Frequently Asked Questions

Financial Development Data (FinData) is a new, annually updated dataset developed by the World Bank Group to measure financial development across up to 218 economies from 1980 onwards. It combines traditional banking and capital market indicators with measures of digital finance and financial inclusion. It also provides composite indices that summarize financial development across economies and over time.

FinData was created to fill a gap in comprehensive, publicly available data on financial systems. The World Bank’s Global Financial Development Database (GFDD) covered over 200 economies from 1960 to 2021 but stopped being updated in September 2022.Meanwhile, financial systems changed significantly, with the rapid expansion of digital finance and mobile money, the growing role of private equity and venture capital, and increasing attention to sustainable finance. Existing databases, designed around traditional banking and capital market indicators, do not adequately capture these developments. FinData  addresses both the need for updated cross-economy data and the need to cover newer areas of finance.

FinData builds on the GFDD’s and FinStats’s legacy and extends its framework. FinStats was a financial sector diagnostic tool used within the World Bank.   Both assessed four dimensions of financial development across two components of the financial system, financial institutions and financial markets (a 4 × 2 framework). FinData splits each component into two segments, resulting in a 4 × 4 framework. It also introduces composite indices to summarize financial development and new indicators on digital finance, private capital markets, and sustainable finance. Its metadata file follows the format established by the GFDD.

FinData is designed for a broad set of users across the public and private sectors. In the public sector, central banks, financial regulators, and ministries of finance can use the data to benchmark performance, identify structural constraints, and inform policy decisions, while foreign investment agencies can draw on it to guide strategies.

In the private sector, analysts and researchers at banks, investment funds, insurance companies, and other financial intermediaries can use FinData to evaluate opportunities and benchmark markets. In academia, economists and development finance scholars can draw on the data to study financial systems, inclusion, economic outcomes.

International organizations, think tanks, and financial media also form a key part of the user base, helping to amplify the database’s policy relevance and public reach.

 FinData is organized around a 4 × 4 framework. Its four segments are banks and non-bank financial institutions (NBFIs), under the financial institutions pillar, and equity and debt markets, under the financial markets pillar. Each segment is assessed along four dimensions: depth, access, efficiency, and stability.

Indicators cover up to 218 economies, though coverage varies by indicator and year. The composite indices are computed for 147 economies.

The indices consider World Bank members, along with Hong Kong SAR, China and Taiwan, China, given their importance as international financial centers. An economy is included only if at least 25 percent of its economy-year observations across the 34 core indicators are available over 2015–2024. The threshold keeps indices from being driven by sparse data and heavy imputation. As a result, the indices cover 147 economies; the excluded economies are listed in the methodology note.

Traditional banking indicators are available from 1980 onward. Digital finance indicators are generally available from about 2011, when the Global Findex began, although some mobile money series begin earlier. The composite indices cover 2000–2024, given that data coverage drops sharply before 2000 across both indicators and economies.

Comparable cross-economy data on NBFIs remain limited, particularly for access, efficiency, and stability. FinData therefore currently measures only NBFI depth, using the assets of insurance companies, pension funds, and mutual funds relative to GDP. Future releases will seek to add indicators that better capture the role and performance of NBFIs.

FinData contains 67 indicators drawn from both publicly available and licensed data sources. These comprise 34 core indicators and 33 supplementary indicators. The website highlights the 34 core indicators because they were selected based on their demonstrated relevance to economic growth, methodological robustness, country and time coverage, and reporting frequency. The core indicators are organized across the four dimensions of financial development: 13 for depth, 10 for access, 6 for efficiency, and 5 for stability. All indicators are defined, sourced, and documented in the accompanying metadata file.

• Traditional financial sector indicators:
  • International Monetary Fund (IMF): International Financial Statistics, Financial Soundness Indicators, and Financial Access Survey.
  • Bank for International Settlements (BIS): credit, debt securities, locational banking, and financial statistics databases.
  • World Federation of Exchanges, Organisation for Economic Co-operation and Development (OECD) Institutional Investors database, and the World Bank’s World Development Indicators.

• Market and institution-level data: Fitch/BankFocus, Bloomberg, London Stock Exchange Group (LSEG), Dealogic, PitchBook, AXCO, White Clarke Global Leasing Report, Factors Chain International, and other specialized providers.

• Household access and usage: the World Bank’s Global Findex.

• Firms: the World Bank Enterprise Surveys.

• Other World Bank datasets: specialized datasets such as Remittance Prices Worldwide.

Each indicator goes through quality control to detect outliers and unreliable observations. The checks examine economy-level growth rates, within-economy outliers, anomalies in the cross-economy distribution for each year, and deviations from the indicator's overall distribution. When a suspect value can be verified against a credible external source, it is corrected using the verified information. When a value cannot be substantiated and remains outside acceptable bounds, it is treated as unreliable and set to missing. Every modification, together with its rationale and supporting reference, is documented in the reproducibility package.

