## _wp1605

## Source details

**Canonical URL:** [_wp1605](https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1605.pdf)

## Other formats

- [Markdown version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1605.pdf.md)
- [Structured JSON version](/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1605.pdf.json)

---

### Introduction: purpose and scope
- Objective: construct indices summarizing how developed financial institutions and financial markets are in terms of depth, access, and efficiency.
- Coverage: 183 advanced, emerging, and low-income developing countries on annual frequency between 1980 and 2013 (33 years).
- Conceptual definition: financial development = combination of depth (size and liquidity of markets), access (ability of individuals and companies to access financial services), and efficiency (ability of institutions to provide financial services at low cost and with sustainable revenues, and the level of activity of capital markets).
- Rationale: single indicators (e.g., private credit/GDP, stock market capitalization/GDP) are inadequate because financial development is multidimensional and includes institutions beyond banks and markets beyond bank lending.

### Index construction and methodological framework
- Three-step approach:
  - (i) normalization of variables;
  - (ii) aggregation of normalized variables into sub-indices for each functional dimension;
  - (iii) aggregation of sub-indices into the final index.
- Indices constructed (nine total):
  - Six lower-level sub-indices: FID, FIA, FIE, FMD, FMA, FME (I = institutions, M = markets; D = depth, A = access, E = efficiency).
  - Two higher-level sub-indices: FI (financial institutions overall) and FM (financial markets overall).
  - Final overall index: FD (financial development index).
- Aggregation choices addressed: selection of data series; treatment of missing data; normalization and outlier treatment; functional form of aggregator; weights in aggregation.
- Methodological reference: OECD Handbook on Constructing Composite Indicators (OECD, 2008).

### Data sources and indicator universe
- Primary data sources:
  - World Bank FinStats 2015 (Feyen, Kibuuka, and Sourrouille, 2014)
  - IMF’s Financial Access Survey
  - Dealogic corporate debt database
  - Bank for International Settlements (BIS) debt securities database
- Key category-level indicators (selected):
  - Financial Institutions — Depth: Private-sector credit to GDP; Pension fund assets to GDP; Mutual fund assets to GDP; Insurance premiums (life + non-life) to GDP — source: FinStats 2015.
  - Financial Institutions — Access: Bank branches per 100,000 adults (FinStats 2015); ATMs per 100,000 adults (IMF Financial Access Survey).
  - Financial Institutions — Efficiency: Net interest margin; Lending-deposits spread; Non-interest income to total income; Overhead costs to total assets; Return on assets; Return on equity — source: FinStats 2015.
  - Financial Markets — Depth: Stock market capitalization to GDP; Stocks traded to GDP; International debt securities of government to GDP (BIS); Total debt securities of financial corporations and nonfinancial corporations to GDP (Dealogic).
  - Financial Markets — Access: Percent of market capitalization outside of top 10 largest companies; Total number of issuers of debt (domestic and external, nonfinancial and financial corporations) per 100,000 adults — source: FinStats 2015.
  - Financial Markets — Efficiency: Stock market turnover ratio (stocks traded to capitalization) — source: FinStats 2015.
- Source note: "Source: IMF staff estimates."

### Missing data treatment, splicing, and specific adjustments
- Two broad missing-data situations:
  - Entire data series unavailable for a country: set to zero (interpreted as market does not exist or access/efficiency very poor).
  - Staggered start of data collection across series: three treatment options considered — (i) treat as missing and exclude series from index average; (ii) treat as zero; (iii) splice indices before and after series becomes available.
- Preferred approach: splicing (applied at raw-data level) to avoid index movements driven solely by addition of new series.
  - Iterative reconstruction uses average growth rates in order: similar series within same sub-index → series for same type of provider → across providers.
  - Two profit indicators excluded from splicing because they span negative and positive ranges.
  - Scope of reconstruction: about 27 percent of the sample is reconstructed through splicing; 32 percent of the sample consists of “missing” markets.
- Specific data adjustments ("butt splicing"):
  - Private sector credit corrected for one-off jumps in banking system coverage for a small number of countries by taking most recent level as representative and merging downward shifts through growth rates; cross-checked with IFS notes.
  - Illustrative source-data jumps preserved from source: Denmark credit to GDP jumps from 30 percent of GDP in 1999 to 135 percent in 2000; Sweden credit to GDP jumps from 40 to 93 percent of GDP in 2001.
- Imputation for recent-year gaps:
  - Where data not available for latest year (example: 2013), values set equal to latest available observations (example: 2012).
  - If a data series is completely not available for a country, the entire series is set at zero.

### Normalization, outlier treatment, and aggregation functional form
- Winsorization:
  - Each series winsorized at the 5th and 95th percentiles using the global distribution across countries and time.
- Min-max normalization:
  - Winsorized indicators normalized between 0 and 1 using min-max procedure; highest (lowest) value across time and countries set to one (zero).
  - For measures where higher values indicate worse performance (net interest margin, lending-deposits spread, noninterest income to total income, overhead costs to total assets), ratings are rescaled so higher normalized values indicate greater financial development.
- Aggregation functional form:
  - Weighted linear average used for sub-indices and higher-level indices (weights from PCA); assumes full compensability.
  - Geometric aggregation considered but rejected due to substantial zero bias in presence of zero/near-zero indicator ratings (example preserved: Luxembourg’s FMD score drops from 0.75 to 0.25 under geometric averaging; ranking falls by 29 places).
- Weights:
  - Principal component analysis (PCA) used to derive weights; sub-index weights are squared factor loadings (sum to 1) from PCA on underlying series.
  - First principal component interpreted as summarizing latent information on degree of financial development and captures between 51 and 92 percent of variance in sub-index data.
  - PCA pooling across all countries (LIDC, EM, AM) and all years (1980-2013).
  - Example weight magnitude preserved: Banking system credit to the private sector has a weight of 0.25 within the depth subcomponent of FI; FI subcomponent weight for that depth subcomponent is less than 0.40 in FI overall.

### Complete procedure (step-by-step)
- (i) Apply missing data treatment to actual data (including splicing where appropriate).
- (ii) Winsorize indicators: set 5th and 95th percentiles at cutoff levels.
- (iii) Normalize indicators to 0-1 via min-max procedure (higher value indicates greater financial depth after rescaling where needed).
- (iv) Construct sub-indices as weighted averages of normalized series, with weights equal to squared factor loadings (sum to 1) from PCA on the underlying series.
- (v) Combine sub-indices into higher-level indices via the same PCA-based weighted averaging and re-normalize.
- Final output: relative ranking of countries on depth, access, and efficiency of financial institutions and financial markets, FI, FM, and overall FD for years 1980–2013.

### Robustness, comparison with traditional measures, and correlations
- Purpose: summarize diverse indicators into indices that reduce dimensionality for empirical work and provide comprehensive assessment across depth, access, and efficiency.
- Correlations with traditional measures (ρ values preserved exactly):
  - Correlation with Private Credit to GDP:
    - FD vs Private credit to GDP ρ = 0.81
    - FI vs Private credit to GDP ρ = 0.82
    - FM vs Private credit to GDP ρ = 0.7
    - FID vs Private credit to GDP ρ = 0.84
    - FIA vs Private credit to GDP ρ = 0.66
    - FIE vs Private credit to GDP ρ = 0.37
    - FMD vs Private credit to GDP ρ = 0.7
    - FMA vs Private credit to GDP ρ = 0.63
    - FME vs Private credit to GDP ρ = 0.51
  - Correlation with Stock Market Capitalization/GDP:
    - FD vs Stock market cap. to GDP ρ = 0.62
    - FI vs Stock market cap. to GDP ρ = 0.52
    - FM vs Stock market cap. to GDP ρ = 0.63
    - FID vs Stock market cap. to GDP ρ = 0.61
    - FIA vs Stock market cap. to GDP ρ = 0.30
    - FIE vs Stock market cap. to GDP ρ = 0.34
    - FMD vs Stock market cap. to GDP ρ = 0.74
    - FMA vs Stock market cap. to GDP ρ = 0.5
    - FME vs Stock market cap. to GDP ρ = 0.35

