversity restrictions on lending delay gap closure.
  - Stronger creditors’ rights associated with lower gaps.
- Insignificant/ambiguous in univariate regressions:
  - Fixed versus floating exchange rate regime not significantly associated with gaps.
  - Size of capital inflows not significantly associated with gaps.
  - Occurrence of a financial crisis in preceding decade not significantly associated with gaps.
  - Competition–gap relationships negative but not statistically significant in some specifications.

- Multivariate robust findings (levels and changes):
  - Associated with LOWER Private Credit Gaps:
    - Lower inflation (Inverse of inflation).
    - Larger remittance share.
    - Higher past growth (Lagged growth).
    - Lower share of government ownership (privatization).
    - Better quality of banking supervision.
    - Stronger creditors' rights.
    - Greater competition and overall financial reforms (effects depend on inclusion of privatization/supervision).
  - Associated with HIGHER Private Credit Gaps (in multivariate contrast):
    - Restrictions on foreign bank entry.
    - Greater exchange rate flexibility (examples: coefficients 2.437**, 2.528* in Table 2).
    - Gross capital inflows (examples: coefficients 0.028*** and 0.027** in Table 2).
  - Interactions and caveats:
    - Overall financial reform variable loses significance when privatization or banking supervision included — privatization and supervision matter most.
    - ICRG risk variables lose significance once macroeconomic, structural, and regulatory factors are controlled for.
  - Changes in gaps (1995–2007) findings:
    - Geographic diversity restrictions discourage lowering of the Private Credit Gap.
    - Asset concentration associated with narrowing gaps over time (Table 3: Asset concentration coefficients -32.470** and -36.255**).
    - Greater banking competition associated with narrowing gaps.
    - Stronger creditors’ rights associated with lower levels of gaps but not with changes over time.
    - Lower economic risk associated with faster gap reduction (Table 3: Economic risk coefficient -2.268*).

### VI. Booms, upper limits, and signals of instability
- Boom definition (Dell’Ariccia et al. (2012), 1980–2008):
  - Boom if either:
    - (i) deviation from a backward-looking rolling cubic trend (t-10 to t) > 1.5 times its standard deviation AND annual growth of private credit to GDP > 10 percent; OR
    - (ii) annual growth of private credit to GDP > 20 percent.
- Sample identification: 139 boom periods (1980–2008).
- Boom outcomes:
  - 34 percent of booms classified as bad (followed by a banking crisis within three years of its end date).
  - 59 percent of booms classified as sub-par (associated with a recession or below-trend medium-term growth).
- Relationship of Private Credit Gap to boom frequency/outcomes:
  - Periods with NEGATIVE gap (actual private credit above benchmark) are more likely to experience a boom.
  - Private credit to GDP 90-100 percent above benchmark is 50 percent more likely to be associated with a credit boom.
  - Extreme negative gaps are almost always bad or sub-par booms (end in low-growth episodes or banking crises).
  - Zero gaps have the lowest incidence of booms, both good and bad/subpar.
  - Large negative changes in the gap (rapid growth relative to benchmark) are associated with higher likelihood of booms; larger negative changes increase probability that the boom will be sub-par or end in crisis.
  - Underperformance (increasing gaps over time) is very rarely associated with booms or bad outcomes.
- Quantified signals and tentative upper limit:
  - Banking system instability much more likely when gaps are highly negative.
  - At negative gaps over 50 percent, probability of a crisis surpasses 10 percent.
  - As gaps approach 90 percent, likelihood of a crisis or subpar macroeconomic performance becomes very high.
  - Rapid deepening above structural changes — e.g., by 30 percent or more over a ten year period — further increases likelihood of instability.
  - Preliminary analysis suggests an upper limit to the financial possibility frontier of at least 50 percent above a country’s structural depth line, related to speed of deepening.

