Tokenized bank deposits could make funding less stable and reduce U.S. banks’ capacity to finance long-duration assets, according to new research from the Federal Reserve Bank of Dallas. Economists Rosie Levy and Srini Ramaswamy argue that faster settlement and greater sensitivity to interest rates could change how deposits behave even while they remain inside the regulated banking system. The concern is not tokenization itself, but whether more portable deposits become less reliable as a source of long-term bank funding.
Unlike stablecoins, tokenized deposits remain claims on banks and can retain protections associated with conventional deposits. Yet their operational characteristics could be different. The Federal Reserve Bank of Dallas analysis says instant settlement could make it easier for yield-sensitive customers to move funds between institutions, potentially shortening deposit duration and increasing deposit-rate betas. Both changes would weaken the maturity transformation that allows banks to fund longer-term loans with relatively stable deposits.
Small Behavioral Changes Could Have Large Balance-Sheet Effects
The Dallas Fed estimates that deposits currently support roughly 80% of the duration risk carried by U.S. banks, measured on a 10-year-equivalent basis. Under one scenario, a 10% reduction in the weighted average life of deposits would cut aggregate maturity-transformation capacity by about $580 billion. A separate scenario in which deposit price sensitivity rises 10% would reduce banks’ appetite for duration risk by approximately $700 billion, assuming a four-year deposit life.
Those numbers are modeling estimates, not projections of an equivalent decline in lending. Banks could respond by changing deposit pricing, issuing more term debt or restructuring their asset portfolios. The likely transmission would be through funding costs and balance-sheet composition rather than an immediate $580 billion or $700 billion contraction in credit. The researchers note that greater reliance on wholesale term funding could ultimately make borrowing more expensive for households and businesses.
Faster deposits could also force banks to maintain larger pools of highly liquid assets. The economists argue that greater uncertainty around real-time outflows could increase demand for reserves and Treasury securities to satisfy liquidity requirements and internal stress tests. That response could improve short-term resilience while leaving less balance-sheet capacity available for longer-duration lending.
Brazil’s Pix Offers an Early Warning on Faster Money
The Dallas Fed points to Brazil’s Pix system as an imperfect but useful comparison. Research using Brazilian regulatory data found that banks with heavier instant-payment usage increased holdings of liquid assets, particularly government bonds, while reducing credit intermediation. The comparison suggests that faster movement of deposits can influence bank behavior even when the underlying money remains conventional bank money.
The policy challenge is therefore broader than payment efficiency. Tokenized deposits could improve settlement speed and programmability while remaining within existing banking safeguards, but those same features may alter assumptions about how long deposits remain with an institution. Supervisors and bank treasury teams may eventually need to recalibrate liquidity stress tests, deposit-duration assumptions and funding strategies if tokenization reaches meaningful scale.
Levy and Ramaswamy do not argue that widespread adoption is inevitable or that tokenized deposits should be discouraged. Their warning is that faster, more programmable money could change the economics of bank funding before it changes the legal definition of a deposit, making liquidity behavior a central issue as tokenized banking moves beyond experimentation.
