Choose the instrument before comparing the numbers
We look for the policy rate or target used by the relevant central bank or monetary authority. A retail deposit rate, a government bond yield or an overnight market average is not a substitute just because it is available from an API. The country details name the selected instrument and link to the source actually used.
Official central-bank and monetary-authority sources come first, followed by permitted international statistics for the same instrument. Some records come through the BIS rather than directly from the national producer. Public secondary sources, including Wikipedia, remain separately labelled references; they do not enter the site's confirmed-current comparisons.
United States: a range and a comparison midpoint
The original federal funds target range is shown before the midpoint. For a single number in comparisons, we calculate (lower bound + upper bound) ÷ 2 and label it as the comparison midpoint. If the source does not provide the range, we do not reconstruct it from a lone number.
The effective federal funds rate, or EFFR, is a separate measure of actual overnight transactions. It is not the arithmetic midpoint of the target range. The Korea–US comparison shows why using the upper bound or midpoint produces different gaps.
Euro area: keep MRO, DFR and MLF separate
The site's representative rate remains the ECB's main refinancing operations rate (MRO). That is our choice for consistency across the directory, not a claim that MRO is the ECB's preferred measure of its policy stance. The deposit facility rate (DFR) is the rate through which the ECB steers its stance. The marginal lending facility rate (MLF) applies to overnight central-bank borrowing against eligible collateral.
These three rates are collected separately and named separately in the country details. We do not derive one by adding an assumed spread to another. See the ECB's definitions and official rate table. Euro-area members share the ECB rate; using the euro alone does not make a country a member.
“Current” describes the evidence we have
A numeric value needs an eligible official source, the right instrument, valid dates and a successful check within the relevant time limit. A failed source check, an expired record or a decision that has not taken effect keeps the value out of current comparisons and numeric rankings. The directory entry itself stays available.
- Automatic official collection: the record must have been successfully checked within 36 hours.
- A direct manual check of current official evidence: the review remains eligible for up to 30 days, subject to the daily-observation limit below. Reading an old publication is not a review of the current rate.
- Daily observation series: BIS observations, Korea's ECOS series and the supported daily feeds for the United States, United Kingdom, Canada, Norway, Sweden and Ukraine also need an observation date no more than 10 days old. This limit applies to manual checks too. The age is measured by UTC calendar dates.
The time limits are the site's operating rules, not guarantees from a data provider. They apply to checks and daily observations. An old decision or effective date can still describe the current rate after a long hold; we do not expire it merely because the central bank has left rates unchanged.
When evidence no longer qualifies, the last successful value and its dates remain as a historical reference. It stops contributing to comparisons, rate ordering and summary statistics automatically. A later successful collection can bring it back only if the new record passes the same checks. Downloading an old observation again does not make it new.
Zero and negative rates are numeric values. Missing data, a source failure, a rate that cannot be publicly verified and a country without a comparable independent policy rate are different states. None is converted to 0%. A country may use another monetary-policy framework without having the single rate this site compares.
A retrieval today can still contain last week's observation
Dates are kept in the form the source supports. We do not fill in a missing decision day from a month label, or use the download date as an effective date.
- Decision date
- The date of the policy decision, when the source identifies it.
- Effective date
- When that decision starts to apply. An announced change can still be in the future.
- Observation date
- The day represented by a time-series value. Repeated observations at the same level do not mean repeated policy decisions.
- Last check or retrieval time
- When the site successfully retrieved and accepted the record. It does not establish that the provider has already published every subsequent decision.
A collection attempt that fails has its own diagnostic time. It does not overwrite the saved value's successful check time. Manually reviewed historical records retain their real review date when an automatic job runs.
The BIS daily policy-rate series is published with a lag. We check its observation date as well as the retrieval time. Two countries can have different valid observation dates; our current comparison is not a reconstructed historical series for a date of your choosing.
What the scheduled checks can catch
Collection is scheduled twice a day, at 09:25 and 18:25 Korean time (00:25 and 09:25 UTC). Test and recovery workflows can also trigger a collection. GitHub may start scheduled jobs late, and providers can publish late or refuse a request. This is a saved-data service, not a live market feed.
Collectors check the expected response structure, series or field identity, numeric format, dates and instrument before accepting a value. Snapshot validation also checks country codes and duplicate entries. Future-effective decisions are kept out of the current comparison until they apply. A blank field or a parsing error cannot silently become zero.
Temporary request failures have bounded retries that respect provider wait instructions. If collection still fails, the last successful record is retained and the failure stays visible. An expired credential or changed data format may need a repair; repeating the request indefinitely would not solve either problem.
