FRED Macro Data

Federal Reserve Economic Data — interest rates, inflation, liquidity, employment, and more as alpha inputs

Overview

ClyptQ integrates with the FRED API (Federal Reserve Economic Data) to provide macro economic indicators as inputs to trading strategies. Macro data is particularly valuable for crypto strategies because crypto markets are highly sensitive to liquidity conditions, interest rate expectations, and risk sentiment.

The FREDCollector supports 38 default indicators across 10 categories, with forward-fill expansion so the latest known value is always available at every timestamp.

Setup

FRED requires a free API key. Get one at https://fred.stlouisfed.org/docs/api/api_key.html.

export FRED_API_KEY="your_api_key_here"

Supported Indicators

Interest Rates & Yield Curve

Series ID	Name	Unit	Frequency
DFF	Federal Funds Rate	%	Daily
DFEDTARU	Fed Funds Upper Target	%	Daily
DFEDTARL	Fed Funds Lower Target	%	Daily
DGS1	1-Year Treasury Rate	%	Daily
DGS2	2-Year Treasury Rate	%	Daily
DGS5	5-Year Treasury Rate	%	Daily
DGS10	10-Year Treasury Rate	%	Daily
DGS30	30-Year Treasury Rate	%	Daily
T10Y2Y	10Y-2Y Treasury Spread	%	Daily
T10Y3M	10Y-3M Treasury Spread	%	Daily

Liquidity & Money Supply

These are critical for crypto — net liquidity (WALCL - RRPONTSYD - WTREGEN) is a key driver of risk asset prices.

Series ID	Name	Unit	Frequency
WALCL	Fed Balance Sheet (Total Assets)	Millions	Weekly
RRPONTSYD	Overnight Reverse Repo (ON RRP)	Billions	Daily
WTREGEN	Treasury General Account (TGA)	Millions	Weekly
TOTRESNS	Total Reserves of Depository Institutions	Billions	Monthly
M1SL	M1 Money Supply	Billions	Monthly
M2SL	M2 Money Supply	Billions	Monthly

Inflation & Expectations

Series ID	Name	Unit	Frequency
CPIAUCSL	Consumer Price Index (CPI)	Index	Monthly
CPILFESL	Core CPI (ex Food & Energy)	Index	Monthly
PCEPI	PCE Price Index	Index	Monthly
PCEPILFE	Core PCE (ex Food & Energy)	Index	Monthly
T5YIE	5-Year Breakeven Inflation	%	Daily
T10YIE	10-Year Breakeven Inflation	%	Daily
MICH	Michigan Inflation Expectations	%	Monthly

Employment

Series ID	Name	Unit	Frequency
UNRATE	Unemployment Rate	%	Monthly
PAYEMS	Total Nonfarm Payrolls	Thousands	Monthly
ICSA	Initial Jobless Claims	Thousands	Weekly
CCSA	Continued Jobless Claims	Thousands	Weekly

Economic Activity

Series ID	Name	Unit	Frequency
GDP	Gross Domestic Product	Billions	Quarterly
INDPRO	Industrial Production Index	Index	Monthly
UMCSENT	Consumer Sentiment (UMich)	Index	Monthly
RSAFS	Advance Retail Sales	Millions	Monthly
HOUST	Housing Starts	Thousands	Monthly

Volatility, Credit & Risk

Series ID	Name	Unit	Frequency
VIXCLS	VIX Volatility Index	Index	Daily
BAMLH0A0HYM2	High Yield Bond Spread (OAS)	%	Daily
BAMLC0A0CM	Investment Grade Bond Spread (OAS)	%	Daily

Dollar, Commodities & Equities

Series ID	Name	Unit	Frequency
DTWEXBGS	Trade Weighted Dollar Index	Index	Daily
DCOILWTICO	WTI Crude Oil Price	USD/barrel	Daily
SP500	S&P 500 Index	Index	Daily
NASDAQCOM	NASDAQ Composite Index	Index	Daily

