Why Deposit Rates Don’t Move One-for-One With the Fed

A simple cumulative deposit beta is the change in a deposit rate divided by the change in a reference rate over the same interval:

Deposit beta = change in deposit rate ÷ change in reference rate

Suppose the reference rate rises 1 percentage point and a bank’s average rate on a defined set of deposits rises 0.35 percentage points. The cumulative beta for that product group and interval is 35%. A beta below 100% means the deposit rate moved by less than the reference rate; it does not tell you why, whether an individual customer received that change, or what will happen in another cycle.

The denominator and population matter. One calculation might compare a bank’s average savings rate with the effective federal funds rate; another might use a different policy-rate measure, product, customer group, or start and end dates. A short window can capture delays, while a full-cycle calculation includes more of the later repricing. Banks and researchers may also use regression-based or scenario betas rather than this simple endpoint ratio. The Federal Reserve’s examination manual describes beta factors as estimates of how product rates change relative to a driver rate and says assumptions should reflect the product and the institution’s analysis. A beta is therefore incomplete unless you know the driver, product, population, and dates.

The Fed’s policy rate guides overnight money-market conditions; it is not a rate that the central bank pays directly to a household’s savings account. A bank decides what rate to offer for each product. It weighs the cost of raising money through deposits against other funding sources, the value of keeping customer relationships, and the rate-sensitive behavior it expects from depositors.

That choice can differ by bank. An institution with abundant deposits that customers rarely move may have less reason to raise its posted savings rate quickly. A bank trying to attract new balances, replace outflows, or compete in a market where customers compare offers may reprice sooner. The offered rate can also reflect account features, service costs, balance tiers, minimums, and the bank’s plans for the rest of its balance sheet. These are commercial and funding decisions; there is no rule that retail deposits must copy a Fed change one-for-one.

The repricing of bank assets and other liabilities also occurs on different schedules. Some loans reset with a short-term benchmark; fixed-rate loans and securities may not. Wholesale borrowing can react more quickly than retail deposits. The Federal Reserve’s 2025 research on U.S. bank funding costs reports that funding betas vary with the length, magnitude, and direction of a policy cycle, and that nondeposit funding costs have generally adjusted more than deposit costs. That is a banking-system research result, not a forecast for a particular account or institution.

Demand deposits and many checking accounts may pay no interest or a low rate, while an interest-bearing checking account can use balance tiers or activity requirements. A regular savings account or a bank money market deposit account (MMDA) can have a variable rate that the bank changes at its discretion under the account terms. Two accounts at the same institution may therefore respond differently to the same policy move.

A certificate of deposit (CD) works differently. Its stated rate is generally fixed for the agreed term, so an existing CD does not ordinarily reset each time the Fed moves. The rate on a new CD, or on a CD renewed at maturity, can change as market conditions and the bank’s funding plans change. Early-withdrawal terms, renewal rules, and special promotional conditions affect the comparison, too.

“Money market” can describe two different products. A bank money market deposit account is a deposit account. A money market mutual fund is an investment fund holding short-term securities; it is not a bank deposit and should not be treated as the same product or as carrying deposit insurance. Its yield responds to market instruments and fund holdings, with different terms and risks. Comparing the two requires looking at what each legally is, how its return is set, and what protections and withdrawal conditions apply.

Even within one account label, a headline rate may depend on the balance band, linked services, new-customer eligibility, or a limited promotional period. A beta computed from a bank-wide average can hide those differences. To understand an individual offer, the account disclosure and rate schedule matter more than a generic industry beta.

Banks care about whether deposits stay put and whether customers bring in or move out balances when another option becomes more attractive. Some depositors value convenience, payment access, branches, bundled services, or a long-standing relationship. Others actively compare rates and can transfer money between institutions or to a different type of cash product. A bank’s pricing response depends partly on which customers and balances it is trying to retain.

Switching is not frictionless. Customers may need to update bill payments and direct deposits, meet minimum balances, wait for transfers, or learn a new institution’s terms. Those frictions can make some balances less rate-sensitive for a time. Digital access and easy transfers can make comparison simpler, but they do not ensure that every account holder moves at the same time or that every bank responds identically.

The March 2024 Federal Reserve Senior Financial Officer Survey illustrates why segment labels matter. Responding banks reported different cumulative betas for retail deposits, wholesale operational deposits, and wholesale non-operational deposits. The survey also asked banks about their pricing rationale; roughly three-fourths of respondents said maintaining deposit balances was the closest fit for setting betas across those categories. That helps explain the role of retention, but does not establish what any current retail customer should be offered.