### VII. Policy implications and conclusions
- Usefulness:
  - Financial possibility frontier useful as a benchmarking heuristic to assess country standing relative to structural depth and peers, and to highlight policy/institutional gaps.
- Policies to reduce Private Credit Gaps (lift “too cold” systems):
  - Market-enabling and market-developing: lower restrictions on lending; privatization (limited government ownership of banks); strengthen creditors’ rights; promote greater competition even where consolidation occurs.
  - Strong supervisory and regulatory frameworks facilitate safe financial deepening.
- Tradeoffs and warning signs:
  - Evidence consistent with an upper threshold beyond which financial deepening becomes “too hot” and instability risks outweigh benefits.
  - Sufficient warning when private credit gap is around 50 percent above benchmark.
  - Lowest probability of bad outcome when actual depth is approximately at structural benchmark.
  - Policymakers need market-harnessing policies to monitor and mitigate emerging stability threats.
- Caveats:
  - Frontier is a heuristic, not derived from a utility-maximizing theoretical framework; general equilibrium models with intermediaries would better quantify trade-offs.
  - Policies treated as exogenous; political economy and implementation constraints matter — “all financial sector policy is local.”

*Excerpted from the IMF working paper content unit titled "_wp1381 - References ................................................................................................................................28" (source PDF content supplied).*

### 1. Univariate Regressions Explaining Levels and Changes in the Private Credit ...................25

### 1. Univariate Regressions Explaining Levels and Changes in the Private Credit ...................25

### Sections included
- 1. Univariate Regressions Explaining Levels and Changes in the Private Credit ...................25
- 2. Multivariate Regressions Explaining Levels of the Private Credit-GDP Gap .....................26
- 3. Multivariate Regressions Explaining the 1995-2005 Change in the Private .......................27

### Figures included
- 1. The Financial Possibility Frontier ........................................................................................19
- 2. Observed Financial Depth Over Time and Across Income Groups ....................................20
- 3. Gaps in Financial Depth Relative to Benchmarks ...............................................................21
- 4. Change in Gaps in Private Credit, 1990 to 2007 .................................................................22
- 5. Share of Underperforming Countries, by Income Level .....................................................23
- 6. Frequency of Credit Booms Related to the Level of the Private Credit ..............................24
- 7. Frequency of Credit Booms Related to the 1995-2005 Change in the Private Credit Gap .24

*Source: _wp1381 - 1. Univariate Regressions Explaining Levels and Changes in the Private Credit ...................25 (canonical URL: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp1381.pdf).*

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

### _wp1381 - References .............................................................................................................

### I. INTRODUCTION — key findings and framing
- Empirical evidence shows a positive, albeit non-linear, relationship between financial system depth, economic growth, and macroeconomic volatility.
- Rapid credit expansion has been associated with higher bank fragility and the likelihood of a systemic banking crisis.
- Core mechanism: maturity and liquidity transformation from short-term savings into long-term investments supports growth but increases susceptibility to shocks; information asymmetries and agency problems create fragility.
- Policy-space and transmission: the financial system plays a critical role in defining policy space and the transmission of fiscal, monetary and exchange rate policies.
- Concept introduced: financial possibility frontier as a constrained optimum level of financial development to gauge relative performance across countries and over time.
- Policy taxonomy tied to position relative to the frontier:
  - Market-developing policies: push the frontier outwards (macroeconomic stability, long-term institution building, overcome small-size/volatile-structure constraints).
  - Market-enabling policies: help a system move toward the frontier (address regulatory barriers, lack of competition).
  - Market-harnessing policies: prevent movement beyond the frontier (regulatory oversight, short-term macroeconomic management).
- Empirical approach: regression-based benchmarking to predict countries’ level of financial development from structural characteristics; relate gaps between predicted and actual levels to macroeconomic, regulatory, and institutional variables.
- Preliminary evidence: overshooting predicted financial development is associated with credit boom-bust episodes.

