Gross portfolio debt liabilities relative to GDP compare the stock of portfolio debt instruments recorded as liabilities with the size of an economy’s annual output. In the World Bank GFDD.DM.10 data used here, the latest available observation is present for 132 economies. Their median ratio is 15.90% of GDP, while the arithmetic mean is 32.03%. The large gap between those two summary measures immediately signals a right-skewed distribution with a relatively small group of very high ratios.
The dataset is not a single-year panel. 110 of 132 observations (83.3%) are from 2020, but 22 economies use an earlier last non-missing observation. The most striking example is Luxembourg: its value of 564.07% is from 2002. For that reason, the map and ranking should be described as latest available observations through 2020, not as a 2020 ranking of all 132 economies.
Table of Contents
What gross portfolio debt liabilities to GDP measures
The numerator covers debt instruments classified as portfolio liabilities. The World Bank description includes bonds, debentures, notes and similar instruments, together with money-market or other negotiable debt instruments. The denominator is gross domestic product. The ratio therefore combines a financial stock measured at a point in time with a flow of production measured over a year.
The word “gross” matters. It refers to the liability side before offsetting it with corresponding portfolio debt assets. The indicator is also not interchangeable with central-government debt. Portfolio debt can involve different issuing sectors and financial instruments, so a country with a high GFDD.DM.10 ratio should not automatically be described as having high government debt, and a low ratio does not imply that its total debt burden is small.
Ratios above 100% are entirely possible. A debt-security position can accumulate over many years, whereas GDP records only one year’s output. In this 132-economy dataset, 10 observations are at or above 100% of GDP. That does not by itself identify financial distress, solvency risk or policy quality. It shows that the recorded portfolio debt liability stock is large relative to annual domestic production.
The world map highlights a strong European concentration at the high end
The map shows a visible cluster of high ratios in parts of Western and Northern Europe. The Netherlands is at 175.62% in 2020, France at 128.56%, the United Kingdom at 111.65% and Belgium at 105.24%. Denmark, Sweden and Norway are all in the high-80% range. Yet nearby economies still differ substantially: Germany is 65.97%, Italy 70.81% and Spain 86.97%.
Elsewhere, the values are more dispersed. Japan records 39.47% in 2020, compared with 4.72% for China and 3.92% for India. Malaysia is at 29.42%, Indonesia at 18.87% and Thailand at 10.43%. In the Americas, Canada reaches 81.48%, the United States 62.20%, Chile 37.57% and Brazil 12.27%.
These geographic patterns are descriptive, not causal. International financial-centre activity, the depth of domestic securities markets, cross-border issuance, the financing mix of corporations and financial institutions, and statistical coverage can all affect the ratio. A map can show where high and low values are concentrated, but it cannot isolate which mechanism produced a particular country’s level. Answering that question would require sector, issuer, currency, maturity and holder information that GFDD.DM.10 does not provide by itself.
The highest latest-available ratios and the Luxembourg caveat
Luxembourg has the highest latest-available ratio in the file at 564.07%, but the observation is from 2002. It should not be placed on the same temporal footing as Ireland at 179.08% in 2020, the Netherlands at 175.62% in 2020 or France at 128.56% in 2020. The underlying collection rule selects one last non-empty observation for each economy, so a stale value can still appear at the top of a latest-available ranking.
| Rank | Economy | Observation year | Share of GDP |
|---|---|---|---|
| 1 | Luxembourg | 2002 | 564.07% |
| 2 | Ireland | 2020 | 179.08% |
| 3 | Netherlands | 2020 | 175.62% |
| 4 | France | 2020 | 128.56% |
| 5 | Mauritius | 2020 | 126.53% |
| 6 | Finland | 2020 | 121.83% |
| 7 | United Kingdom | 2020 | 111.65% |
| 8 | Belgium | 2020 | 105.24% |
| 9 | Cyprus | 2020 | 100.98% |
| 10 | Austria | 2020 | 100.82% |
If attention is restricted to the 2020 rows, Ireland and the Netherlands remain above 170%. France, Mauritius, Finland, the United Kingdom, Belgium, Cyprus and Austria are also at or above 100%. Europe is therefore heavily represented among the high-ratio observations. That does not establish a universal threshold for “too much” debt. The indicator covers a specific set of portfolio debt securities, not every liability in the economy, and it says nothing directly about net external positions or the quality of the assets that may sit on the other side of balance sheets.
A median of 15.90% tells a different story from the 32.03% mean
Across all 132 economies, the first quartile is 6.13%, the median 15.90% and the third quartile 33.32%. The mean of 32.03% is substantially higher because the distribution has a long upper tail. The old Luxembourg observation is the most extreme value, but it is not the only high ratio: 10 economies reach at least 100%, and another 13 sit between 50% and 100%.
The frequency breakdown makes the shape clearer. 17 economies are below 1% of GDP, 29 are from 1% to under 10%, 63 are from 10% to under 50%, 13 are from 50% to under 100%, and 10 are at 100% or more. The 10–50% interval contains the largest number of observations. Reporting only the mean would therefore give too much influence to a small number of exceptionally high values; the median, quartiles and bins provide a more faithful description of a typical economy in this sample.