Accepted data is used to generate the related saved pages together, and publication checks that the updated site is reachable. Automated source-terms checks compare selected terms text and legal links with a reviewed baseline. A detected difference is a candidate for review, not a finding that a licence was violated or that a change is safe. The checks do not automatically approve new terms.
Read a country comparison as a country comparison
The comparison tool subtracts the second selected country's representative rate from the first: A − B. The unit is percentage points, written as %p on the Korean site. US comparisons use the labelled target-range midpoint. Reversing the countries reverses the sign; it does not change the underlying rates.
The median and the count of countries at or above 10% use only eligible current official numeric records, with the same weight for each country or territory. A confirmed 0% or negative rate is included. Secondary references, historical values, pending decisions and missing numbers are excluded.
Countries sharing a policy rate are still separate directory entries. Several euro-area countries can therefore contribute the same MRO to the calculation. These figures are not an average across independent central banks, an economy-weighted measure or an investment return. The number of searchable countries is also different from the number with eligible current data.
Sources keep their own reuse conditions
We extract permitted rate fields and dates, format them for the site and calculate our own midpoints, gaps and summaries. The explanations here describe FURRY RATES. They are not official translations of provider publications, and a link or attribution does not imply a provider's approval.
BIS data source: BIS Central bank policy rates and the national producers identified in each record. Use is subject to the BIS statistical data terms. Our English explanations are not official BIS translations. No additional charge is made for these statistics, and no BIS endorsement or affiliation is implied.
ECB statistics are identified as ECB data and remain subject to the ECB's reuse policy. Our choice of a representative rate and our comparison calculations are our own.
© 2026 Federal Reserve Bank of New York. Content from the New York Fed subject to the Terms of Use at newyorkfed.org. The federal funds target range is subject to those Terms of Use. The New York Fed is not responsible for publication of the federal funds target range by FURRY RATES, does not sanction or endorse any particular republication, and has no liability for your use.
The full provider notices (Korean, with original licence wording) remain available, along with the source and instrument directory (Korean). The historical Brazil SGS notice below documents the earlier data use.
Korea: ECOS observations and a direct manual check
Source: Bank of Korea Economic Statistics System (ECOS), BOK Base Rate indicator. The BOK-produced series 722Y001 / D / 0101000 reports daily observations in annual percent. The 24 October 2026 observation of 3% was directly checked in the official UI on 26 October at 12:54:51 UTC. It takes precedence over the older 8 October BIS observation; older BIS data cannot overwrite it. The observation is not a new decision or effective date.
Automatic ECOS collection is not active: no formal API key is configured. The prepared server-side route requires an issued key under the Open API terms, rejects the public sample key and excludes credentials from public files. Until then, BIS and eligible time-limited manual evidence remain available. Manual ECOS records also expire when the observation exceeds ten days. The former BOK homepage collector stays disabled.
The ECOS statistical-information use guidelines in the footer permit reuse, processing and redistribution of BOK-produced statistics, including commercial use, with attribution and without arbitrary alteration or distortion. We extract the rate and observation date, reformat their display and add our own explanations and comparisons. We reproduce no institutional prose, images or logos, and imply no BOK endorsement.
Historical Brazil SGS 432 data
From 19 September to 13 October 2026, the site used Banco Central do Brasil's SGS 432: Copom Selic target rate. These are daily observations in annual percent, not Copom decision dates. Direct SGS collection is disabled. The current Brazil series comes separately from BIS.
The historical SGS data comes from Banco Central do Brasil and is subject to the Open Database License 1.0 (ODbL). The full licence text remains available. No BCB endorsement is implied.
Since 24 October 2026, the retained SGS data and processing method have been excluded from public distribution and kept privately. They are not used for current collection, comparisons or rankings. Private retention does not establish that obligations arising from past use have ended or been met.
Current BIS data, other providers' material, articles, software, images and marks retain their respective terms. The Korean notice describes the same change.
Map, font and library licences
The map geometry is based on Natural Earth's public-domain data. The following notices cover the webfont, map conversion and libraries:
- Nunito — SIL Open Font License 1.1
- D3 — ISC licence
- TopoJSON — BSD 3-Clause licence
- world-atlas — ISC licence
Each licence applies to the named material. This page does not grant a blanket licence over FURRY RATES articles, code, branding or original artwork.
Reporting a number or explanation that looks wrong
Please include the country or page, the value and dates shown, and the official source you are comparing it with. The contact page (Korean) explains where to send a correction. We check the instrument and timing before changing a record. Our editorial policy describes how we prepare the guides and use writing tools.