*References cited in this section: Levine (2005); Beck (2012); Beck and de la Torre (2007); Beck et al. (2008); Al Hussainy et al. (2011); Arcand et al. (2012); Dabla-Norris and Srivisal (2013); Detragiache and Demirguc-Kunt (2005); Claessens et al. (2011); Rioja and Valev (2004a, 2004b); Aghion et al. (2005).*

### II. THE FINANCIAL POSSIBILITY FRONTIER — mechanisms, constraints, and typology of positions
- Market frictions that constrain financial deepening: information, enforcement, and transactions costs.
- Fixed transaction costs at transaction, client, institution, and system levels create economies of scale and explain higher intermediation costs in smaller financial systems.
- Risk constraints: contract-specific/idiosyncratic risks and systemic risks increase costs and limit supply; macroeconomic uncertainty and weak contract enforcement exacerbate agency problems.
- State variables (invariant or slow-moving determinants) that impose an upper limit on financial deepening include:
  - Structural variables: income, savings, market size, population density, age dependency ratios.
  - Macroeconomic management and credibility (degree of fiscal discipline).
  - Legal, contractual, and information frameworks (enforceability of contracts, credit registries, accounting and auditing standards, debtor/collateral information sharing).
  - Prudential oversight.
  - Available technology and infrastructure (transportation and communications quality).
  - Socio-economic factors (conflict, financial illiteracy, degree of informality).
- Definition: the financial possibility frontier is the maximum sustainable depth, outreach, or breadth of a financial system achievable at a given point in time, given state variables and market frictions.
- Possible country positions relative to the frontier:
  - Frontier low relative to peers due to structural/state variable deficiencies (e.g., low population density, small market size, informality).
  - System lies below the frontier due to demand-side constraints (self-exclusion, lack of viable projects) and supply-side constraints (lack of competition, regulatory restrictions, weak creditor information).
  - System moves beyond the frontier (overshooting): unsustainable expansion, credit boom-bust cycles, often associated with weak regulatory/supervisory frameworks and governance problems.

### III. TAXONOMY OF FINANCIAL SECTOR POLICIES — policy objectives and instruments
- Market-developing policies (push out the frontier; long-term):
  - Legal changes and upgrading macroeconomic performance (especially fiscal).
  - Regional integration and foreign bank entry to mitigate small market size constraints (with careful risk management).
  - Strengthen informational and contractual frameworks and market infrastructure.
  - Note: benefits materialize over the longer term.
- Market-enabling policies (move system toward frontier; short- to medium-term):
  - Foster greater competition (example: expansion of micro- and consumer lending).
  - Remove regulatory impediments and reform tax policies.
  - Open infrastructures (payment systems, credit registries) to more institutions; force platform sharing where appropriate.
  - Address coordination failures, first-mover disincentives, obstacles to risk distribution/sharing.
  - Use limited government interventions that create incentives for private sector entry without shifting undue risks/costs to government (e.g., partial credit guarantees, joint platforms).
- Market-harnessing (market-stabilizing) policies (prevent overshoot):
  - Risk oversight and management: regulatory framework, macroeconomic and macro-prudential management.
  - Upgrade regulation to mitigate risks from new non-bank providers and cross-border integration.
  - Calibrate pace of financial liberalization to supervisory capacity.
  - User-side measures: financial literacy programs and consumer protection to reduce household over-indebtedness.
- Caveats:
  - Paths of financial deepening are not necessarily replicable across countries; reforms’ relative importance and cost-benefit tradeoffs vary widely with country-specific circumstances.