### Summary statistics of indices (selected exact values from Table 5)
- All countries (Var.Obs Mean Median St. Dev. Min Max) preserved:
  - FD6222: 0.23 0.16 0.21 0.00 1.00
  - FI6222: 0.31 0.26 0.23 0.00 1.00
  - FM6222: 0.15 0.03 0.22 0.00 1.00
  - FID6222: 0.20 0.11 0.23 0.00 1.00
  - FIA6222: 0.23 0.12 0.27 0.00 1.00
  - FIE6222: 0.48 0.53 0.23 0.00 1.00
  - FMD6222: 0.14 0.04 0.22 0.00 1.00
  - FMA6222: 0.15 0.00 0.24 0.00 1.00
  - FME6222: 0.15 0.01 0.28 0.00 1.00

### Key empirical findings and country-level patterns (selected)
- Evolution 1980–2013:
  - Financial development progressed noticeably in Advanced Economies (AEs) and Emerging Markets (EMs), and to a lesser extent in Low-Income Developing Countries (LIDCs).
  - The gap between AEs and EMs widened between the mid-1990s and early 2000s, and declined after the global financial crisis.
- Cross-group variation:
  - EMs are closer to AEs in financial markets development than in financial institutions development.
  - Despite lower depth, efficiency of EM and LIDC financial institutions is relatively high.
  - Access is particularly low in LIDCs.
- Country examples and exact scores (select observations preserved):
  - United Kingdom market capitalization outside of top 10 companies in 2013: UK 30 percent; Australia 50 percent; Korea 38 percent.
  - Corporate issuance per 100,000 adults: UK 0.6; Australia 0.9; Korea 1.
  - Stock market turnover: UK 84 percent; Australia 85 percent; Korea 139 percent.
  - Branches and ATMs per 100,000 adults: Trinidad and Tobago 13 and 41; St. Kitts and Nevis 55 and 107.
  - Efficiency (net interest margins and overhead costs): Trinidad and Tobago 5 and 4 percent; St. Kitts and Nevis 0.7 and 1.3 percent.
- Top Financial Development Index (selected top entries preserved from Annex 1):
  - 1 Switzerland 0.951; 2 Australia 0.890; 3 United Kingdom 0.882; 4 United States 0.877; 5 Spain 0.860; 6 Korea, Republic of 0.854; 7 Canada 0.847; 8 Japan 0.827; 9 Hong Kong 0.827; 10 Italy 0.785.
- Financial Markets Depth (top 10 preserved from Annex 3):
  - 1 Sweden 0.996; 2 Canada 0.987; 3 United Kingdom 0.973; 4 United States 0.971; 5 Switzerland 0.970; 6 Spain 0.908; 7 Australia 0.904; 8 Netherlands 0.902; 9 Singapore 0.895; 10 Korea, Republic of 0.890.
- Financial Markets Access top entries preserved:
  - 1 Norway 1.000; 2 Ireland 1.000; 3 Luxembourg 1.000; 4 Switzerland 0.977; 5 Austria 0.908.
- Financial Markets Efficiency top entries preserved:
  - Saudi Arabia 1.000; Turkey 1.000; China, Mainland 1.000; Italy 1.000; Hong Kong 1.000; Korea, Republic of 1.000; Spain 1.000; United States 1.000; Japan 0.950; Germany 0.874.
- Large set of economies record 0.000 in one or more market indices, concentrated among small island states and fragile/post-conflict economies (many examples preserved in Annex listings).

### Robustness, limitations, and cautions
- Data limitations:
  - Some institutions and activities (for example, shadow banks) and some user-side access measures (for example, mobile banking, detailed payments indicators) lack sufficiently long country-time coverage and are excluded.
  - Domestic sovereign debt outstanding volumes excluded due to low BIS reporting coverage.
  - Bid-ask spread in sovereign bond market not used because average coverage is 37 countries starting only in 2000.
- Conceptual caveats:
  - Indices rely on proxy variables where direct measures are unavailable.
  - Indices capture characteristics (depth, access, efficiency) but not underlying drivers (institutional, regulatory, legal frameworks) or outcomes (financial stability measures).
  - Efficiency measures may reflect government controls (e.g., on lending and deposit rates) that could inflate efficiency ratings.
  - Higher FD ranking may not necessarily be positive; could indicate a financial system stretched beyond structural and regulatory capabilities.
- Use cases:
  - Indices intended as summary tools to facilitate empirical work and cross-country comparisons, to be complemented by disaggregated data from FinStats or GFDD for detailed investigations.

*Source: IMF staff paper (excerpt from _wp1605 - References, methodology and introduction sections).*

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

### _wp1605 - References .............................................................................................................

### Introduction: purpose and scope
- Objective: construct indices summarizing how developed financial institutions and financial markets are in terms of depth, access, and efficiency.
- Coverage: 183 advanced, emerging, and low-income developing countries on annual frequency between 1980 and 2013 (33 years).
- Rationale: single indicators (e.g., private credit/GDP, stock market capitalization/GDP) are inadequate because financial development is multidimensional and includes institutions beyond banks (investment banks, insurance companies, mutual funds, pension funds, venture capital firms, etc.), and markets beyond bank lending (stocks, bonds, wholesale money markets).
- Conceptual definition: financial development = combination of depth (size and liquidity of markets), access (ability of individuals and companies to access financial services), and efficiency (ability of institutions to provide financial services at low cost and with sustainable revenues, and the level of activity of capital markets).

### Index construction and methodological framework
- Three-step approach:
  - (i) normalization of variables;
  - (ii) aggregation of normalized variables into sub-indices for each functional dimension; 
  - (iii) aggregation of sub-indices into the final index.
- Methodological reference: OECD Handbook on Constructing Composite Indicators (OECD, 2008).
- Number of indices constructed: nine indices in total.
  - Six lower-level sub-indices: FID, FIA, FIE, FMD, FMA, FME (I = institutions, M = markets; D = depth, A = access, E = efficiency).
  - Two higher-level sub-indices: FI (financial institutions overall) and FM (financial markets overall).
  - Final overall index: FD (financial development index).
- Aggregation choices addressed in the paper: selection of data series; treatment of missing data; normalization and outlier treatment; functional form of aggregator; weights in aggregation.

### Data sources and indicators
- Primary data sources:
  - World Bank FinStats 2015 (Feyen, Kibuuka, and Sourrouille, 2014)
  - IMF’s Financial Access Survey
  - Dealogic corporate debt database
  - Bank for International Settlements (BIS) debt securities database
- Indicator universe and selection:
  - The underlying literature and databases include many indicators (for example, "there are 105 distinct indicators in GFDD and 46 indicators in FinStats"), but the paper selects variables that cover a sufficiently wide range of countries and time.
  - Total set of key indicators chosen is summarized in Table 1 (in the source).
- Coverage constraints and tradeoffs:
  - Some potentially useful indicators were excluded due to insufficient country-time coverage.
  - Example: World Bank Global Financial Inclusion (Global Findex) database provides user-side access data (including mobile banking) but is only available for 2011 and 2014 and thus not usable for the full sub-index time series.

### Indicators used for sub-indices (high-level)
- Financial institutions depth:
  - Bank credit to private sector (standard)
  - Assets of mutual fund and pension fund industries
  - Size of life and non-life insurance premiums (chosen over insurance company assets because it covers more countries)
- Financial institutions access:
  - Number of bank branches per 100,000 adults
  - Number of ATMs per 100,000 adults
  - Other candidate measures (bank accounts per 1,000 adults, percent of firms with line of credit, mobile money usage) were not included due to limited coverage
- Financial institutions efficiency:
  - Net interest margin (bank net interest revenue as a share of average interest-bearing assets)
  - Lending-deposit spread
  - Non-interest income to total income
  - Overhead costs to total assets
  - Return on assets and return on equity
- Financial markets: stock and bond market indicators for depth, access, and efficiency (details and full list in the source tables and annexes).

### Missing data treatment and specific data adjustments
- Missing data treatment:
  - Missing data reconstruction is applied iteratively.
  - The reconstruction no longer uses the data on profit growth.
- Specific adjustments for data breaks ("butt splicing"):
  - Private sector credit data corrected for one-off jumps in banking system coverage in a small number of countries by taking the most recent level as most representative and merging downward shifts through growth rates; cross-checked with IFS notes.
  - Example source-data jumps (illustrative of the problem in original source series):
    - Denmark: credit to GDP jumps from 30 percent of GDP in 1999 to 135 percent in 2000.
    - Sweden: credit to GDP jumps from 40 to 93 percent of GDP in 2001.
  - These corrections do not affect gradual credit buildups during booms (e.g., Thailand in late 1990s, Cyprus and Iceland in 2000s) or jumps in crisis/hyperinflation episodes (e.g., Argentina and Brazil at the end of 1980s and in early 1990s).
- Changes relative to prior release (Staff Discussion Note):
  - Number of issuers (an indicator for financial market access) is now scaled by population to be consistent with financial institutions access measures; this changes relative rankings for some countries (countries with larger populations receive lower scores under the new scaling).
  - Database updated for more recent releases of FinStats and Dealogic data.

