The same deposit, with a shorter waiting time
At the end of July, seasonally adjusted US M2 stood at 23.218 trillion dollars. Demand deposits accounted for 7.088 trillion dollars and other liquid deposits for 10.416 trillion dollars. Those sums show the scale of the liability base on which bank lending rests. The tokenized deposit examined by the Dallas Fed does not remove the account from the bank's balance sheet; instant payment and transfer can change how long that account is expected to remain there.[2], [1]
A deposit may be withdrawable on demand in the contract, yet most accounts do not disappear overnight in practice. Dallas Fed researchers describe that behaviour with two variables: the account's weighted average life and its beta, the share of a market-rate change passed through to the deposit rate. If tokenization shortens average life and raises beta, the relatively patient funding that helps a bank carry fixed-rate long-term loans shrinks. Banks could instead change pricing to keep customers in place, leaving both effects modest.[1]
Speed travels across both sides of the balance sheet
The paper's sensitivity exercise makes the scale of transmission visible. A 10 per cent fall in deposits' weighted average life reduces banking-system duration capacity by about 580 billion dollars in 10-year equivalents. A 10 per cent rise in deposit beta produces an estimate of about 700 billion dollars. These are scenarios under stated assumptions, not losses that have already occurred. The direction is still clear: a faster and more rate-sensitive liability leaves less room to carry an asset of the same long maturity.[1]
A bank can close that gap in three places. It can hold more reserves and US Treasuries against unexpected outflows, issue longer-term debt in place of deposits, or shorten the maturity of its loan portfolio. The first route dedicates more of the balance sheet to assets that can become cash immediately. The second moves funding costs closer to wholesale-debt pricing. The third carries the adjustment to households and firms seeking credit. Greater use of intraday Federal Reserve credit and the discount window could ease the pressure; how far banks would use those facilities remains uncertain.[1]
The first signal is in composition, not the total
The first test of this mechanism is not whether M2 alone rises or falls. In July, M2 increased from 23.115 trillion dollars in June to 23.218 trillion dollars, while demand and other liquid deposits continued to form a large base. Total deposits could stay unchanged as tokenization spreads, with money merely shifting into accounts that move faster. The useful distinction is therefore between the total stock and deposit composition: if the liquid-account share, bank holdings of reserves and Treasuries, and term-debt issuance move together, the balance-sheet adjustment becomes visible.[2], [1]
The credit-side result arrives later. Banks may retain deposits through pricing and keep faster payments from reaching lending terms; that is the strongest counter-scenario. If funding becomes shorter-lived while liquid-asset holdings or term borrowing rise, however, cost and maturity choices travel into credit. Tokenized deposits would not have created new money in that case. They would have changed how patiently existing money sits on a bank balance sheet, redrawing the boundary for long-term lending with it.[1]