Where Bank Liquidity Cushions Housing Markets: A 145-Economy Ranking

Home prices in New Hampshire set a record in July, and prices are rising in roughly 80% of US metro areas — the latest leg of a housing market that has climbed almost without interruption since 2020. Every one of those transactions rests on a bank balance sheet. Mortgages are the largest single asset class in most banking systems, which makes one number unusually important right now: bank liquid reserves as a share of assets — the cash and near-cash cushion that stands between a housing correction and a banking crisis.

The World Bank tracks that ratio for 145 economies. The spread is enormous: Libya holds liquid reserves equal to 214.1% of bank assets, while Norway holds just 1.1%. Where a country sits on that spectrum says a lot about how its banks would weather a downturn in the most credit-intensive market of all.

The Data: the world’s largest liquidity cushions

At the top of the ranking sit banking systems holding far more in liquid form than they have lent out. This is rarely a sign of strength — it usually signals credit markets that have stopped working.

RankEconomyLiquid Reserves (% of assets)Year
1Libya214.1%2024
2Solomon Islands117.0%2024
3Haiti116.5%2024
4Mozambique106.5%2024
5Afghanistan86.9%2020
6Tonga81.5%2024
7Ghana70.7%2024
8Sudan70.6%2022
9South Sudan65.4%2024
10Sao Tome and Principe61.7%2024

Source: World Bank WDI (FD.RES.LIQU.AS.ZS).

In these economies, extreme liquidity is a symptom of broken lending, not prudent buffers. When banks cannot find creditworthy borrowers, reserves pile up in government paper and central bank deposits. Ghana is the instructive case: 70.7% liquidity sits alongside a 20.6% nonperforming loan ratio — the banks are holding cash because the loan book is too damaged to grow. For more on that dynamic, see our report on bad loans and bank stress.

Housing-heavy economies: where the cushion is thin

The countries where housing finance actually matters — deep mortgage markets, high home-ownership, heavy household debt — sit at the opposite end of the ranking. Their banks run lean because they are fully loaned out, mostly into real estate.

EconomyLiquid Reserves (% of assets)Private Credit (% of GDP)NPL RatioYear
Japan35.6%194.6%1.2%2024
Israel26.4%69.8%1.0%2024
United States14.1%201.2%0.8%2024
Australia8.3%129.3%0.7%2024
Chile8.3%104.0%2.1%2024
New Zealand6.2%142.6%—2024
Sweden6.0%125.5%0.4%2024
Hong Kong6.0%231.0%1.2%2024
India5.3%40.0%1.7%2024
Denmark3.7%144.1%0.9%2024
South Korea3.6%160.3%0.2%2024
Norway1.1%128.8%0.4%2024

Sources: World Bank WDI — liquid reserves (FD.RES.LIQU.AS.ZS), private credit (FS.AST.PRVT.GD.ZS), NPLs (FC.ABN.NONV.ZS). Years shown are the latest observation for each indicator.

Norway and South Korea mark the extreme: banks with barely 1–4% of assets in liquid form, yet private credit at 129–160% of GDP — much of it housing. These systems run on the assumption that loans keep performing and that the central bank will always be there as backstop. Their NPL ratios (0.4% and 0.2%) are among the lowest on Earth, which is why the model has worked. But it is a model with no spare tyre: a housing downturn that pushes defaults up would force these banks to sell assets into a falling market. Our explainer on what nonperforming loans mean for bank stability covers the mechanics.

The US story: reserves deployed as prices climbed

The United States shows the housing–liquidity dynamic in motion. US bank liquid reserves rose from 15.4% of assets in 2020 to a pandemic peak of 22.3% in 2021, then fell steadily as banks redeployed cash into credit — 19.0% in 2022, 16.5% in 2023, and 14.1% in 2024 — while home prices climbed to records. The cushion was spent financing the same market whose price level it was meant to protect against.

YearUS Liquid Reserves (% of assets)
202015.4%
202122.3%
202219.0%
202316.5%
202414.1%

Source: World Bank WDI (FD.RES.LIQU.AS.ZS).

This is not inherently dangerous — the 2024 ratio is still above its pre-pandemic level, and US NPLs remain low at 0.8%. But it illustrates the central tension of housing finance: the more a banking system lends into a rising property market, the thinner the liquidity cushion it keeps for the turn. For the context on how lending rates shape that borrowing, see our analysis of the cost of credit.

Analysis: liquidity is the price of lending into a boom

The data describes a consistent inverse relationship: economies with the deepest credit markets run the thinnest liquidity buffers, and vice versa. Hong Kong is the sharpest example — the world’s highest private-credit ratio (231% of GDP) paired with just 6.0% liquid reserves. Japan is the outlier: 35.6% liquidity alongside 194.6% of GDP in private credit, a legacy of decades of weak loan demand after its own housing bust in the early 1990s.

That inverse relationship is the reason liquidity ratios must be read alongside credit quality and capital. A bank with 6% liquidity, a clean loan book, and strong capital can be safer than one with 40% liquidity and a deteriorating book. The combination of capital adequacy and loan quality determines whether a thin cushion ever gets tested.

Context: the housing question

The July price data extends a pattern that has defined the post-pandemic cycle: housing wealth keeps rising while the banking systems that finance it hold progressively less cash. For borrowers, the relevant cost side is covered in our piece on where a car loan is possible and the broader how the world borrows report. For the banking-system side, the key indicators to watch are liquidity, NPLs, and capital — and how they move together if prices turn.

Sources & method

All figures come from FinStatGlobe’s derived country datasets, built from the World Bank World Development Indicators: bank liquid reserves to assets (FD.RES.LIQU.AS.ZS, 145 economies), domestic credit to the private sector (FS.AST.PRVT.GD.ZS), and nonperforming loans to gross loans (FC.ABN.NONV.ZS). Rankings use each country’s latest observation, which ranges from 2020 to 2024 depending on reporting schedules; years are cited alongside every figure. Time series show annual observations. No projections are used. For how the snapshot, derivation, and ranking steps work, see the methodology page. For the baseline ranking of bank liquidity without the housing lens, see our earlier report on how liquid the world’s banks are.