### IV. BENCHMARKING FINANCIAL SYSTEMS — empirical specification and stylized results
- Empirical strategy: use a large cross-country panel spanning over a 40 year period to estimate time-varying benchmark levels of financial development and measure gaps.
- Benchmark regression specified (as in Al Hussainy et al. (2011)):
  - FD_{i,t} = β X_{i,t} + ε_{i,t}        (1)
  - FD is the log of an indicator of financial development; X is an array of structural country-specific factors.
- Structural factors included:
  - log of GDP per capita and its square (to account for non-linearities).
  - log of population (proxy for market size).
  - log of population density (proxy for ease of service provision).
  - log of the age dependency ratio (demographics and savings behavior).
  - other fundamentals: an off-shore center dummy, a transition country dummy, and an oil-exporting country dummy; time dummies for global factors.
- Benchmark prediction: regression results predict benchmark level FD_{i,t}^B for each country-year.
- Gap definition: Gap = FD^B − FD
  - A positive gap indicates under-performance; a negative gap indicates over-performance.
  - Text reproduction from source: "B ititit GapFDFD"
- Stylized empirical findings (1990 to 2009 examples):
  - Banking sector depth (Private Credit to GDP, extended by banks and other financial institutions) — median changes:
    - In the median LIC, increased from 14 to 23 percent.
    - In high-income countries, increased from 41 to 98 percent of GDP.
    - In middle-income countries, increased from 22 to 37 percent.
  - Stock market capitalization:
    - LICs: increased from 5 percent of GDP in 1990 to 16 percent by 2009.
    - Middle- and high-income countries: increased from 20 to 40 percent of GDP over the same period.
  - Market turnover ratios in LICs: turnover increased from 2 to [text truncated in source].

*Italic source attribution: Excerpted from the IMF working paper content unit titled "_wp1381 - References ................................................................................................................................28" (source PDF content supplied).*

### 4.5 percent of GDP over this period).

### _wp1381 - 4.5 percent of GDP over this period).

### Benchmarking financial deepening by income group
- Comparing observed trends to structural benchmarks gauges whether cross-country differences reflect structural versus policy-related factors.
- Figure 3 findings (by income group, median gaps over time):
  - Low-Income Countries (LICs):
    - Private Credit to GDP: gap was just over 1 percent in 1990 and became negative over the subsequent three decades; by 2009 the median LIC was outperforming its benchmark by about 2 percent.
    - Stock Market Capitalization: positive gaps of 4 percent in 1990 turned into a negative gap of 7 percent by 2009.
    - Stock Market Turnover: by 2009 a positive gap of almost 3 percent persisted.
  - High-Income Countries:
    - Private Credit to GDP: eliminated a 25 percent gap in the run up to the crisis.
    - Stock Market Capitalization: gaps eliminated from 2000 onward.
    - Stock Market Turnover: gap reduced modestly (by 5 percent), remaining at a positive 6 percent in 2009.
  - Middle-Income Countries:
    - Private Credit to GDP: gaps remained virtually unchanged.
    - Stock Market Capitalization: gap turned negative in 2000 and dropped to -12 percent by 2009.
- Crisis impact:
  - The global crisis had a more significant impact on non-LICs, particularly high-income countries, where actual Private Credit to GDP in the median country fell markedly below its benchmark level in 2008-09.
- Data coverage caveat:
  - Stock market indicators have limited LIC coverage: only five LICs reported in 1990, increasing to 20 by 2006.

### Cross-country heterogeneity in financial deepening (1990–2007 focus)
- Heterogeneity:
  - Financial deepening over 1990-2007 was quite heterogeneous across countries; somewhat less so among LICs than other income groups.
- Private Credit to GDP gap changes (1990-2007):
  - LICs: changes ranged between -40 and +30 percent, with several countries lowering gaps by up to 20 percent.
  - Non-LICs: ranges much larger; some gaps closed or widened by over 100 percent.
- Distributional patterns:
  - More LICs lowered than increased their gaps in Private Credit to GDP.
  - For non-LICs, roughly the same number increased as lowered their gaps.
- Expected sample property:
  - By construction, the share of countries over- or under-performing their benchmarks should approach 50 percent for the sample period as a whole, but these shares vary by income levels and over time.
- Trends in share underperforming (Figure 5 summary):
  - Downward trend in share of underperforming countries for Bank Deposits to GDP and Stock Market Capitalization to GDP, especially in recent years and for low- and middle-income countries.
  - Opposite trend for Private Credit to GDP and Stock Market Turnover, reflecting the impact of the global financial crisis on high-income countries.