### Robustness, comparison with traditional measures, and use cases
- Purpose of indices:
  - Summarize diverse indicators into several easy-to-use indices to allow comprehensive assessment of strengths and deficiencies across depth, access, and efficiency for institutions and markets.
  - Facilitate empirical work by reducing dimensionality from many indicators to a small set of indices.
- Comparisons:
  - The paper shows how the new indices compare with traditional measures of financial development (e.g., private credit to GDP, stock market capitalization/GDP) and documents stylized facts on financial development across the globe.
- Limitations and caveats:
  - Reliance on proxy variables where direct measures are unavailable.
  - Some dimensions (especially nonbank institutions access and certain user-side access measures) are imperfectly captured due to data availability constraints.
  - The indices are intended as summary tools that should be complemented by disaggregated data from FinStats or GFDD for detailed investigations.

*Source: IMF staff paper (excerpt from _wp1605 - References, methodology and introduction sections).*

### 1. Data Sources

### 1. Data Sources

### Category-level indicators and data sources
- Financial Institutions — Depth
  - Private-sector credit to GDP: FinStats 2015
  - Pension fund assets to GDP: FinStats 2015
  - Mutual fund assets to GDP: FinStats 2015
  - Insurance premiums, life and non-life to GDP: FinStats 2015
- Financial Institutions — Access
  - Bank branches per 100,000 adults: FinStats 2015
  - ATMs per 100,000 adults: IMF Financial Access Survey
- Financial Institutions — Efficiency
  - Net interest margin: FinStats 2015
  - Lending-deposits spread: FinStats 2015
  - Non-interest income to total income: FinStats 2015
  - Overhead costs to total assets: FinStats 2015
  - Return on assets: FinStats 2015
  - Return on equity: FinStats 2015
- Financial Markets — Depth
  - Stock market capitalization to GDP: FinStats 2015
  - Stocks traded to GDP: FinStats 2015
  - International debt securities of government to GDP: BIS debt securities database
  - Total debt securities of financial corporations to GDP: Dealogic corporate debt database
  - Total debt securities of nonfinancial corporations to GDP: Dealogic corporate debt database
- Financial Markets — Access
  - Percent of market capitalization outside of top 10 largest companies: FinStats 2015
  - Total number of issuers of debt (domestic and external, nonfinancial and financial corporations): FinStats 2015
- Financial Markets — Efficiency
  - Stock market turnover ratio (stocks traded to capitalization): FinStats 2015

- Source note: "Source: IMF staff estimates."

### Summary statistics of the underlying data (Table 2)
- Financial Institutions — Depth
  - FID1 Private sector credit to GDP: CodeName Obs Mean Median St. Dev. Min Max shown as "5,3284330390.30319"
  - FID2 Pension fund assets to GDP: "942208280.00157"
  - FID3 Mutual fund assets to GDP: "97287105190.005,232"
  - FID4 Insurance premiums (life + non-life) to GDP: "3,3713230.0118"
- Financial Institutions — Access
  - FIA1 Bank branches per 100,000 adults: "1,7221813180.1398"
  - FIA2 ATMs per 100,000 adults: "1,5164028430.01290"
- Financial Institutions — Efficiency
  - FIE1 Net interest margin: "3,3915440.0244"
  - FIE2 Lending-deposits spread: "4,7508680.0392"
  - FIE3 Non-interest income to total income: "3,5273937160.01100"
  - FIE4 Overhead costs to total assets: "3,4194330.0448"
  - FIE5 Return on assets: "3,434113-10921"
  - FIE6 Return on equity: "3,422121445-1,792192"
- Financial Markets — Depth
  - FMD1 Stock market capitalization to GDP: "2,5174526570.00549"
  - FMD2 Stocks traded to GDP: "2,312285580.000756"
  - FMD3 International debt securities of government to GDP: "1,56484100.00398"
  - FMD4 Total debt securities of financial corporation to GDP: "1,7512531030.0001,912"
  - FMD5 Total debt securities of nonfinancial corporation to GDP: "2,229156250.000341"
- Financial Markets — Access
  - FMA1 Percent of market capitalization outside of top 10 largest companies: "669555319  1499"
  - FMA2 Total number of issuers of debt (domestic and external, fin. and non-fin. corporations) per 100,000 adults: "1,8040.30.10.608"
- Financial Markets — Efficiency
  - FME1 Stock market turnover ratio (value traded/stock market capitalization): "2,3134322570.01581"

- Source note: "Source: IMF staff estimates."

### Conceptual choices and excluded indicators
- Efficiency sub-index excludes banking system concentration ratios (Herfindahl index or share of top three banks) for conceptual reasons; literature findings on whether concentration increases or decreases efficiency are mixed (referenced: Čihák and Schaeck, 2010; Berger et al., 2004).
- Financial market depth uses:
  - Stock market size (capitalization) and activity (stocks traded).
  - Outstanding volumes of international sovereign debt securities and corporate debt securities (Dealogic corporate securities data used for corporate debt; corporate data based on nationality principle).
- Excluded data and rationale:
  - Domestic sovereign debt outstanding volumes excluded due to low country coverage in BIS voluntary reporting (18 countries at best).
  - Bid-ask spread in sovereign bond market (Bloomberg) not used because average coverage is 37 countries (20 percent of country sample) starting only in 2000.
  - Variables capturing drivers or outcomes of financial development are excluded (e.g., World Bank Doing Business indicators on ease of getting credit, protection of minority investors, time and cost to enforce contracts, ease of resolving insolvency) as are financial stability indicators (z-scores, capital adequacy or liquidity ratios, frequency of banking crises).
- Financial market access measures:
  - Stock market access proxied by percent of market capitalization outside top 10 largest companies (higher concentration implies greater access difficulties for smaller issuers).
  - Bond market access proxied by number of distinct issuers of debt (domestic and external, financial and nonfinancial) per 100,000 adults; repeat issuance by the same company in a year is counted once.
  - Acknowledged data limitation: scaling issuers by the total pool of potential issuers is preferable but not possible; Dealogic reports only number of companies that issue. World Bank listed domestic companies data cover about 60 percent of the country-year sample; correlation between that indicator and population size is 60 percent, so population size is a relatively good proxy.

### Treatment of missing data
- General tradeoff: broader coverage versus data availability; missing-data extent varies considerably across indicators and over time.
- Data availability is stronger for the most recent twenty years than earlier periods.
- Strong coverage for private credit, debt issuance, and financial institutions efficiency measures.
- Weaker coverage for non-bank financial institutions and other financial markets measures, especially in the LIDC sample.
- Specific missing-data reasons:
  - Financial institutions access measures missing pre-2004 due to lack of comprehensive collection.
  - Market nonexistence: lack of data can indicate that markets are missing (e.g., few LIDCs have domestic stock markets).
- Imputation approach:
  - Where data are not yet available for the latest year (e.g., 2013), values are set equal to the latest available observations (e.g., 2012). Example given for stock market capitalization and stocks traded sourced from World Bank Development Indicators, which are only available until [text truncated in source].

*Source: IMF staff estimates.*

### 2012. If the data series is completely not available for a country, the entire series is set at zero,

### _wp1605 - 2012. If the data series is completely not available for a country, the entire series is set at zero,

### Missing data: cases and treatment options
- Two broad missing-data situations are identified:
  - Entire data series unavailable for a country (set to zero, indicating market does not exist or access/efficiency very poor).
  - Staggered start of data collection across series (e.g., credit to GDP available from the 1960s; financial access data start in 2004).
- Three treatment options for the staggered-start case:
  - (i) Treat the data as truly missing: exclude the series from the index average when unavailable.
  - (ii) Treat the data as zero: assume absence of data implies market does not exist or is very poor.
  - (iii) Splice indices before and after the series becomes available.

