How To Use FXMacroData Through Pinkfish

Pinkfish is a lightweight Python backtester for daily OHLC research, portfolio notebooks, and simple strategy loops. FXMacroData fits into that workflow when the research question starts with daily FX reference rates rather than equities or ETFs.

Pinkfish is deliberately direct: fetch data, add indicators, loop over rows, and call buy or sell at the price your rule chooses. FXMacroData adds the FX reference-rate input so the same style can be used for simple currency research.

Why Pinkfish Works for Daily FX Tests

Pinkfish is strongest when the dataset is small, daily, and easy to inspect. Its repository describes it as a backtester and spreadsheet library for stocks, ETFs, portfolios, and swing-trading strategies. The distinguishing pattern is simple: the strategy decides whether to use the open, close, or another price from the current row.

That maps naturally to daily FX reference rates. The FXMacroData fetch helper reads the val field from the FXMacroData forex endpoint and copies it into Pinkfish's OHLC-style columns. The resulting dataframe behaves like a daily bar series even though the source is a reference-rate series.

Workflow Shape

You can inspect the underlying REST call directly before using Pinkfish. Public examples should send the API key in the X-API-Key header:

curl -H "X-API-Key: YOUR_API_KEY" "https://api.fxmacrodata.com/v1/forex/EUR/USD?start_date=2024-01-01&end_date=2026-06-30&limit=5000"

Step 1: Install Pinkfish

Install Pinkfish from the open-source repository so the FXMacroData fetch helper is available:

git clone https://github.com/fja05680/pinkfish.git
cd pinkfish
python -m venv venv
source venv/bin/activate
pip install setuptools
python setup.py develop

On Windows, activate the environment with venv\Scripts\activate. If you already keep Pinkfish in a research environment, update that checkout before running the new helper.

Step 2: Fetch FXMacroData into Pinkfish

Use fetch_fxmacrodata_timeseries() with a fixed start and end date. Passing api_root is optional, but it makes the public API host explicit.

import os
from pinkfish.fetch import fetch_fxmacrodata_timeseries

eurusd = fetch_fxmacrodata_timeseries(
    "EUR/USD",
    "2024-01-01",
    "2026-06-30",
    api_key=os.environ["FXMD_API_KEY"],
    api_root="https://api.fxmacrodata.com/v1",
    use_cache=False,
)

print(eurusd.tail())

The helper accepts pair strings such as EUR/USD, EURUSD, EUR-USD, or EUR_USD. It normalizes the pair, writes a CSV cache under fxmacrodata-cache, and returns a pandas dataframe indexed by date.

For repeatable research, fetch once with use_cache=False, review the dataframe, then rerun with the cache enabled. That keeps parameter tests from quietly changing because the input window moved.

Step 3: Run a Small Strategy Loop

The snippet below shows the shape of a minimal Pinkfish strategy after the FXMacroData dataframe has been fetched. It adds a simple moving-average regime and trades the reference close when the regime turns positive or negative.

import datetime as dt
import pinkfish as pf

start = dt.datetime(2024, 1, 1)
end = dt.datetime(2026, 6, 30)

ts = pf.select_tradeperiod(eurusd, start, end, check_fields=["close"])
ts["regime"] = pf.CROSSOVER(ts, timeperiod_fast=20, timeperiod_slow=60)
ts, start = pf.finalize_timeseries(ts, start, dropna=True)

tlog = pf.TradeLog("EURUSD")
dbal = pf.DailyBal()

Then loop through the rows and let Pinkfish record trades and daily balances:

pf.TradeLog.cash = 100000

for i, row in enumerate(ts.itertuples()):
    date = row.Index.to_pydatetime()
    close = row.close
    end_flag = pf.is_last_row(ts, i)

    if tlog.shares == 0 and row.regime > 0:
        tlog.buy(date, close)
    elif tlog.shares != 0 and (row.regime < 0 or end_flag):
        tlog.sell(date, close)

    dbal.append(date, close)

Finally, inspect the logs and stats in the normal Pinkfish style:

trade_log = tlog.get_log()
daily_balance = dbal.get_log(trade_log)
stats = pf.stats(ts, trade_log, daily_balance, 100000)

print(trade_log.tail())
print(stats)

Backtest Guardrails

Pinkfish makes it easy to test an idea quickly, which is exactly why the data discipline matters. Treat the FXMacroData fetch as an experiment input, not as a live dependency inside the strategy loop.

If the strategy starts to depend on intraday execution, changing symbol universes, broker events, or live order handling, Pinkfish may no longer be the right layer. Keep Pinkfish for compact daily research, then move the surviving idea into a more complete engine if the execution assumptions become more complex.

Related Links

• Open source FXMacroData integrations
• FXMacroData API quickstart
• FXMacroData MCP server
• FX sessions dashboard

Sources and References

• Pinkfish open-source repository
• Pinkfish fetch module
• Pinkfish FXMacroData fetch test
• FXMacroData OpenAPI schema

With that setup, Pinkfish becomes a fast notebook layer for testing daily FX reference-rate ideas. Fetch the data once, keep the cache stable, and use the result to decide whether the idea deserves a heavier research stack.