### Interpretation and limitations of the benchmarking model
- The benchmarking model:
  - Tracks progress of a financial system over time relative to structural characteristics and comparator countries.
  - Is a relative, not absolute measure: depends on the distribution within the sample used for benchmarking.
- Model specification:
  - Findings can be sensitive to model specification and the set of explanatory variables included.
- Suggested extensions:
  - Future work could extend benchmarking to include longer-term institutional and macroeconomic variables that impact financial system performance.
- Note on contemporaneous increases:
  - If all countries in the sample increase financial development indicators in a given year, the benchmark for the country in question will also increase.

### Explaining gaps: focus, approach, and variables
- Focus:
  - Determinants of Private Credit to GDP, chosen for broad coverage across countries and time and its robust relation to economic growth.
- Empirical approach:
  - Cross-country regressions on both levels of the gap averaged over 2000 to 2007 and the change in the Private Credit Gap between 1995 and 2007.
  - Exclusion of the recent crisis from the sample period to draw more general conclusions.
  - Initial univariate regressions used to identify simple correlations (reported in Table 1).
- Explanatory variable groups:
  - Macroeconomic variables:
    - Exchange rate regime flexibility indicator ranging from zero (hard peg) to 8 (freely floating).
    - Inverse of inflation as proxy for macroeconomic stability.
    - Lagged growth.
    - Dummy for banking crises.
    - Remittance flows as a share of GDP.
    - Gross capital inflows relative to GDP.
  - Market structure:
    - Indicator of foreign bank entry restrictions.
    - Five-bank concentration ratio.
    - Share of government ownership.
    - Share of foreign-owned banks.
    - Lerner index of market power, averaged across banks within a country.
  - Regulatory policy variables:
    - Requirements to geographically diversify lending.
    - Indicators of the state of financial reform from Abiad, et al (2008) including credit controls, privatization, quality of bank supervision, and an overall financial reform index.
  - Institutional variables:
    - ICRG indicators of financial, economic and political risk.
    - World Bank’s creditor rights index.

### Key empirical findings on determinants of gaps
- Macroeconomic correlates:
  - Countries with lower inflation rates tend to over-perform (lower gaps).
  - Higher remittance inflows are associated with over-performance (lower gaps).
  - More rapid previous growth associated with over-performance (lower gaps).
- Market structure and ownership:
  - Lower share of government-owned banks associated with over-performance (lower gaps).
  - Fewer restrictions on foreign bank entry tend to be associated with more rapid gap reduction over time.
- Regulatory and supervisory variables:
  - Higher quality and strength of banking supervision is associated with better performance (lower gaps).
  - Existence of geographical diversity restrictions on bank lending tends to delay closing of gaps over time.
- Institutional variables:
  - All institutional variables are significantly associated with either the level or changes in gaps.
  - Countries with lower overall risk have lower levels of gaps and succeed in closing existing gaps more rapidly.
  - Political and economic risk indices are particularly significant.
  - Stronger creditors’ rights tend to be associated with lower gaps.

### Insignificant or ambiguous determinants
- Variables found not to be significant determinants of gaps (univariate regressions):
  - Fixed versus floating exchange rate regime does not appear significantly associated with gaps.
  - Size of capital inflows not significantly associated with gaps.
  - Occurrence of a financial crisis in the preceding decade not significantly associated with gaps.
- Competition and concentration:
  - Regressions show a negative relationship between competition and gaps, but this is not statistically significant.
  - Some evidence that more concentrated banking systems narrowed their gaps more rapidly between 1995 and 2007.

*Source: _wp1381 - 4.5 percent of GDP over this period).*

### 2007. Other measures of financial reform, including the composite index, are also associated

### _wp1381 - 2007. Other measures of financial reform, including the composite index, are also associated with lower gaps, but not significantly.