### Splicing: rationale and implementation
- Splicing is preferred to avoid index movements driven solely by addition of new series (Figures 2 and 3 examples).
- Conceptual rationale:
  - Prevents artificial drops or jumps in aggregate FD index when a series is introduced late.
  - Allows informed judgment whether missing data imply non-existence of a market or simply lack of historical measurement.
- Practical implementation in the dataset:
  - Splicing is applied at the level of raw data.
  - Identify series with missing earlier-year observations.
  - Fill missing data retrospectively, starting from the first available observation and applying the average growth rate of other indicators with data available for previous years.
  - Iterative procedure uses growth rates in this order:
    - similar series within the same sub-index (e.g., financial institutions depth),
    - series for the same type of provider (e.g., financial institutions),
    - across providers (from financial institutions to financial markets).
  - Two profit indicators are excluded from the splicing procedure because they span negative and positive ranges and would overstate movement.
- Scope of reconstruction in practice:
  - About 27 percent of the sample is reconstructed through splicing.
  - 32 percent of the sample consists of “missing” markets (Figure 4).
- Noted exception where splicing might be inappropriate:
  - A rare “big bang” financial development event that creates a sustained, large market immediately (example: typical first-time sovereign issuance averaged four percent of GDP in the last ten years).

### Normalization and treatment of outliers
- Winsorization:
  - Each series is winsorized with the 5th and 95th percentiles set as cutoff levels to prevent extreme values from distorting 0-1 indicators.
  - Global distribution (across countries and time) is used to determine cutoff levels.
- Min-max normalization:
  - Winsorized indicators are normalized between 0 and 1 using min-max procedures (equations 1 and 2).
  - Highest (lowest) value across time and countries is set to one (zero); other values measured relative to these extremes.
  - For series where higher values indicate worse performance (net interest margin, lending-deposits spread, noninterest income to total income, overhead costs to total assets), ratings are rescaled so that higher normalized values indicate greater financial development.
- Note: The Human Development Index is cited as an example of min-max normalization; alternative normalization methods are noted (standardization, min-max, distance to a reference point).

### Functional form of the aggregator: linear versus geometric
- Aggregation approach used:
  - Weighted linear average for sub-indices (weights from principal component analysis).
  - Sub-indices re-normalized to range [0, 1]; higher-level indices built similarly, culminating in the FD index.
- Rationale for linear aggregation:
  - Best suited for data with a significant share of zero or near-zero observations.
  - Assumes full compensability (perfect substitutes): poor performance on some indicators can be offset by high values on others.
  - Simpler to implement and interpret; indicator contribution to FD index is its weight.
- Geometric aggregation considerations:
  - Allows imperfect substitutability; would penalize unequal indicator distributions.
  - In this dataset, geometric averaging introduces substantial zero bias because zero/near-zero indicator ratings multiplicatively drive the index toward zero.
  - Example consequences:
    - Luxembourg’s FMD score drops from 0.75 to 0.25 under geometric averaging; one indicator received a normalization rating of 0.003.
    - Luxembourg’s ranking falls by 29 places under geometric aggregation.
  - Geometric averaging would require different normalization (distance to a reference point), replacements for zero raw data with minima, and shifting ROA/ROE into nonnegative territory.

### Weights: principal component analysis (PCA)
- Weighting method:
  - PCA used to derive weights so as not to prejudge indicator importance.
  - Sub-index weights are squared factor loadings (sum to 1) from PCA on underlying series.
  - Factor loadings on the first principal component are chosen as weights.
- Interpretation and empirical properties:
  - First principal component interpreted to summarize latent information on degree of financial development.
  - First principal component embodies between 51 and 92 percent of the variance in the sub-index data (Table 4).
  - PCA pooling done across all countries (LIDC, EM, AM) and all years (1980-2013).
  - Example weight magnitudes:
    - Banking system credit to the private sector has a weight of 0.25 within the depth subcomponent of FI.
    - FI subcomponent weight for that depth subcomponent is less than 0.40 in FI overall.
- Purpose of weighting:
  - Correct for overlapping information between correlated indicators; weights are not measures of theoretical importance.

### Complete procedure summary (step-by-step)
- (i) Apply missing data treatment to actual data (including splicing where appropriate).
- (ii) Winsorize indicators: set 5th and 95th percentiles at cutoff levels.
- (iii) Normalize indicators to 0-1 via min-max procedure (higher value indicates greater financial depth after rescaling where needed).
- (iv) Construct sub-indices as weighted averages of normalized series, with weights equal to squared factor loadings (sum to 1) from PCA on the underlying series.
- (v) Combine sub-indices into higher-level indices via the same PCA-based weighted averaging and re-normalize.
- Final output:
  - Relative ranking of countries on depth, access, and efficiency of financial institutions and financial markets, on the development of financial institutions and markets (FI, FM), and on the overall level of financial development (FD).
  - Dataset span referenced for PCA pooling: years 1980-2013.

*Source: IMF staff estimates.*

### 2013. Financial market development is low in Africa, and more advanced in Russia and China. See

### _wp1605 - 2013. Financial market development is low in Africa, and more advanced in Russia and China. See

### Indices and measurement approach
- The Financial Development (FD) indices incorporate information on a broader range of financial development features for a wider array of financial agents compared with traditional measures (private credit to GDP and stock market capitalization to GDP).
- The indices are correlated with traditional measures but the correlation is not one for one; the indices contain more information (Figures 10 and 11).

### Principal component weights (sub-index structure)
- Reported principal component (PC) contributions by sub-index labels (as presented):
  - PC1: DepthAccess Efficiency DepthAccess Efficiency FI FM FD PC 1 values include: 0.7001 0.8824 0.5364 0.5896 0.6698 ... 0.6749 0.7685 0.8595
  - PC2: 0.1288 0.1176 0.2676 0.1937 0.3302 0.2180 0.1523 0.1405
  - PC3: 0.0983 0.0949 0.1007 0.1071 0.0792
  - PC4: 0.0728 0.070 0.0752
  - PC5: 0.0181 0.0408
  - PC6: 0.013
- Sub-index grouping: Financial Institutions, Financial Markets, and sub-components Depth, Access, Efficiency.

### Summary statistics of the Financial Development Index (Table 5)
- Variables and summary stats (Var.Obs Mean Median St. Dev. Min Max) as reported (values preserved exactly):
  - FD6222: 0.23 0.16 0.21 0.00 1.00
  - FI6222: 0.31 0.26 0.23 0.00 1.00
  - FM6222: 0.15 0.03 0.22 0.00 1.00
  - FID6222: 0.20 0.11 0.23 0.00 1.00
  - FIA6222: 0.23 0.12 0.27 0.00 1.00
  - FIE6222: 0.48 0.53 0.23 0.00 1.00
  - FMD6222: 0.14 0.04 0.22 0.00 1.00
  - FMA6222: 0.15 0.00 0.24 0.00 1.00
  - FME6222: 0.15 0.01 0.28 0.00 1.00
  - FD3026: 0.23 0.21 0.17 0.00 0.85
  - FI3026: 0.30 0.29 0.19 0.00 0.87
  - FM3026: 0.15 0.07 0.19 0.00 0.90
  - FID3026: 0.18 0.13 0.18 0.00 0.99
  - FIA3026: 0.23 0.17 0.22 0.00 1.00
  - FIE3026: 0.47 0.54 0.25 0.00 0.95
  - FMD3026: 0.13 0.05 0.18 0.00 0.90
  - FMA3026: 0.16 0.04 0.21 0.00 1.00
  - FME3026: 0.16 0.03 0.26 0.00 1.00
  - FD2312: 0.11 0.10 0.07 0.00 0.39
  - FI2312: 0.18 0.18 0.12 0.00 0.61
  - FM2312: 0.03 0.00 0.07 0.00 0.52
  - FID2312: 0.07 0.05 0.08 0.00 0.50
  - FIA2312: 0.08 0.03 0.14 0.00 1.00
  - FIE2312: 0.42 0.47 0.22 0.00 1.00
  - FMD2312: 0.03 0.01 0.07 0.00 0.50
  - FMA2312: 0.01 0.00 0.05 0.00 0.50
  - FME2312: 0.04 0.00 0.16 0.00 1.00
- Note labels preserved: FD = financial development; FI = financial institutions; FM = financial markets; FID = financial institutions depth; FIA = financial institutions access; FIE = financial institutions efficiency; FMD = financial markets depth; FMA = financial markets access; FME = financial markets efficiency.
- Group labels in table: Emerging Markets, Low-Income and Developing Countries, All countries, Advanced Markets (as presented).