Very small positive ratios are not the same as zero or missing data
At the bottom of the file, Moldova is about 0.0029%, Tonga 0.0261%, the Kyrgyz Republic 0.0471% and the Maldives 0.0831%. These are tiny positive observations, not zeros. Rounding every small value to one decimal place would erase meaningful distinctions between a recorded liability position, a true zero and an unavailable observation. The same principle applies to the map: economies without a compatible polygon are not filled with zero.
| Rank | Economy | Observation year | Share of GDP |
|---|---|---|---|
| 1 | Moldova | 2020 | 0.00% |
| 2 | Tonga | 2020 | 0.03% |
| 3 | Kyrgyz Republic | 2020 | 0.05% |
| 4 | Maldives | 2010 | 0.08% |
| 5 | Tanzania | 2019 | 0.11% |
| 6 | Tunisia | 2011 | 0.12% |
| 7 | Cabo Verde | 2020 | 0.17% |
| 8 | Bosnia and Herzegovina | 2020 | 0.22% |
| 9 | Guinea | 2019 | 0.22% |
| 10 | Sao Tome and Principe | 2020 | 0.23% |
A low GFDD.DM.10 value does not mean that an economy has little debt overall. Bank loans, trade credit, direct-investment liabilities, non-marketable public debt and other categories can be important even when portfolio debt securities are small. The denominator matters as well: a large GDP can reduce the ratio for a given liability stock. The measure should therefore not be treated as a comprehensive leverage score or a sovereign-credit rating.
The observation years range from 2002 to 2020
The year distribution is 2020: 110, 2019: 12, 2018: 3, 2016: 1, 2013: 1, 2011: 2, 2010: 1, 2008: 1, 2002: 1. The dominance of 2020 makes the map broadly representative of conditions around that year, but the earlier values are not trivial. Luxembourg is from 2002, Azerbaijan from 2008, the Maldives from 2010, and several other economies use 2011, 2013, 2016, 2018 or 2019 data. A strict contemporaneous comparison should therefore filter the dataset to the 110 observations dated 2020.
“Latest available” is useful when the objective is geographic coverage. It allows more economies to appear in one comparison even when reporting histories differ. The trade-off is that it cannot answer a time-specific question with the same precision as a balanced panel. If the research question is whether portfolio debt liabilities increased or decreased, a full annual time series should be used instead of comparing these one-row snapshots.
What this ratio can and cannot tell you
The ratio is useful for scale. Dividing by GDP allows economies with very different absolute sizes to be compared on a common basis. The United States records 62.20% in 2020, Canada 81.48% and Australia 77.17%. Those percentages answer a different question from the absolute dollar value of outstanding portfolio debt liabilities.
The indicator does not report maturity, interest cost, currency denomination, issuer sector, investor residence, refinancing risk or repayment capacity. It does not show expected bond returns, forecast interest rates or tell an investor which country is attractive. Nor can it identify whether a high ratio is mainly the result of government securities, corporate issuance or financial-sector intermediation. Those questions require additional datasets.
For policy or balance-sheet analysis, GFDD.DM.10 is best read alongside central-government debt, private credit, external assets and liabilities, current-account balances, foreign-exchange reserves and the currency and maturity structure of debt. The value of this page is narrower: it makes the cross-country scale of portfolio debt liabilities visible and makes the data limitations explicit rather than turning a single ratio into a judgment about financial health.
Source and calculation method
The definition follows the World Bank Global Financial Development metadata for GFDD.DM.10. The World Bank describes the debt liabilities as bonds, debentures, notes and related instruments plus money-market or negotiable debt instruments, and identifies the IMF’s International Financial Statistics as the underlying raw source. Economy-level values are available from the World Bank indicator page for GFDD.DM.10.
This analysis uses one latest non-missing observation for each of 132 economies. The mean, median, quartiles, frequency bins and rankings are calculated directly from those rows; missing values are never replaced with zero. For the choropleth, country codes are joined to Natural Earth low-resolution polygons, with France and Norway matched by name where the boundary file lacks standard ISO-3 codes. That produces 115 mapped economies, while all 132 remain in the statistical calculations and tables.
Frequently Asked Questions
Can gross portfolio debt liabilities exceed 100% of GDP?
Yes. The liability measure is a stock at a point in time, while GDP is a one-year flow, so the recorded portfolio debt stock can exceed annual GDP.
Is this the same as the government-debt-to-GDP ratio?
No. GFDD.DM.10 covers portfolio debt instruments recorded as liabilities and is not equivalent to total central-government debt.
Are all 132 observations from 2020?
No. 110 observations are from 2020 and the remaining values are earlier latest-available observations. The complete set is latest available through 2020.
Related Articles
Green Map creates custom-edited map images using open geographic data sources such as geoBoundaries, Natural Earth, OpenStreetMap, and government open data. These maps are edited visual materials, not raw data files, and are provided for education, documents, presentations, and graphic reference.
