Early access is open. Free sample datasets are live now. Download a sample

Data product · Early access

Historical Exchange Rates: Daily FX Data Download

Daily exchange rates for every pair we carry, as open, high, low, close or as the official ECB fixing, with the cut-off time written on every row so your month-end numbers match your accountant's.

Coverage

What you get

Item Detail
Product code FX-DAILY
Coverage Major, minor and exotic pairs, plus ECB euro reference rates
Granularity One row per pair per day
History Full history on the Free plan, delayed by one day
Delivery REST time series, CSV and JSON files
Update speed Daily; same-day close from Starter

FX never closes on weekdays, so a "daily close" depends on where you cut the day: 17:00 New York, midnight UTC, or the ECB's afternoon fixing in Frankfurt. We write the cut-off in day_close_time and the origin in source on every row.

Schema

Schema

Field Type Required Description
symbol string yes Pair, e.g. EURUSD
base_currency string yes ISO 4217 code
quote_currency string yes ISO 4217 code
date string yes Trading day, YYYY-MM-DD
open number optional First mid of the day (empty for single fixings)
high number optional Highest mid
low number optional Lowest mid
close number yes Closing mid, or the fixing rate when source is a central bank
average number optional Average mid over the day
close_bid number optional Closing bid
close_ask number optional Closing ask
tick_count integer optional Number of price updates in the day; FX has no central volume, so vendors that show "volume" mean this
day_close_time string yes Cut-off used, e.g. 17:00 America/New_York or 00:00 UTC
source string yes marketdata_composite, or a fixing source such as ecb

Example record:

{
  "symbol": "EURUSD",
  "base_currency": "EUR",
  "quote_currency": "USD",
  "date": "2026-11-18",
  "open": 1.09541,
  "high": 1.09827,
  "low": 1.09488,
  "close": 1.09693,
  "average": 1.09661,
  "close_bid": 1.09690,
  "close_ask": 1.09696,
  "tick_count": 184233,
  "day_close_time": "17:00 America/New_York",
  "source": "marketdata_composite"
}

Sample data

Sample data

Two real samples, free to download now, both from the European Central Bank's euro reference rates.

8 majors: 60 business days of EUR against USD, GBP, JPY, CHF, AUD, CAD, CNY and INR, 480 rows. First 10:

date base quote pair rate
2026-11-18 EUR USD EURUSD 1.0968
2026-11-18 EUR GBP EURGBP 0.8329
2026-11-18 EUR JPY EURJPY 174.85
2026-11-18 EUR CHF EURCHF 0.9243
2026-11-18 EUR AUD EURAUD 1.6294
2026-11-18 EUR CAD EURCAD 1.569
2026-11-18 EUR CNY EURCNY 7.3126
2026-11-18 EUR INR EURINR 102.4905
2026-11-17 EUR USD EURUSD 1.0989
2026-11-17 EUR GBP EURGBP 0.83305

Source: European Central Bank

All currencies: 30 business days of EUR against every currency the ECB publishes, 870 rows. First 10:

date base quote pair rate
2026-11-18 EUR USD EURUSD 1.0968
2026-11-18 EUR JPY EURJPY 174.85
2026-11-18 EUR CZK EURCZK 24.386
2026-11-18 EUR DKK EURDKK 7.4744
2026-11-18 EUR GBP EURGBP 0.8329
2026-11-18 EUR HUF EURHUF 368.4
2026-11-18 EUR PLN EURPLN 4.4118
2026-11-18 EUR RON EURRON 5.3467
2026-11-18 EUR SEK EURSEK 11.143
2026-11-18 EUR CHF EURCHF 0.9243

Source: European Central Bank

The samples use a simple five-column layout (date, base, quote, pair, rate). In the API, the same values come back as close with source set to ecb.

Use cases

Use cases

Accounting and invoicing. Many companies book foreign-currency transactions at the ECB reference rate for the day. Pull source = ecb rows and you have an auditable rate with a named origin.

Backtesting carry and trend strategies. Composite OHLC with a consistent day_close_time gives clean daily returns. Mixing cut-offs from different sources adds noise that looks like signal.

Converting a price history. To restate a US stock in euros, divide its close by the EURUSD close, joined on date.

Code

Code examples

Example: EUR/USD daily history.

curl "https://api.marketdata.co/v1/forex/daily?symbol=EURUSD&start=2015-01-01" \
  -H "Authorization: Bearer $MARKETDATA_API_KEY"

Example: load the free ECB sample and build a USD-based table.

import pandas as pd

ecb = pd.read_csv("ecb-reference-rates.csv", parse_dates=["date"])
wide = ecb.pivot(index="date", columns="quote", values="rate")
usd_based = wide.div(wide["USD"], axis=0)
print(usd_based[["GBP", "JPY", "CHF"]].tail())

Pricing

Pricing

Plan Historical exchange rates
Free, $0 Full daily history for all pairs, delayed one day, 1,000 calls a day; ECB fixings with credit
Starter, $29/mo Same-day close, closing bid and ask, 300 calls a minute
Business, $249/mo Commercial use

Files: the ECB reference rates are free. Composite OHLC history is planned at $19 per pair for all years, or $149 for every pair we carry, sold one-off with a yearly refresh. Prices are early access and may change.

Across 20 providers, many give daily rates away. Paid plans run from $12.50 to $450 a month, median $49.99, one-off files from $50 to $990, and one forex specialist charges £149 a month.

Side by side

How we compare

20 of the 50 providers we studied sell daily FX history. It's easy to find free; what's hard to find is a source that tells you when its day ends. We give the full history away on the Free plan, sell clean files at a listed price per pair, and use the same symbol and date fields as our stock and crypto data.

FAQ

Frequently asked questions

How far back does the history go?

On the Free plan you get the full history we carry, delayed by one day. The ECB's euro rates start in 1999, when the euro launched; composite history start dates per pair will be published before launch.

What time is the daily close?

It's on every row in day_close_time. Our composite uses 17:00 New York, the usual FX convention, and ECB rows use the ECB's own fixing.

Why don't you show volume?

Spot FX has no central exchange, so there's no true volume. tick_count counts price updates instead, which is what other vendors usually mean by FX "volume".

Can I use the ECB sample commercially?

Yes. The ECB allows reuse with credit, so keep the line "Source: European Central Bank" with the data.

Keep reading