### Correlations with traditional measures (Figures 10 and 11)
- Correlation with Private Credit to GDP (ρ values preserved):
  - FD vs Private credit to GDP ρ = 0.81
  - FI vs Private credit to GDP ρ = 0.82
  - FM vs Private credit to GDP ρ = 0.7
  - FID vs Private credit to GDP ρ = 0.84
  - FIA vs Private credit to GDP ρ = 0.66
  - FIE vs Private credit to GDP ρ = 0.37
  - FMD vs Private credit to GDP ρ = 0.7
  - FMA vs Private credit to GDP ρ = 0.63
  - FME vs Private credit to GDP ρ = 0.51
- Correlation with Stock Market Capitalization/GDP (ρ values preserved):
  - FD vs Stock market cap. to GDP ρ = 0.62
  - FI vs Stock market cap. to GDP ρ = 0.52
  - FM vs Stock market cap. to GDP ρ = 0.63
  - FID vs Stock market cap. to GDP ρ = 0.61
  - FIA vs Stock market cap. to GDP ρ = 0.30
  - FIE vs Stock market cap. to GDP ρ = 0.34
  - FMD vs Stock market cap. to GDP ρ = 0.74
  - FMA vs Stock market cap. to GDP ρ = 0.5
  - FME vs Stock market cap. to GDP ρ = 0.35

### Country examples and interpretation of rankings
- Advanced market comparisons (2013, Annex 1 data cited):
  - United Kingdom has the deepest financial markets among the four-country example but ranks lower on financial market access and efficiency versus Korea and Australia.
    - Market capitalization outside of top 10 companies in 2013: UK 30 percent, Australia 50 percent, Korea 38 percent.
    - Corporate issuance per 100,000 adults: UK 0.6, Australia 0.9, Korea 1.
    - Stock market turnover: UK 84 percent, Australia 85 percent, Korea 139 percent.
  - Hong Kong: ranks highly on financial market efficiency but overall FM indicator is reduced by lower depth and access.
- Caribbean example:
  - Trinidad and Tobago vs St. Kitts and Nevis:
    - Branches and ATMs per 100,000 adults: Trinidad and Tobago 13 and 41; St. Kitts and Nevis 55 and 107.
    - Efficiency (net interest margins and overhead costs): Trinidad and Tobago 5 and 4 percent; St. Kitts and Nevis 0.7 and 1.3 percent.
  - Result: Trinidad and Tobago receives a lower FD rating compared to St. Kitts and Nevis due to lower ratings on financial institutions access and efficiency despite larger institutions.

### Evolution of FD over time (1980–2013)
- Broad patterns (Figure 12):
  - Financial development progressed noticeably in Advanced Economies (AEs) and Emerging Markets (EMs), and to a lesser extent in Low-Income Developing Countries (LIDCs).
  - The gap between AEs and EMs widened between the mid-1990s and early 2000s, reflecting rapid growth in AEs’ financial systems during the "Greenspan Era."
  - The gap between AEs and EMs declined after the global financial crisis, reflecting deleveraging in AEs.
- Country-group average FD index axis preserved as shown (0.0 to 1.0 scale across 1980–2013).

### Cross-group and within-group variation (Figures 13–14)
- Peer group averages show:
  - EMs are closer to AEs in financial markets development than in financial institutions development.
  - Despite lower depth, efficiency of EM and LIDC financial institutions is relatively high.
  - Access is particularly low in LIDCs.
- Selected country 2013 FD values (Figure 14 and Annex 1 highlights):
  - Examples where EMs surpass some AEs: Malaysia, Brazil, South Africa have higher FD than New Zealand and Greece.
  - Examples where some EMs fall below some LIDCs: Tunisia and Armenia have lower FD than Mongolia and Bangladesh.
- Annex 1 top FD rankings (first 46 entries as presented; first three columns are Financial Development Index, Financial Institutions Index, Financial Markets Index with values):
  - Financial Development Index top entries: 1 Switzerland 0.951; 2 Australia 0.890; 3 United Kingdom 0.882; 4 United States 0.877; 5 Spain 0.860; 6 Korea, Republic of 0.854; 7 Canada 0.847; 8 Japan 0.827; 9 Hong Kong 0.827; 10 Italy 0.785; ... (list continues as shown in Annex 1).
  - Financial Institutions Index top entries: 1 Switzerland 1.000; 2 Luxembourg 0.893; 3 France 0.892; 4 United Kingdom 0.892; 5 Canada 0.890; ... (as presented).
  - Financial Markets Index top entries: 1 United States 0.903; 2 Korea, Republic of 0.902; 3 Switzerland 0.883; 4 Australia 0.873; 5 Hong Kong 0.869; ... (as presented).

### Caveats and limitations
- Data coverage limitations:
  - Not possible to find sufficiently extensive country and time period data on some institutions and activities (example: shadow banks).
  - Indicators for different forms of financial payments (credit transfers, direct debits, mobile banking) are not available on a sufficiently long time horizon for inclusion.
  - Diversity of financial intermediaries and organizational complexity are not incorporated.
- Conceptual caveats:
  - The FD index captures characteristics of financial systems (depth, access, efficiency) but does not include underlying drivers (institutional, regulatory, legal frameworks) or outcomes (financial stability measures).
  - Some measures may overstate true financial development; efficiency measures could reflect government controls (e.g., on lending and deposit rates) that may inflate efficiency ratings.
  - Higher FD ranking may not necessarily be positive; it may indicate that a financial system is stretched beyond structural and regulatory capabilities, with potential negative implications for growth and stability.
- The FD indices represent progress toward more comprehensive measurement and will be improved as new information becomes available.

*Source: IMF staff estimates and IMF staff calculations (as presented in the source PDF).*

### Annex 1. 2013 Country Rankings on Financial Development (ctd.)

### Annex 1. 2013 Country Rankings on Financial Development (ctd.)

### Financial Development Index / Financial Institutions Index / Financial Markets Index (selected rows as presented)
- 47  Jordan0.414 47  Slovak Republic0.547 47  Barbados0.328
- 48  Peru0.410 48  Estonia0.546 48  Jordan0.312
- 49  Croatia0.406 49  Panama0.539 49  Bahrain0.311
- 50  Mexico0.396 50  Grenada0.538 50  Peru0.288
- 51  India0.392 51  St. Lucia0.536 51  Egypt0.281
- 52  Morocco0.390 52  Lebanon0.535 52  Kazakhstan0.267
- 53  Mauritius0.389 53  Czech Republic0.533 53  Slovenia0.267
- 54  Bulgaria0.380 54  Barbados0.532 54  Indonesia0.259
- 55  St. Kitts and Nevis0.366 55  Morocco0.528 55  Moldova0.250
- 56  Philippines0.365 56  Peru0.524 56  Oman0.249
- 57  Czech Republic0.360 57  China, Mainland0.511 57  Morocco0.243
- 58  Panama0.342 58  Jordan0.509 58  Argentina0.225
- 59  Brunei Darussalam0.336 59  Costa Rica0.503 59  Bahamas, The0.216
- 60  Mongolia0.335 60  Latvia0.499 60  Bangladesh0.213
- 61  Estonia0.330 61  Dominica0.491 61  Mauritius0.208
- 62  Trinidad & Tobago0.328 62  Lithuania0.491 62  Jamaica0.187
- 63  Indonesia0.322 63  Ecuador0.489 63  Sri Lanka0.185
- 64  Lebanon0.321 64  Namibia0.488 64  Iran, I. Rep. Of0.182
- 65  Argentina0.314 65  Trinidad & Tobago0.488 65  Brunei Darussalam0.181
- 66  Slovak Republic0.314 66  Brunei Darussalam0.485 66  Czech Republic0.181
- 67  Kuwait0.313 67  Hungary0.484 67  Kuwait0.174
- 68  Antigua & Barbuda0.312 68  Cape Verde0.480 68  Trinidad & Tobago0.161
- 69  Kazakhstan0.311 69  Turkey0.474 69  Papua New Guinea0.155
- 70  Bahrain0.304 70  Macedonia, FYR0.468 70  Cote D'Ivoire0.138
- 71  Latvia0.298 71  Bosnia and Herzegovina0.464 71  Panama0.138
- 72  Oman0.297 72  U.A.E.0.449 72  Pakistan0.129
- 73  Moldova0.297 73  Kuwait0.447 73  Croatia0.120
- 74  Seychelles0.295 74  Qatar0.446 74  Estonia0.107
- 75  St. Lucia0.288 75  Mexico0.443 75  Mongolia0.105
- 76  Costa Rica0.284 76  Guatemala0.443 76  Vietnam0.103
- 77  Egypt0.280 77  Belize0.436 77  Lebanon0.101
- 78  Lithuania0.273 78  Ukraine0.429 78  Botswana0.091
- 79  Grenada0.272 79  Venezuela0.426 79  Latvia0.090
- 80  Sri Lanka0.270 80  Georgia0.426 80  Liberia0.088
- 81  Namibia0.269 81  Vanuatu0.419 81  Laos0.088
- 82  Ecuador0.258 82  El Salvador0.417 82  Burundi0.085
- 83  Ukraine0.257 83  Armenia0.416 83  St. Kitts and Nevis0.081
- 84  Bangladesh0.256 84  Fiji0.411 84  Ukraine0.080
- 85  Venezuela0.255 85  St. Vincent and the Gren0.402 85  Uzbekistan0.079
- 86  Macedonia, FYR0.251 86  Uruguay0.402 86  Venezuela0.079
- 87  Iran, I. Rep. Of0.249 87  Tunisia0.400 87  Slovak Republic0.074
- 88  Dominica0.248 88  Argentina0.398 88  Tunisia0.074
- 89  El Salvador0.247 89  Saudi Arabia0.396 89  El Salvador0.071
- 90  Guatemala0.244 90  Albania0.393 90  Kenya0.071
- 91  Cape Verde0.243 91  Suriname0.390 91  Bulgaria0.071
- 92  Georgia0.239 92  Macao SAR, China0.388 92  Honduras0.065