Assume, purely for illustration, that a selected policy-rate measure increases by 500 basis points over a defined cycle and that a defined group of retail deposit rates increases by 175 basis points over the same dates. The cumulative beta is:

175 basis points ÷ 500 basis points = 0.35, or 35%

This is a hypothetical example, not a description of current bank offers or a prediction. It means the measured deposit-rate change was 35% of the selected reference-rate change for that population and period. It does not mean every account rose by 35% of the Fed’s next move. If the comparison window ended earlier, began later, or covered a different product mix, the calculated beta could differ.

A separate simple-interest example shows why the account rate matters to a saver. On a constant $10,000 balance for one year, before fees and taxes, 0.5% annual interest is $50, while 2.0% is $200. The arithmetic assumes the balance stays unchanged and does not model compounding, changing rates, taxes, fees, or withdrawal timing. It is not a current quote or a promise of return.

Rates are often advertised as annual percentage yields (APYs), which incorporate the stated rate and compounding under the account’s terms. The example above deliberately uses simple annual interest so the multiplication is transparent. Do not compare a simple rate with an APY as if they were the same measure; use the same annualized convention and balance assumptions when comparing offers.

The March 2024 Senior Financial Officer Survey asked bank officers about cumulative deposit betas from March 2022 through March 2024 and expectations through September 2024. Among respondents taking retail deposits, the reported average cumulative retail beta for March 2022–March 2024 was 40%, and the average expected cumulative beta through September 2024 was 42%. The corresponding historical averages were 58% for wholesale operational deposits and 77% for wholesale non-operational deposits. These were bank-officer reports for defined segments; the 42% figure was an expectation at that time, not an observed result or a current consumer rate. The survey received responses from 92 of 98 banks invited.

An earlier Federal Reserve study, published in 2013, examined eleven deposit rates across more than 2,500 branches of about 900 institutions over ten years. It found that deposit rates were downward-flexible and upward-sticky on average, with meaningful differences by product, bank size, and branch. That historical research can help explain why savings rates may lag in a tightening period and move down when market rates fall, but it is not a current rate table.

The 2025 Federal Reserve paper studies bank funding costs across policy cycles and finds that cycle length and size help explain pass-through, with modestly greater transmission in loosening cycles in its aggregate results. It also finds that nondeposit liabilities tend to adjust more than deposits. The studies cover different periods, data, products, and definitions. Taken together, they argue against treating one beta as a permanent property of every account: the result depends on the interval, institution, deposit segment, and method.

Banks pay interest and incur other costs to fund loans, securities, and cash assets. Net interest margin (NIM) broadly compares interest income on earning assets with interest expense on funding, relative to earning assets. If asset yields rise sooner or faster than deposit costs, NIM may widen for a time, all else equal. If deposit costs catch up, nondeposit funding becomes more expensive, or asset yields reset downward, the effect can reverse or be offset.

That “all else equal” condition matters. The timing of loan resets, fixed-rate assets, securities valuations, deposit outflows, hedges, credit losses, operating expenses, and capital needs can all affect bank results. A low retail deposit beta by itself does not show that a bank is more profitable or safer. A high beta by itself does not show that a bank is in trouble. Deposit pricing is only one part of a balance-sheet and risk-management picture.

The Federal Reserve’s Commercial Bank Examination Manual discusses how banks use driver rates and beta assumptions in interest-rate-risk analysis, including the need to consider competition and historical deposit behavior when modeling non-maturity deposits such as savings and money market deposit accounts. Those assumptions are tools for analyzing scenarios; they are not guarantees of how actual customer rates or bank earnings will evolve.

Deposit rates do not necessarily rise and fall symmetrically. In an easing cycle, banks may lower some variable deposit rates as market rates fall, while existing fixed-term CDs keep their contract rate until maturity. In another setting, competitive pressure or deposit outflows may keep selected rates higher for longer. The 2013 study’s evidence of downward-flexible and upward-sticky rates and the 2025 cycle-level results both show why direction and measurement window matter, even though they study different data and definitions.

For a household comparing accounts, a beta is background context, not a shopping score. Compare the actual annualized yield for the balance you expect to hold; whether it is variable or fixed; minimums and balance tiers; monthly fees; withdrawal limits; introductory-rate end dates; CD maturity and early-withdrawal terms; and how quickly an advertised rate can change. Also distinguish a bank deposit account from an investment product such as a money market mutual fund. The Fed’s rate move alone cannot tell you which account fits your needs, and this article is general education rather than personal financial advice.

For related context, bank reserves and customer deposits are different balance-sheet concepts. The sources below describe bank funding transmission and deposit pricing; none provides a live quote for a particular depositor.

Sources: Federal Reserve, “Monetary Policy and Bank Funding Costs” (September 2025); Federal Reserve, “March 2024 Senior Financial Officer Survey Results”; Federal Reserve, “Sticky Deposit Rates” (October 2013); Federal Reserve, Commercial Bank Examination Manual, Section 3000.