### Multivariate regressions and determinants of Private Credit Gaps
- Data and approach:
  - Eight different multivariate specifications presented to deal with shrinking country sample when many regressors are included.
  - Dependent variable: Private Credit-GDP gap (difference between predicted/benchmark and actual levels); also regressions for changes in the gap between 1995 and 2007.
- Robust multivariate findings (levels and changes):
  - Associated with LOWER Private Credit Gaps:
    - Lower inflation (Inverse of inflation).
    - Larger remittance share.
    - Higher past growth (Lagged growth).
    - Lower share of government ownership (privatization component captures this).
    - Better quality of banking supervision.
    - Stronger creditors' rights (creditors' rights coefficients significant in Table 2).
    - Greater competition and overall financial reforms (though effects depend on inclusion of privatization or bank supervision).
  - Associated with HIGHER Private Credit Gaps (in multivariate, contrasting univariate results):
    - Restrictions on foreign bank entry.
    - Greater exchange rate flexibility (Exchange rate regime coefficients positive and significant in some specs: e.g., 2.437**, 2.528* in Table 2).
    - Gross capital inflows (Gross inflows: coefficients include 0.028*** and 0.027** in Table 2).
  - Interactions and caveats:
    - The significance of the overall financial reform variable disappears when privatization or banking supervision are included — indicating privatization and supervision matter most for depth.
    - All ICRG risk variables lose significance once macroeconomic, structural, and regulatory factors are controlled for.
- Results for changes in gaps (1995–2005 / 1995–2007):
  - Restrictions on geographic diversity discourage lowering of the Private Credit Gap.
  - Concentration in the banking system (Asset concentration) tends to be associated with narrowing gaps over time (Table 3: Asset concentration coefficients -32.470** and -36.255**).
  - Greater banking sector competition also associated with narrowing gaps — implying pure consolidation that reduces competition is unlikely to generate substantial financial depth.
  - Stronger creditors’ rights, while robustly associated with lower levels of gaps, do not seem to influence changes in gap levels over time.
  - Evidence that countries with lower economic risk tended to lower their gaps more rapidly over time (Table 3: Economic risk coefficient -2.268*).

### Robustness and alternative indicators
- Robustness tests (available on request) re-ran regressions using other financial sector indicators:
  - Stock market capitalization to GDP.
  - Stock market turnover.
  - Interest rate margin.
- While coefficient sizes and significance vary, the broad findings are confirmed.

### Financial Possibility Frontier, booms, and upper limits
- Boom definition (Dell’Ariccia et al. (2012), applied 1980–2008 sample):
  - Boom if either: (i) deviation from a backward-looking rolling cubic trend (t-10 to t) is greater than 1.5 times its standard deviation AND annual growth rate of private credit to GDP exceeds 10 percent; OR (ii) annual growth rate of private credit to GDP exceeds 20 percent.
  - Sample identification: 139 boom periods (1980–2008).
  - Boom classification in sample:
    - 34 percent of booms are classified as bad (followed by a banking crisis within three years of its end date).
    - 59 percent of booms are classified as sub-par (associated with a recession or below-trend medium-term growth).
- Relationship between Private Credit Gap and boom frequency/outcomes:
  - Periods with a NEGATIVE gap (actual private credit to GDP above benchmark) are more likely to experience a boom episode.
  - Having a level of private credit to GDP of 90-100 percent above the benchmark is 50 percent more likely to be associated with a credit boom.
  - Such extreme negative gaps are almost always bad or sub-par booms (end in low-growth episodes or banking crises).
  - Zero gaps (actual depth close to structural benchmark) have the lowest incidence of booms, both good and bad/subpar.
  - Large negative changes in the gap (rapid growth relative to benchmark) are associated with higher likelihood of boom episodes; the larger the negative change, the higher the likelihood the boom will be sub-par or end in a crisis.
  - Underperformance (increase in gaps over time) is very rarely associated with booms or bad outcomes.
- Quantified signals of instability and a tentative upper limit:
  - Banking system instability is much more likely when gaps are highly negative (country lies significantly above structural depth line).
  - At negative gaps of over 50 percent, the probability of a crisis surpasses 10 percent.
  - As gaps approach 90 percent, the likelihood of a crisis or subpar macroeconomic performance becomes very high.
  - Rapid deepening above structural changes — e.g., by 30 percent or more over a ten year period — further increases the likelihood of instability.
  - Preliminary analysis points to an upper limit to the financial possibilities frontier of at least 50 percent above a country’s structural depth line, which is related to the speed of deepening.