### Financial Development Index (lower ranks, selected)
- 93  Tunisia0.239
- 94  Jamaica0.238
- 95  Vietnam0.236
- 96  Bosnia and Herzegovina0.236
- 97  Uruguay0.231
- 98  Belize0.223
- 99  Armenia0.220
- 100  Botswana0.219
- 101  Honduras0.217
- 102  Fiji0.216
- 103  Vanuatu0.212
- 104  Bolivia0.211
- 105  Romania0.205
- 106  St. Vincent and the Gren0.203
- 107  Albania0.200
- 108  Pakistan0.197
- 109  Uzbekistan0.197
- 110  Suriname0.197
- 111  Macao SAR, China0.196
- 112  Serbia0.190
- 113  Bhutan0.189
- 114  Papua New Guinea0.187
- 115  Kenya0.187
- 116  Samoa0.180
- 117  Nepal0.176
- 118  Paraguay0.171
- 119  Cote D'Ivoire0.168
- 120  Azerbaijan0.160
- 121  Maldives0.159
- 122  Dominican Republic0.158
- 123  Laos0.158
- 124  Aruba0.157
- 125  Guyana0.154
- 126  Tonga0.152
- 127  Belarus0.151
- 128  Angola0.151
- 129  Djibouti0.148
- 130  Swaziland0.146
- 131  Nigeria0.138
- 132  Libya0.136
- 133  Lesotho0.136
- 134  Gabon0.133
- 135  Nicaragua0.129
- 136  Mozambique0.128
- 137  Algeria0.128
- 138  Zambia0.128

### Financial Institutions Index (selected lower and bottom rows)
- 100  Romania0.346
- 101  Bhutan0.346
- 102  India0.344
- 103  Botswana0.342
- 104  Philippines0.342
- 105  Oman0.340
- 106  Moldova0.338
- 107  Serbia0.334
- 108  Nepal0.323
- 109  Maldives0.314
- 110  Iran, I. Rep. Of0.312
- 111  Aruba0.312
- 112  Uzbekistan0.310
- 113  Tonga0.301
- 114  Kenya0.299
- 115  Paraguay0.298
- 116  Bangladesh0.294
- 117  Dominican Republic0.293
- 118  Bahrain0.291
- 119  Belarus0.289
- 120  Azerbaijan0.287
- 121  Guyana0.283
- 122  Jamaica0.283
- 123  Swaziland0.282
- 124  Angola0.276
- 125  Egypt0.274
- 126  Libya0.270
- 127  Lesotho0.268
- 128  Pakistan0.261
- 129  Algeria0.251
- 130  Djibouti0.251
- 131  Nicaragua0.250
- 132  Gabon0.247
- 133  Sao Tome and Principe0.238
- 134  Micronesia, Fed. Sts.0.234
- 135  Syria0.230
- 136  Cambodia0.229
- 137  Nigeria0.225
- 138  Kyrgyz Republic0.225

### Financial Markets Index (selected entries and bottom zeros)
- 93  Costa Rica0.061
- 94  Romania0.059
- 95  Uruguay0.055
- 96  Lithuania0.051
- 97  Nigeria0.049
- 98  Georgia0.048
- 99  Namibia0.043
- 100  Djibouti0.043
- 101  Serbia0.042
- 102  Paraguay0.041
- 103  Guatemala0.040
- 104  Mozambique0.036
- 105  Uganda0.036
- 106  Zambia0.035
- 107  St. Lucia0.034
- 108  Bolivia0.031
- 109  Azerbaijan0.031
- 110  Macedonia, FYR0.030
- 111  Bhutan0.029
- 112  Ghana0.029
- 113  Nepal0.026
- 114  Turkmenistan0.025
- 115  Angola0.023
- 116  Ecuador0.022
- 117  Malawi0.022
- 118  Guyana0.021
- 119  Kyrgyz Republic0.021
- 120  Dominican Republic0.020
- 121  Armenia0.020
- 122  Yemen0.018
- 123  Niger0.018
- 124  Tanzania0.016
- 125  Gabon0.016
- 126  Fiji0.016
- 127  Ethiopia0.015
- 128  Mauritania0.011
- 129  Seychelles0.011
- 130  Belarus0.010
- 131  Cambodia0.010
- 132  Chad0.010
- 133  Madagascar0.009
- 134  Swaziland0.006
- 135  Senegal0.006
- 136  Belize0.006
- 137  Sierra Leone0.006
- 138  Nicaragua0.006
- 139  Burkina Faso0.005
- 140  Guinea0.004
- 141  Cameroon0.003
- 142  Togo0.003
- 143  Bosnia and Herzegovina0.003
- 144  Algeria0.002
- 145  Lesotho0.002
- 146  Rwanda0.002
- 147  Albania0.002
- 148  Cape Verde0.002
- 149  Sudan0.000
- 150  Tajikistan0.000
- 151  Libya0.000
- 152  Mali0.000
- 153  Congo, Dem. Rep. of0.000
- 154  French Polynesia0.000
- 155  South Sudan0.000
- 156  Guinea-Bissau0.000
- 157  Timor Leste0.000
- 158  Comoros0.000
- 159  Equatorial Guinea0.000
- 160  Marshall Islands0.000
- 161  C.A.R.0.000
- 162  Congo, Republic of0.000
- 163  Eritrea0.000
- 164  Haiti0.000
- 165  Gambia, The0.000
- 166  Solomon Islands0.000
- 167  Kiribati0.000
- 168  Myanmar0.000
- 169  Syria0.000
- 170  Micronesia, Fed. Sts.0.000
- 171  Sao Tome and Principe0.000
- 172  Tonga0.000
- 173  Aruba0.000
- 174  Maldives0.000
- 175  Samoa0.000
- 176  Macao SAR, China0.000
- 177  Suriname0.000
- 178  St. Vincent and the Gren0.000
- 179  Vanuatu0.000
- 180  Dominica0.000
- 181  Grenada0.000
- 182  Antigua & Barbuda0.000
- 183  Antigua & Barbuda0.000

