European Central Bank

What You're Getting

ECB Statistical Data Warehouse (SDW) covers euro area and EU member state monetary, financial, and economic data: interest rates (MRO, deposit facility, lending rates), exchange rates, monetary aggregates (M1/M2/M3), balance sheets of monetary financial institutions, bank lending surveys, securities holdings, TARGET2 balances, and HICP inflation. Daily, monthly, quarterly, and annual series. Free, comprehensive, and API-first.

Ingestion Strategy

ECB Data Portal API (SDMX 2.1) is excellent — well-documented, reliable, no auth required. Key flows: EXR (exchange rates), IRS (interest rate statistics), BSI (balance sheet items), ICP (HICP). Request series by dataflow + key (frequency.currency.denomination.type), and filter by startPeriod.

import requests, pandas as pd, io

# Daily EUR/USD exchange rate
url = "https://data-api.ecb.europa.eu/service/data/EXR/D.USD.EUR.SP00.A"
params = {
    "startPeriod": "2020-01-01",
    "format": "csvdata"
}
r = requests.get(url, params=params)
df = pd.read_csv(io.StringIO(r.text))
# Columns: FREQ, CURRENCY, CURRENCY_DENOM, EXR_TYPE, EXR_SUFFIX, TIME_PERIOD, OBS_VALUE

Rate Limits & API Details

• No authentication; no published rate limit — polite use (1 req/sec) is sufficient
• SDMX 2.1 base: data-api.ecb.europa.eu/service
• Dataflow list: /dataflow/ECB/all
• Exchange rates updated by ~3pm CET on business days

Schema Stability

ECB SDMX dataflow codes are very stable. Key dimensions are well-documented. Exchange rate reference currency is always EUR — build your FX pipeline around EUR as base. Historical series go back to 1999 (euro introduction) for most monetary series; some go back further via legacy ERM data.

Data Quality Gotchas

• EUR reference date: ECB rates are reference rates, not transaction rates — mid-market, not bid/ask
• Methodology revision: M3 definition and BSI reporting changed in 2014; use care when joining pre- and post-2014 series
• Business day gaps: daily series have gaps for TARGET2 non-business days — handle missing dates explicitly
• National contributions: euro area aggregates and individual country contributions are separate series

Tool Compatibility

• Python: ecb PyPI package; pandasdmx for full SDMX; direct API calls are straightforward
• R: ECB CRAN package
• Airbyte: no official connector
• SDMX JSON format supported by pandasdmx; CSV format is simplest for warehouse loads

Benchmarking Approach

EUR/USD rate for a known date should match ECB published reference rate to 4 decimal places. Euro area M3 annual growth rate should appear within 30 days of month end. Alert if daily exchange rate update is more than 2 business days late. M3 level for current month should be within 5% of prior month (extreme deviations warrant a data quality check).