### Policy implications and conclusions
- Concept and utility:
  - The financial possibility frontier is introduced as a benchmarking concept to assess a country’s standing relative to its structural depth frontier and peers with similar characteristics, highlighting policy and institutional gaps.
- Policy measures that can reduce Private Credit Gaps (pull countries out of “too cold” status):
  - Market-enabling and market-developing policies, such as:
    - Lower restrictions on lending.
    - Limited government ownership of banks (privatization).
    - Strong creditors’ rights.
  - Promoting greater competition in banking can enhance financial deepening even if the system becomes more concentrated.
  - Stronger supervisory and regulatory frameworks can facilitate financial deepening.
- Tradeoffs and warning signs:
  - Evidence consistent with an upper threshold to financial deepening beyond which system becomes “too hot” and instability risks outweigh benefits.
  - Sufficient warning when private credit gap is around 50 percent above benchmark.
  - Lowest probability of a bad outcome when actual depth is approximately at its structural benchmark.
  - Policymakers need market-harnessing policies to monitor and mitigate potential/emerging stability threats.
- Caveats:
  - The financial possibility frontier is a heuristic, not derived from a utility-maximizing theoretical framework; general equilibrium models with financial intermediaries would be important for quantifying policy impacts and trade-offs.
  - The paper treats policies as exogenous tools; political economy considerations (policy implementation constraints) are important — “all financial sector policy is local.”

*Italic: Source — _wp1381 - 2007. Other measures of financial reform, including the composite index, are also associated with lower gaps, but not significantly.*

### References

### _wp1381 - References

### Databases, Tools, and Benchmarking
- Abiad, Abdul, Enrica Detragiache, and Thierry Tressel, 2008, “A New Database of Financial Reforms,” IMF Working Paper 08/266 (Washington: International Monetary Fund).
- Al Hussainy, Ed, Andrea Coppola, Erik Feyen, Alain Ize, Katie Kibbuka, and Haocong Ren, 2011, Fin Stats 2011: A Ready-to-Use Tool to Benchmark Financial Sectors Across Countries and Over Time (Washington: World Bank).
- Beck, Thorsten, Erik H. B. Feyen, Alain Ize, and Florencia Moizeszowicz, 2008, “Benchmarking Financial Development,” Policy Research Working Paper No. 4638 (Washington: World Bank).
- Djankov, Simeon, Caralee McLiesh, and Andrei Shleifer, 2007, “Private Credit in 129 Countries,” Journal of Financial Economics, Vol. 84, pp. 299-329.

### Empirical Studies on Finance and Growth
- Aghion, Philippe, Peter Howitt, and David Mayer-Foulkes, 2005, “The Effect of Financial Development on Convergence: Theory and Evidence.” Quarterly Journal of Economics, Vol. 120, pp. 173–222.
- Arcand, Jean Louis, Enrico Berkes, and Ugo Panizza, 2011, “Too Much Finance?” IMF Working Paper 12/161 (Washington: International Monetary Fund).
- Barajas, Adolfo, Ralph Chami, and Seyed Reza Yousefi, 2012, “The Finance and Growth Nexus Re-examined: Do All Countries Benefit Equally?” unpublished IMF Working Paper (Washington: International Monetary Fund).
- Beck, Thorsten, 2012, “The Role of Finance in Economic Development – Benefits, Risks, and Politics,” in The Oxford Handbook of Capitalism, ed. by Dennis Müller (U.K.: Oxford University Press).
- Beck, Thorsten, Ross E. Levine, and Norman Loayza, 2000, “Finance and the Sources of Growth,” Journal of Financial Economics, Vol. 58(1–2), pp. 261–300.
- Rioja, Felix, and Neven Valev, 2004, “Finance and the Sources of Growth at Various Stages of Economic Development.” Economic Inquiry Journal, Vol. 42, pp. 127–140.
- ________, 2005, “Does One Size Fit All? A Reexamination of the Finance and Growth Relationship,” Journal of Development Economics, Vol. 74, pp. 429–447.
- Levine, Ross, 2005, “Finance and Growth: Theory and Evidence,” in Handbook of Economic Growth, ed. by P. Aghion and S. Durlauf (Elsevier Science, The Netherlands).