*Source: IMF staff estimates.*

---

### Annex 2. 2013 Country Rankings on Financial Institutions Depth, Access, Efficiency

### Financial Institutions Depth / Access / Efficiency (top ranks)
- 1  Ireland1.000 1  St. Kitts and Nevis1.000 1  Greece0.784
- 2  Denmark1.000 2  Brazil1.000 2  New Zealand0.751
- 3  United Kingdom1.000 3  Portugal1.000 3  Japan0.749
- 4  Switzerland1.000 4  Spain1.000 4  China, Mainland0.747
- 5  Hong Kong0.979 5  Luxembourg0.994 5  Australia0.735
- 6  Singapore0.945 6  Switzerland0.990 6  Qatar0.729
- 7  Canada0.929 7  Bulgaria0.954 7  Malaysia0.726
- 8  Malaysia0.894 8  Italy0.948 8  Sweden0.723
- 9  South Africa0.890 9  France0.922 9  Estonia0.717
- 10  United States0.817 10  Russian Federation0.919 10  Bahrain0.712
- 11  Australia0.813 11  Belgium0.913 11  Korea, Republic of0.711
- 12  Sweden0.802 12  Bahamas, The0.878 12  Malta0.710
- 13  France0.777 13  Croatia0.877 13  Norway0.703
- 14  Japan0.773 14  United States0.870 14  Kuwait0.699
- 15  Netherlands0.746 15  Japan0.869 15  Barbados0.697
- 16  Germany0.741 16  Slovenia0.867 16  Finland0.694
- 17  Korea, Republic of0.724 17  Iceland0.836 17  Oman0.693
- 18  Austria0.707 18  Australia0.835 18  Singapore0.692
- 19  Luxembourg0.699 19  Seychelles0.823 19  Libya0.691
- 20  Iceland0.680 20  Antigua & Barbuda0.806 20  Spain0.690

### Financial Institutions Depth (selected mid and lower ranks)
- 21  Malta0.680
- 22  Israel0.678
- 23  Belgium0.674
- 24  Finland0.658
- 25  Chile0.638
- 26  Spain0.629
- 27  Italy0.622
- 28  New Zealand0.612
- 29  Portugal0.604
- 30  Brazil0.585
- 31  Norway0.573
- 32  Cyprus0.555
- 33  Thailand0.515
- 34  Morocco0.471
- 35  St. Lucia0.470
- 36  Mauritius0.467
- 37  Trinidad & Tobago0.461
- 38  Bahamas, The0.432
- 39  Barbados0.422
- 40  China, Mainland0.413
- 41  Jordan0.399
- 42  Croatia0.379
- 43  Poland0.373
- 44  Namibia0.370
- 45  Greece0.366
- 46  Hungary0.365
- 47  Slovenia0.359
- 48  Antigua & Barbuda0.350
- 49  Grenada0.347
- 50  El Salvador0.320

### Financial Institutions Access (selected)
- 1  St. Kitts and Nevis1.000
- 2  Brazil1.000
- 3  Portugal1.000
- 4  Spain1.000
- 5  Luxembourg0.994
- 6  Switzerland0.990
- 7  Bulgaria0.954
- 8  Italy0.948
- 9  France0.922
- 10  Russian Federation0.919
- 11  Belgium0.913
- 12  Bahamas, The0.878
- 13  Croatia0.877
- 14  United States0.870
- 15  Japan0.869
- 16  Slovenia0.867
- 17  Iceland0.836
- 18  Australia0.835
- 19  Seychelles0.823
- 20  Antigua & Barbuda0.806

### Financial Institutions Efficiency (selected top and mid)
- 1  Greece0.784
- 2  New Zealand0.751
- 3  Japan0.749
- 4  China, Mainland0.747
- 5  Australia0.735
- 6  Qatar0.729
- 7  Malaysia0.726
- 8  Sweden0.723
- 9  Estonia0.717
- 10  Bahrain0.712
- 11  Korea, Republic of0.711
- 12  Malta0.710
- 13  Norway0.703
- 14  Kuwait0.699
- 15  Barbados0.697
- 16  Finland0.694
- 17  Oman0.693
- 18  Singapore0.692
- 19  Libya0.691
- 20  Spain0.690

### Financial Institutions Depth / Access / Efficiency (selected lower and bottom rows)
- 100  Togo0.143 100  China, Mainland0.316 100  Bosnia and Herzegovina0.541
- 101  Aruba0.142 101  Finland0.313 101  St. Lucia0.538
- 102  Saudi Arabia0.137 102  Maldives0.308 102  Romania0.536
- 103  Seychelles0.136 103  Moldova0.297 103  India0.534
- 104  Ecuador0.135 104  Dominican Republic0.289 104  Ireland0.530
- 105  Swaziland0.133 105  Bolivia0.285 105  Maldives0.529
- 106  Malawi0.133 106  Sri Lanka0.284 106  Hungary0.529
- 107  Dominican Republic0.133 107  Bhutan0.278 107  Swaziland0.528
- 108  Oman0.131 108  Azerbaijan0.276 108  Papua New Guinea0.528
- 109  Moldova0.131 109  El Salvador0.273 109  Laos0.523
- 110  Armenia0.123 110  Belarus0.266 110  Congo, Dem. Rep. of0.523
- 111  Djibouti0.123 111  Botswana0.246 111  Costa Rica0.522
- 112  Qatar0.122 112  Angola0.244 112  Slovenia0.521
- 113  Albania0.118 113  Marshall Islands0.241 113  St. Vincent and the Gren0.516
- 114  Mozambique0.112 114  Micronesia, Fed. Sts.0.238 114  Turkey0.515
- 115  Nicaragua0.111 115  Paraguay0.226 115  Cambodia0.514
- 116  Cambodia0.111 116  Oman0.221 116  Benin0.513
- 117  Samoa0.110 117  Jamaica0.213 117  Ecuador0.511
- 118  Senegal0.110 118  Aruba0.213 118  Gabon0.510
- 119  Guyana0.109 119  Philippines0.207 119  Kiribati0.509
- 120  Sri Lanka0.109 120  India0.198 120  Angola0.508
- 121  Georgia0.109 121  Swaziland0.191 121  Belarus0.507
- 122  Guatemala0.107 122  Kyrgyz Republic0.190 122  Lesotho0.506
- 123  Brunei Darussalam0.105 123  Gabon0.188 123  Senegal0.504
- 124  Bhutan0.102 124  Guyana0.186 124  Azerbaijan0.504
- 125  Papua New Guinea0.100 125  Nicaragua0.156 125  Kenya0.502
- 126  Zambia0.099 126  Vietnam0.150 126  Chad0.501
- 127  Maldives0.095 127  Libya0.147 127  Antigua & Barbuda0.500
- 128  Egypt0.094 128  Kiribati0.136 128  Nicaragua0.499
- 129  Cote D'Ivoire0.093 129  Nepal0.135 129  Peru0.496
- 130  Belarus0.089 130  Pakistan0.134 130  Grenada0.495
- 131  Nigeria0.086 131  Solomon Islands0.133 131  Bolivia0.494
- 132  Mauritania0.086 132  Nigeria0.132 132  Mozambique0.491
- 133  Iran, I. Rep. Of0.082 133  Tajikistan0.125 133  Paraguay0.488
- 134  Burkina Faso0.081 134  Equatorial Guinea0.121 134  Eritrea0.486
- 135  Benin0.081 135  Bangladesh0.120 135  Micronesia, Fed. Sts.0.485
- 136  Angola0.079 136  Laos0.117 136  Tanzania0.484
- 137  Ghana0.077 137  Egypt0.110 137  Argentina0.484
- 138  Azerbaijan0.077 138  Kenya0.110 138  Ukraine0.484

