Skip to contents

boe 0.3.0

Yield curves: historical archive and panel helper

  • boe_curve() gains from, to, frequency, and cache_ttl_h arguments. Setting any of from / to, or frequency = "monthly", routes the request through the BoE archive zips, which extend back to ~1979 for nominal gilts, ~1985 for real, ~2000 for the commercial bank liability curve, and ~2009 for OIS. Default behaviour (from = NULL, to = NULL, frequency = "daily") is unchanged: the function still returns the latest published month from the latest-yield-curve-data.zip endpoint.
  • boe_curve() gains a fifth curve type, "blc" (commercial bank liability curve). BLC is only published in the historical archive zip, so requests for it always route through the archive path regardless of from / to.
  • New boe_curve_panel(curve, measure, frequency, from, to, maturities): wide-format wrapper that returns one row per date and one numeric column per pillar maturity. The default pillar set is c(0.5, 1, 2, 5, 10, 20), which aligns exactly with the BoE half-year grid. Pillars outside a curve’s published range trigger a warning and are dropped.
  • boe_curve() and boe_curve_panel() gain a segment argument. segment = "short" returns the separately fitted short end of the curve (monthly maturity steps from one month out to five years) for every curve type; segment = "standard" (the default) is unchanged. Short-end history extends as far back as the Bank published it (to 1979 for nominal gilts, later for OIS); periods with no short-end sheet are skipped rather than erroring. The panel’s default pillars become c(0.5, 1, 2, 3, 5) when segment = "short".
  • boe_curve_panel() now keeps pillar column labels aligned with the maturities they match when an off-grid pillar is dropped; previously a dropped pillar could shift the labels so a surviving column carried the wrong name.
  • Provenance: boe_tbl queries from boe_curve() now record source = "latest" or "archive" and the source_url so the data carries its own audit trail.
  • Internal: archive zips cache for 30 days (vs. 24 hours for the latest-month zip); per-period workbooks within an archive are concatenated transparently, with content-based maturity-row detection for older layouts.

Fixes: Monetary Policy Report release resolution

  • boe_mpr_forecasts() no longer fails with an HTTP 404 when the latest scheduled release does not exist at the guessed URL. The Bank’s publication month drifts between years (the second 2026 report was published in April, not May), and the data archive filename changed from chart-slides-and-data to charts-slides-and-data during 2025. Release selection now enumerates recent months, verifies each archive exists with a lightweight request, and handles both filename variants, instead of assuming a fixed February / May / August / November calendar.
  • From the April 2026 report the Bank moved to a scenario-based “Scenario Projections Databank” with a transposed layout, which this function cannot parse yet. Automatic selection now skips such releases, falls back to the most recent compatible release, and warns. Requesting a scenario-format release explicitly via month / year raises a clear error rather than a parsing failure. Full support for the scenario format is planned for a future release.

boe 0.2.0

CRAN release: 2026-04-27

New: monetary policy data

  • New boe_mpc_decisions(from, to): history of MPC rate-change events derived from the daily Bank Rate series. Returns date, new rate, previous rate, change in basis points, and direction.
  • New boe_mpc_votes(): full MPC voting record from June 1997, parsed from BoE’s published mpcvoting.xlsx. Long format with one row per (meeting, member) including a dissent flag.
  • New boe_mpr_forecasts(series, month, year): Monetary Policy Report forecast paths for CPI inflation, GDP growth, GDP level, unemployment, and Bank Rate. Parses the Projections Databank workbook from the per-release MPR zip. Defaults to the latest published quarterly release; older releases are accessible via month / year (post-2025 format only).

New: search and discovery

  • New boe_series exported dataset: a 52-row catalogue of every BoE series wrapped by the package, with code, title, category, frequency, unit, start date, and seasonal-adjustment flag.
  • New boe_search(query, category, frequency): keyword + filter search over boe_series. Case-insensitive substring match against title and code.
  • New boe_browse(category, frequency): filter-only view of boe_series. Equivalent to boe_search() with no keyword.

New: yield-curve depth (Anderson-Sleath)

  • New boe_curve(curve, measure): full Anderson-Sleath fitted yield curves at all maturities (typically 0.5 to 25 or 40 years), covering nominal gilt, real gilt, implied inflation, and OIS curves. Both spot and forward measures available where published. Latest month of daily data; archive coverage planned.
  • References Anderson and Sleath (2001, BoE Working Paper 126).
  • Adds readxl to Suggests (used only by boe_curve(); lazily required at call time).

New: cache helpers

Provenance

  • New boe_tbl S3 class. All boe_*() functions now return data frames carrying provenance metadata (series codes, date range, frequency, function called, fetch timestamp). Subclasses data.frame so downstream operations are unaffected.
  • New print.boe_tbl() method shows a one-line provenance header above the data frame body, mirroring the fred_tbl pattern in the fred package.

boe 0.1.2

CRAN release: 2026-03-17

  • Removed non-existent pkgdown URL from DESCRIPTION.

boe 0.1.1

  • Examples now cache to tempdir() instead of the user’s home directory, fixing CRAN policy compliance for \donttest examples.
  • Cache directory is now configurable via options(boe.cache_dir = ...).

boe 0.1.0

CRAN release: 2026-03-12