### Financial Sector Structure, Institutions, and Access
- Beck, Thorsten and Augusto de la Torre, 2007, “The Basic Analytics of Access to Financial Service: Financial Markets, Institution and Instruments,” World Bank Publications, Vol. 17, pp. 79-117 (Washington: World Bank).
- Beck, T. Demirgüç-Kunt, and Vojislav Maksimovic, 2005, “Financial and Legal Constraints to Firm Growth: Does Firm Size Matter?” Journal of Finance, Vol. 2, pp. 137-177.
- La Porta, Rafael, Florencio Lopez-de-Silanes, and Andrei Shleifer, 2002 “Government Ownership of Commercial Banks,” Journal of Finance, Vol. 57, pp. 265–301.
- Merton, Robert, and Zvi Bodie, 2005, “The Design of Financial Systems: Towards a Synthesis of Function and Structure,” NBER Working Papers No. 01/2004 (Cambridge, Massachusetts: National Bureau of Economic Research).
- De la Torre, Augusto, Erik Feyen, and Alain Ize, 2013, “Financial Development: Structure and Dynamics.” World Bank Economic Review, forthcoming.

### Financial Stability, Cycles, and Macrofinancial Policy
- Claessens, Stijn, Ahyan Kose, and Marco Terrones, 2011, “Financial cycles: What? How? When?” IMF Working Paper 11/76 (Washington: International Monetary Fund).
- Claessens, Stijn, Gergely Dobos, Daniela Klingebiel, and Luc Laeven, 2003, “The Growing Importance of Networks in Finance and its Effects of Competition,” in Innovations in Financial and Economic Networks, ed. by Anna Nagurney, Edward Elgar Publishers (Northampton, MA, USA).
- Dell’Ariccia, Giovanni, Deniz Igan, Luc Laeven, Hui Tong, Bas Bakker, and Jerome Vandenbussche, 2012, “Policies for Macrofinancial Stability: How to Deal With Credit Booms,” Staff Discussion Note 12/06 (Washington: International Monetary Fund).
- Dabla-Norris, Era and Srivisal Narapong, 2013, “Revisiting the Link between Finance and Macroeconomic Volatility.” IMF Working Paper 13/29 (Washington: International Monetary Fund).
- Demirguc-Kunt, Asli, and Enrica Detragiache, 2005. “Cross-Country Empirical Studies of Systemic Banking Distress: A Survey,” IMF Working Paper 05/96 (Washington: International Monetary Fund)
- Boyd, John, Ross Levine, and Bruce D. Smith, 2001, “The Impact of Inflation on Financial Sector Performance,” Journal of Monetary Economics, Vol. 47, pp. 221-48.

### Political Economy and Governance of Finance
- Haber, Stephen, and Enrico Perotti, 2008, “The Political Economy of Financial Systems,” Tinbergen Institute Discussion Paper, (Tinbergen, Amsterdam).
- Quintyn, Marc, and Geneviève Verdier. 2010, “Mother, Can I Trust the Government? Sustained Financial Deepening—A Political Institutions View,” IMF Working Paper 10/210 (Washington: International Monetary Fund).

### Foundational and Survey Works
- Fry, Maxwell, 1988, Money, Interest, and Banking in Economic Development (Baltimore, MD: Johns Hopkins University Press).

*References list as provided in _wp1381 - References*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2013/_wp1381.pdf_