### Financial Institutions Depth / Access / Efficiency (bottom rows and zeros)
- 139  Laos0.071 139  Ghana0.106 139  Zambia0.484
- 140  Sao Tome and Principe0.069 140  Zambia0.102 140  Samoa0.482
- 141  Tonga0.068 141  Gambia, The0.096 141  Nigeria0.481
- 142  Mali0.068 142  Lesotho0.094 142  Mali0.481
- 143  Tanzania0.066 143  Cambodia0.093 143  Brazil0.479
- 144  Cameroon0.066 144  Rwanda0.091 144  Mauritania0.479
- 145  Syria0.065 145  Algeria0.088 145  United States0.479
- 146  Pakistan0.064 146  Cote D'Ivoire0.082 146  Sudan0.473
- 147  Gabon0.063 147  Djibouti0.082 147  Haiti0.471
- 148  Algeria0.060 148  Mali0.082 148  Kyrgyz Republic0.461
- 149  Ethiopia0.060 149  Mozambique0.081 149  Belize0.461
- 150  Congo, Republic of0.059 150  Syria0.078 150  Ghana0.460
- 151  Macao SAR, China0.057 151  Togo0.076 151  Niger0.459
- 152  Uganda0.057 152  Senegal0.076 152  Honduras0.457
- 153  Haiti0.056 153  Timor Leste0.071 153  Guinea0.448
- 154  Gambia, The0.055 154  Papua New Guinea0.062 154  Tonga0.436
- 155  Rwanda0.054 155  Malawi0.059 155  Cote D'Ivoire0.434
- 156  Solomon Islands0.052 156  Congo, Republic of0.058 156  Togo0.430
- 157  Tajikistan0.052 157  Benin0.058 157  Kazakhstan0.426
- 158  Niger0.051 158  Tanzania0.056 158  Dominican Republic0.422
- 159  Comoros0.051 159  Uganda0.055 159  Solomon Islands0.416
- 160  Madagascar0.049 160  Sudan0.054 160  Madagascar0.412
- 161  Eritrea0.047 161  Liberia0.053 161  Burundi0.395
- 162  Burundi0.044 162  Comoros0.052 162  Uzbekistan0.395
- 163  C.A.R.0.043 163  Mauritania0.050 163  Uganda0.384
- 164  Kyrgyz Republic0.042 164  Guinea-Bissau0.048 164  Gambia, The0.381
- 165  Sudan0.038 165  Yemen0.044 165  Jamaica0.372
- 166  Libya0.037 166  Burundi0.041 166  Russian Federation0.345
- 167  Guinea-Bissau0.031 167  Burkina Faso0.040 167  Rwanda0.344
- 168  Equatorial Guinea0.026 168  Cameroon0.037 168  Sierra Leone0.344
- 169  Myanmar0.024 169  Ethiopia0.034 169  Liberia0.336
- 170  Sierra Leone0.023 170  Sierra Leone0.034 170  Tajikistan0.335
- 171  Chad0.023 171  Myanmar0.031 171  C.A.R.0.296
- 172  Guinea0.021 172  Haiti0.030 172  Malawi0.293
- 173  Yemen0.021 173  Madagascar0.028 173  Congo, Republic of0.242
- 174  Congo, Dem. Rep. of0.019 174  Guinea0.026 174  Sao Tome and Principe0.222
- 175  Liberia0.012 175  Niger0.024 175  Iran, I. Rep. Of0.175
- 176  Uzbekistan0.011 176  South Sudan0.019 176  Comoros0.093
- 177  Turkmenistan0.009 177  C.A.R.0.016 177  Serbia0.089
- 178  South Sudan0.003 178  Congo, Dem. Rep. of0.011 178  Equatorial Guinea0.074
- 179  French Polynesia0.000 179  Chad0.011 179  Timor Leste0.067
- 180  Timor Leste0.000 180  French Polynesia0.000 180  South Sudan0.055
- 181  Marshall Islands0.000 181  Eritrea0.000 181  French Polynesia0.000
- 182  Kiribati0.000 182  Turkmenistan0.000 182  Guinea-Bissau0.000
- 183  Micronesia, Fed. Sts.0.000 183  Bahrain0.000 183  Marshall Islands0.000

*Source: IMF staff estimates.*

### Annex 3. 2013 Country Rankings on Financial Markets Depth, Access, Efficiency

### Annex 3. 2013 Country Rankings on Financial Markets Depth, Access, Efficiency

### Top-ranked countries (selected highlights)
- Financial Markets Depth (top 10):
  - 1 Sweden 0.996
  - 2 Canada 0.987
  - 3 United Kingdom 0.973
  - 4 United States 0.971
  - 5 Switzerland 0.970
  - 6 Spain 0.908
  - 7 Australia 0.904
  - 8 Netherlands 0.902
  - 9 Singapore 0.895
  - 10 Korea, Republic of 0.890
- Financial Markets Access (top 10):
  - 1 Norway 1.000
  - 2 Ireland 1.000
  - 3 Luxembourg 1.000
  - 4 Switzerland 0.977
  - 5 Austria 0.908
  - 6 Australia 0.835
  - 7 Malta 0.832
  - 8 U.A.E. 0.764
  - 9 Korea, Republic of 0.754
  - 10 Hong Kong 0.737
- Financial Markets Efficiency (top 10):
  - 1 Saudi Arabia 1.000
  - 1 Turkey 1.000
  - 1 China, Mainland 1.000
  - 1 Italy 1.000
  - 1 Hong Kong 1.000
  - 1 Korea, Republic of 1.000
  - 1 Spain 1.000
  - 1 United States 1.000
  - 9 Japan 0.950
  - 10 Germany 0.874

### Middle-range performers (selected ranks and values)
- Financial Markets Depth (ranks 11–40, selected):
  - 11 Finland 0.820
  - 12 Malaysia 0.817
  - 13 Hong Kong 0.815
  - 14 France 0.811
  - 15 Japan 0.757
  - 20 Thailand 0.700
  - 25 Philippines 0.626
  - 30 Austria 0.535
  - 35 Barbados 0.491
  - 40 U.A.E. 0.418
- Financial Markets Access (ranks 11–40, selected):
  - 11 United Kingdom 0.708
  - 12 Greece 0.700
  - 13 Canada 0.687
  - 14 Qatar 0.684
  - 15 Singapore 0.681
  - 20 New Zealand 0.592
  - 25 Slovenia 0.560
  - 30 Brunei Darussalam 0.500
  - 35 Sweden 0.500
  - 40 Mexico 0.444
- Financial Markets Efficiency (ranks 11–40, selected):
  - 11 Russian Federation 0.834
  - 12 Australia 0.806
  - 13 United Kingdom 0.800
  - 14 Finland 0.795
  - 15 Moldova 0.763
  - 20 France 0.632
  - 25 South Africa 0.523
  - 30 Israel 0.437
  - 35 Greece 0.361
  - 40 Iceland 0.269

### Low-ranked and zero-value countries (selected)
- Financial Markets Depth (ranks with low values and zeros):
  - 100 Ghana 0.064
  - 110 Zambia 0.047
  - 120 Mauritania 0.030
  - 130 Moldova 0.023
  - 140 Armenia 0.009
  - 150 Libya 0.000
  - 160 C.A.R. 0.000
  - 170 Micronesia, Fed. Sts. 0.000
  - 180 Vanuatu 0.000
  - 190 (beyond listed) many small economies recorded 0.000
- Financial Markets Access (ranks with low values and zeros):
  - 100 Tanzania 0.006
  - 110 South Sudan 0.000
  - 120 Eritrea 0.000
  - 130 Guinea 0.000
  - 140 Liberia 0.000
  - 150 Nicaragua 0.000
  - 160 Laos 0.000
  - 170 Serbia 0.000
  - 180 Seychelles 0.000
  - 190 (beyond listed) many economies recorded 0.000
- Financial Markets Efficiency (ranks with low values and zeros):
  - 100 Armenia 0.007
  - 110 Swaziland 0.000
  - 120 Sierra Leone 0.000
  - 130 Niger 0.000
  - 140 Liberia 0.000
  - 150 Cambodia 0.000
  - 160 Tonga 0.000
  - 170 St. Vincent and the Gren 0.000
  - 180 Seychelles 0.000
  - 190 (beyond listed) many economies recorded 0.000

### Notable patterns and exact-score clusters
- Multiple countries achieved a perfect score of 1.000 in Financial Markets Access and Financial Markets Efficiency (examples preserved exactly as listed):
  - Financial Markets Access: Norway 1.000; Ireland 1.000; Luxembourg 1.000.
  - Financial Markets Efficiency: Saudi Arabia 1.000; Turkey 1.000; China, Mainland 1.000; Italy 1.000; Hong Kong 1.000; Korea, Republic of 1.000; Spain 1.000; United States 1.000.
- Several countries have identical or very close scores across dimensions (examples preserved exactly as listed):
  - Sweden: Depth 0.996; Access 0.500; Efficiency 0.695.
  - United States: Depth 0.971; Access 0.665; Efficiency 1.000.
  - Hong Kong: Depth 0.815; Access 0.737; Efficiency 1.000.
- Large group of countries recorded 0.000 in one or more dimensions; these include small island states and fragile/post-conflict economies (entries preserved exactly as listed, e.g., C.A.R. 0.000 across dimensions in parts of the listing).

*Source: IMF staff estimates.*

---


_Source: https://www.imf.org/-/media/websites/imf/imported-full-text-pdf/external/pubs/ft/wp/2016/_wp1605.pdf_