What Global Liquidity is
Global Liquidity is the balance-sheet capacity behind global credit: central bank money plus the government bonds that credit is secured on, times their reuse. BlockPhi estimates it every Friday from the balance sheets of five central banks and 17 government debt markets, times a collateral multiplier; for the week ending 9 October 2026 the series stood at about $198 trillion. This is a model estimate, and past results do not indicate future results.
I What it is
Global Liquidity is the flow of credit and savings through financial markets. In practice it is balance-sheet capacity: central bank money plus the government bonds that banks, funds and dealers borrow against, multiplied by how many times that collateral is reused. It funds trades and rolls over debt.
In 2011 the Committee on the Global Financial System, which the BIS hosts, split global liquidity in two (CGFS Papers No 45). Official liquidity is ultimately provided by central banks. Private liquidity is created to a large degree through the cross-border operations of banks and other financial institutions.
Five drivers outside bank deposits
- Collateral-based lending. Since 2008 wholesale lending has run largely on collateral, and the collateral used most is US Treasuries. Much of that lending refinances existing debt rather than funding new investment.
- Non-bank lenders and cash pools. Broker-dealers, funds and the cash pools of large companies and institutions lend against high-quality collateral, largely outside bank regulation.
- Offshore dollars. Dollars held and lent outside the United States, in the Eurodollar market, sit outside the Fed’s jurisdiction and outside M2. FX swaps create dollar funding that M2 does not count. A dollar that crosses a border can become collateral for a foreign lender and be lent again.
- Central bank and Treasury operations. The Treasury General Account, the Fed’s reverse repo facility and the mix of bills and bonds the Treasury issues add or drain liquidity with no change in the policy rate.
- The collateral multiplier. The same Treasury is pledged, borrowed against and pledged again. How many times depends on the haircut lenders apply, and the haircut depends on bond volatility. When volatility falls, the same collateral supports more credit.
II What it is not
Not Global M2. M2 counts cash and retail deposits, mostly current account and savings balances at retail banks. Global Liquidity counts the balance-sheet capacity of every credit provider, banks and non-banks alike. M2 has no equivalent of the collateral multiplier.
Not the BIS global liquidity indicators. Those measure credit to non-bank borrowers, with a focus on foreign-currency credit in US dollars, euros and yen, on a quarterly clock.
Not a policy rate. Central banks talk in rates. Markets run on balance sheets. A rate cut changes the price of credit. It does not by itself add balance-sheet capacity.
III Why global, not just the Fed
Liquidity is fungible. It crosses borders and settles wherever the return is. Fed net liquidity is one line of it. The balance sheets of the People’s Bank of China, the European Central Bank, the Bank of Japan and the Bank of England are others, and so are the government bond markets that lenders borrow against.
Who is driving
Two central banks are usually given the most weight. The Federal Reserve moves financial asset prices, because the world runs on a dollar standard. The People’s Bank of China moves the pace of China’s economy and, through its industrial footprint, global activity and commodity markets. The BlockPhi Global Liquidity Model carries both balance sheets, with those of the European Central Bank, the Bank of Japan and the Bank of England.
The dollar
Cross-border liquidity moves with the dollar. A strong dollar drains it; a weak dollar feeds it. When dollars flow into emerging markets, local central banks buy them to hold their currencies down and create money against them, so domestic credit expands on top of the inflow. That is the emerging market and commodity channel.
The Global Liquidity Model is revalued every Friday in US dollars, at current exchange rates and, for six markets, current government bond yields. A weaker dollar raises its level and a stronger dollar lowers it.
IV How it moves every asset
In the collateral-based reading of credit, liquidity reaches assets in sequence: central bank reserves reprice bonds first, then corporate credit, then equities as institutions reallocate towards riskier assets, and then Bitcoin.
The sequence describes position in the chain, not timing. Published estimates of how far liquidity leads each market vary with the measure they are taken on. The lead-lag record reconciles the published estimates against the measure each was taken on, and the research page shows the one direction test run on the current series.
The debt that has to be refinanced
Debt is why liquidity reaches every asset. Most outstanding debt is refinanced rather than repaid, and refinancing needs balance-sheet capacity, which is liquidity. In this reading, when debt grows faster than liquidity, refinancing is at risk, and the crises of recent decades read as refinancing crises: 2008, Asia in 1997, the US repo market in 2019.
What central banks defend
Whatever their mandates say, central banks have defended the sovereign debt market when it broke. In late 2022, when the UK gilt market sold off, the Bank of England bought gilts in an emergency operation that ran from 28 September to 14 October 2022.
The Global Liquidity Model made its 2022 low, about $155 trillion, in the week ending 21 October 2022, one week after those purchases ended. That week is part of the reconstructed history, and the figure is a model estimate. Most of the fall in the series’ collateral base between 25 February and 21 October 2022 came from the stronger dollar.
V Why Bitcoin sits at the end of the chain
Bitcoin sits at the outer edge of institutional allocation, the end of that chain. It has no earnings to miss, no chief executive and no debt to default on.
Its supply follows a fixed schedule, to a maximum of 21 million coins. The supply of liquidity has no such limit: on the Global Liquidity Model, in US dollars, Global Liquidity grew about 7.3% a year between the weeks ending 3 January 2020 and 9 October 2026, a model estimate.
Gold is the older protection against monetary inflation, but central banks buy, hold and sell it, so its price carries their policy. No balance sheet decides Bitcoin’s supply.
Bitcoin’s rise can be read in two parts: adoption, the S-curve of new holders, and allocation, how much existing holders and institutions hold as they expect more monetary inflation. The Global Liquidity Model tracks the liquidity behind the second, not the adoption behind the first. On this reading, liquidity bears more on Bitcoin’s cycle than on its adoption trend.
VI How BlockPhi measures it
The Global Liquidity Model is a composite that measures the balance-sheet capacity behind global credit: the balance sheets of five central banks plus 17 government debt markets, revalued every Friday in US dollars and multiplied by a collateral multiplier keyed to the US ten-year Treasury yield. Published every Friday after the US close.
On the Global Liquidity Model, Global Liquidity for the week ending 9 October 2026 was $198.20 trillion. The collateral base was $104.84 trillion: $26.06 trillion of central bank money and $78.77 trillion of sovereign debt. The collateral multiplier was 1.8905. Figures are rounded, so the parts may not add exactly. Model estimate as of 9 October 2026. Past results do not indicate future results.
The series follows the BIS split. Central bank money stands for the official part. The government bonds that banks, funds and dealers borrow against stand for the private part. The series is closer to that definition than M2 is. It is not the BIS’s own measure.
The multiplier has two parameters, keyed to the US ten-year Treasury yield, estimated once and frozen. It is named for collateral reuse but does not measure reuse, haircuts, bond volatility, dealer capacity rules or margin.
Recorded live since 31 July 2026. Weeks before 31 July 2026 are reconstructed with the current method, each input entering on a lag that follows its usual release schedule. They are not a record of what was known on each date.