Bank Gap by State & Metro — Methodology

Two independently sourced numbers, shown together: a real, FDIC-deposit-weighted local weighted APY for that specific geography, and the existing site-wide best available APY (the same top nationwide rate used everywhere else on SwitchWize). The difference between them, applied to a realistic typical household balance for that geography, is the per-household annual gap shown on each page.

For each state or metro, FDIC Summary of Deposits gives the real branch-reported deposit total for every FDIC-insured institution present there. Each institution's share of that local deposit base is used to weight its own currently observed savings rate (from the same rate-tracking pipeline every other page on this site reads). The result is a realistic answer to "what does the typical bank someone in this area actually banks with pay," not a simple average across every bank regardless of size.

Only institutions with both a real deposit-share figure and a current, recently observed rate are included in the weighted average. An institution present in the deposit data but never rate-tracked is excluded from the local rate calculation — its deposits still count toward the geography's total local deposit base (see Coverage below), it just cannot contribute a rate that was never observed.

Source: S&P Capital IQ Pro; SNL Financial Data. Calculations: FDIC. Savings and interest checking account rates are based on the $2,500 product tier, while money market and certificate of deposit rates represent an average of the $10,000 and $100,000 product tiers. Account types and maturities published are those most commonly offered by the banks and branches for which FDIC has data (on-tenor maturities).

Several online-first banks are FDIC-chartered at a single address (their headquarters), so Summary of Deposits attributes their entire nationwide online deposit base to wherever that one address happens to be. Left uncorrected, that would inflate the "local" rate of whichever state hosts that charter with a rate that was never actually a local option — and that same rate is already captured in the nationwide best-available figure.

Institutions with fewer than 4 branches nationally (by the same FDIC data) are excluded from the local-APY weighting for this reason, though their deposits still count toward the local total deposit base. This threshold was set by inspecting real branch-count data, not chosen arbitrarily: every covered institution at 3 or fewer national branches was a recognized online-first bank, with no false positives found against genuinely local or regional banks in the same data.

Rather than a geo-aggregate dollar total (see above), each page shows what the rate gap costs on a typical household balance for that geography's own income level. That balance is not guessed: each geography's real Census median household income is matched to the corresponding national income-percentile bracket in the Federal Reserve's own 2022 Survey of Consumer Finances microdata, and that bracket's real median liquid-balance figure is used — the same method already published and shipped for the State Cash Cost Index, reused here rather than re-derived.

A geography is published only if institutions with both real deposit share and a current observed rate (after the online-charter exclusion above) account for at least 40% of that geography's total local deposit base, across at least 2 distinct institutions. Below that bar, the local weighted APY would not be a trustworthy representation of the local market, so the geography is left out entirely rather than shown with a low-confidence number.

This means coverage is currently partial — 17 geographies are published today out of all 50 states and several hundred metro areas evaluated. That number grows as more banks are added to SwitchWize's own rate-tracking coverage; it is not something this project shortcuts by loosening the gate.

Deposit share and Census data are reference data that changes slowly (FDIC Summary of Deposits is published roughly annually; Census ACS annually) and is refreshed on that same cadence. The local weighted APY, best-available APY, and resulting gap are recomputed daily against the current reference data and SwitchWize's live rate observations, and published as a versioned, checksummed snapshot — each page always reads the latest snapshot, never recomputes on the fly.