Central bank policy rates explained

Understand what the central bank sets

A central bank uses its policy-rate setting to influence short-term interest rates and wider borrowing conditions. This can affect spending and inflation over time. The setting is a policy decision. It is not a household borrowing rate or a promise that every market interest rate will move by the same amount.

Unlike a household survey statistic, a policy rate comes from official decisions and operational publications. A policy committee assesses economic conditions and chooses whether to raise, lower, or leave its setting unchanged. The central bank publishes the decision and, where relevant, instructions for implementing it.

The instrument differs across monetary systems. The Federal Reserve specifies a target range for the federal funds rate. The Bank of England publishes Bank Rate as a single rate. A target range has two bounds; it should not be silently converted to a point. An observed overnight market rate is also distinct from the central bank's target.

Distinguish decision dates and unchanged rates

An announcement date identifies when the central bank communicated a decision. An effective date identifies when the setting applies. They can differ. The Federal Reserve's implementation notes, for example, state when operational changes take effect. Use the effective date when the question is which setting applied on a particular day.

An observed date records when a source reports a setting without establishing an effective date. Do not turn that date into an announcement date. Narwhal sets exactly one of effective_on and observed_on, and populates announced_on only when the source shows it.

No change means a decision to keep the setting unchanged. It matters when reviewing decisions even though a chart of rate levels stays flat. A level series alone cannot establish whether a flat interval contains one decision, several decisions, or no meeting. Use the official decision record to interpret meetings; do not infer meeting counts from repeated values.

When comparing countries, identify the instrument and target form first. Also align the applicable dates. Two rates reported on the same calendar day can refer to different decisions or different kinds of policy instrument.

Work through an illustrative range change

All numbers and dates in this worked example are illustrative. Suppose a target range falls from 3.75%–4.00% to 3.50%–3.75%. Both bounds fall by 0.25 percentage points, equivalent to 25 basis points. The range remains 0.25 percentage points wide.

The new midpoint is (3.50 + 3.75) / 2 = 3.625%. That is a calculated summary of the range. It is neither an observed market rate nor a separately announced point target. Preserve both official bounds even if an application also displays the midpoint.

Suppose the illustrative decision is announced on 10 December and takes effect on 11 December. The earlier range still applies on 10 December. If a later meeting leaves the new range unchanged, the setting remains 3.50%–3.75%, but the official decision still belongs in a review of that meeting.

Request a policy-rate observation

Use GET /v1/economics/{country}/central-bank/policy-rate. The optional date parameter makes an as-of request: Narwhal returns the most recent source date at or before the requested date. This example requests USA as of an illustrative date. Inspect the returned date rather than assuming it equals the requested date.

Use /v1/economics/{country}/central-bank/policy-rate/history for history. Its start and end bounds are inclusive. It accepts limit from 1 to 1,000 and cursor for pagination. These are observation requests; an as-of date does not by itself establish what information was publicly known at that time.

curl --request GET \
  --url "https://api.narwhalapi.com/v1/economics/USA/central-bank/policy-rate?date=2025-12-11" \
  --header "Authorization: Bearer $NARWHAL_API_KEY"

Read the target form and source dates

This illustrative response identifies the country, authority, currency, and instrument. A rate_type of range uses lower_bound and upper_bound in percent. A point target instead uses rate and qualifier. Keep the qualifier with the rate so its meaning survives in your application.

Here effective_on is populated and observed_on is null. The null announced_on does not supply an announcement date. Do not fill it from the worked example above or infer it from the effective date. Retain the decimal strings when storing the two bounds.

{
  "country": "USA",
  "authority": "federal_reserve",
  "currency": "USD",
  "instrument": "federal_funds_target_range",
  "rate_type": "range",
  "lower_bound": "3.50",
  "upper_bound": "3.75",
  "unit": "percent",
  "effective_on": "2025-12-11",
  "observed_on": null,
  "announced_on": null
}

Compare the setting with expectations

Open Fed rate odds below to see what prediction markets expect for the next Fed decision. Those expectations answer a different question from the official setting returned by the policy-rate endpoint. Keep the current target and the expected next decision clearly labelled.

Use the Economics API reference for request details. “Inverted yield curve: meaning and shapes” explains how market yields vary by maturity and why they should not be treated as policy targets.