Debt-market depth is measured by outstanding international public and private debt securities, together with total bond issuance. The debt-market indices therefore reflect access to international debt funding as well as domestic bond market development.

Yes, FinData includes selected sustainable finance indicators.

FinData publishes 16 financial development indices on a 0–100 scale, where higher values indicate greater financial development. They fall into three groups:

• Overall and pillar indices: Overall Financial Development (OFD), the Financial Institutions and Financial Markets indices (the two pillars), and the depth, access, and efficiency components of each pillar.
• Segment indices: Banks, NBFIs, Equity Markets, and Debt Markets.
• Dimension indices: Depth, Access, and Efficiency.

These indices are built from 10 segment-dimension indices, such as Bank-Depth or Equity Markets-Access. Because NBFIs are currently measured for depth only, three published indices each equal one underlying segment-dimension index: Financial Institutions-Access equals Bank-Access, Financial Institutions-Efficiency equals Bank-Efficiency, and the NBFI index equals NBFI-Depth.

FinData indices are constructed as follows:

1. Each indicator is winsorized at the 5th and 95th percentiles of its pooled distribution across all economies and years, to limit the influence of outliers.
2. Each winsorized indicator is then rescaled to 0–1 using min-max normalization over the same pooled sample. Indicators where lower values indicate greater financial development, such as net interest margins, cost-to-income ratios, and bid-ask spreads, are inverted so higher values consistently indicate greater financial development.
3. The available normalized indicators within each segment-dimension cell are averaged to form a segment-dimension index, which is then min-max normalized again across all economies and years.
4. Higher-level indices are simple averages of the available indices one level below. They are not re-normalized, so each can be interpreted directly as the equal-weight average of its available components.

At each stage, an index is computed if at least one relevant component is available. If none are available, the index remains missing. For publication, the indices are multiplied by 100 and expressed on a 0–100 scale.

FinData follows a minimum-necessary-imputation principle, using the least invasive method that materially improves coverage or reliability. Values are first restricted to economically meaningful ranges. Then:

1. Internal gaps are filled by linear interpolation.
2. Gaps at the start or end of a series are reconstructed using the growth of closely related indicators in the same segment and dimension. Annual growth factors are bounded between 0.80 and 1.25, reciprocal limits that treat proportional changes symmetrically in opposite directions.
3. Remaining gaps at the end of a series may be filled by carrying forward the most recent available or reconstructed value for up to three years.

Any remaining gaps are left missing. Bank return on assets, which can be positive or negative, is excluded from growth-based reconstruction and is filled only through interpolation of internal gaps and carry-forward of trailing gaps.

Several indicators come from surveys that are not conducted every year. The Global Findex, for example, is fielded roughly every three years. Linear interpolation between survey waves is therefore applied without a maximum gap length, so that values change gradually rather than jumping between waves.

Equal weights keep aggregation transparent and easy to interpret: each index is the simple average of its available components. FinData leaves observations missing when data remain unavailable and aggregates the components that are available. Using weights from principal component analysis would therefore require an additional choice of estimation sample and the resulting weights could change as the sample evolves. Equal weighting avoids this additional source of variation and preserves a transparent aggregation structure.

Stability is part of the framework, and FinData reports stability indicators for banks, equity markets, and debt markets. However, stability captures the resilience of a financial system to shocks, which is distinct from its level of financial development. A more developed financial system is not necessarily more stable, a stable system is not necessarily a developed one, and rapid financial deepening can be associated with a build-up of vulnerabilities. Averaging stability with depth, access, and efficiency could therefore obscure these differences. Stability indicators are reported separately so that financial development and financial stability can be assessed alongside one another.

Yes. Winsorization and normalization use the pooled distribution across all economies and years, placing index values on a common scale and supporting comparisons across economies and over time. However, because the indices are constructed using the components available for each economy-year, their underlying composition can vary. Comparisons should therefore be interpreted with these differences in coverage in mind.

Yes. Because winsorization and normalization bounds are calculated using the pooled distribution across all economies and years, new observations, particularly at the extremes, can shift these bounds and therefore change the index values for earlier years. The winsorization bounds are also reviewed periodically as the pooled distribution evolves. We recommend using a single FinData release throughout an analysis and citing the data vintage used.

Yes. A reproducibility package, independently reviewed by the World Bank’s DEC Reproducibility Team, documents the construction of the indices, including every modification made to individual data points. 

FinData's financial development indices and underlying core indicators can be downloaded from this website. On the Home page, select a segment, dimension, and year, then use the Download button to export the data. 

Please cite the database as: World Bank. 2026. Financial Development Data (FinData). Washington, DC: World Bank. https://www.worldbank.org/en/findata

For the methodology, please cite: Bertay, A. C., E. Feyen, D. S. Mare, J. Marzluf, and D. M. Sourrouille. 2026. "The Financial Development Data: Methodology for Measuring Financial Development." World Bank, Washington, DC.