Symbol Aliases

The FREDSymbolMapper provides human-readable aliases for common series:

from clyptq.data.collectors.fred.symbol_mapper import FREDSymbolMapper

FREDSymbolMapper.to_native("FED_FUNDS_RATE")  # -> "DFF"
FREDSymbolMapper.to_native("CPI")             # -> "CPIAUCSL"
FREDSymbolMapper.to_native("VIX")             # -> "VIXCLS"
FREDSymbolMapper.to_native("YIELD_CURVE_10Y2Y")  # -> "T10Y2Y"

Usage

Direct Collection

from clyptq.data.collectors.fred.collector import FREDCollector
from datetime import datetime

collector = FREDCollector(api_key="your_key")

# Collect specific indicators
data = collector.collect_historical(
    symbols=["DFF", "T10Y2Y", "VIXCLS", "WALCL"],
    start=datetime(2020, 1, 1),
    end=datetime(2024, 12, 31),
)
# data["DFF"] -> DataFrame with 'value' and 'exists' columns

# Filter by category
rate_indicators = collector.get_available_symbols(category="interest_rate")
# -> ["DFF", "DFEDTARU", "DFEDTARL", "DGS1", "DGS2", ...]

In a TradingDataSpec

from clyptq.apps.trading.spec.observation.alternative import MacroIndicatorSpec as MacroSpec

macro = MacroSpec(
    indicators=["DFF", "T10Y2Y", "VIXCLS", "WALCL", "RRPONTSYD", "WTREGEN"],
)
# Produces FIELD:fred:macro:DFF, FIELD:fred:macro:T10Y2Y, etc.

Collect and Save to Storage

collector = FREDCollector(api_key="your_key", storage=my_storage)

results = collector.collect_and_save(
    symbols=["DFF", "CPIAUCSL", "M2SL", "VIXCLS"],
    start=datetime(2015, 1, 1),
    end=datetime(2024, 12, 31),
)
# results -> {"DFF": 2500, "CPIAUCSL": 120, "M2SL": 120, "VIXCLS": 2500}

Resolution and Forward-Fill

FRED data has a FIXED resolution type. Each indicator is released at its natural frequency (daily, weekly, monthly, quarterly). Between releases, the previous value is forward-filled so that at any given timestamp the latest known data is available.

This reflects reality: GDP does not change daily — the market trades on the last known GDP value until the next release.

Using Macro Data as Alpha Inputs

Macro indicators are powerful conditioning variables for crypto alphas. Here are common patterns:

Net Liquidity Signal

class NetLiquidityAlpha(AlphaOperator):
    """Long crypto when net liquidity is expanding."""

    def __init__(self, walcl_input, rrp_input, tga_input, window=20):
        super().__init__(inputs=[walcl_input, rrp_input, tga_input])
        self._liq_window = window
        self._validate_lookback()

    def compute_signal(self, data):
        fed_bs = data[0].value   # WALCL (Fed balance sheet)
        rrp = data[1].value      # RRPONTSYD (reverse repo)
        tga = data[2].value      # WTREGEN (Treasury General Account)

        # Net Liquidity = Fed BS - RRP - TGA
        net_liq = fed_bs - rrp - tga
        liq_momentum = self._delta(net_liq, self._liq_window)

        return self._rank(liq_momentum)

Yield Curve Regime Filter

class YieldCurveRegimeAlpha(AlphaOperator):
    """Reduce exposure when yield curve inverts (recession signal)."""

    def __init__(self, spread_input, close_input, window=10):
        super().__init__(inputs=[spread_input, close_input])
        self._regime_window = window

    def compute_signal(self, data):
        spread = data[0].value   # T10Y2Y (10Y-2Y spread)
        close = data[1].value

        # Trend of yield curve spread
        spread_trend = self._regbeta(spread, self._regime_window)
        ret = self._ret(close)

        # Momentum weighted by yield curve regime
        mom = self._mean(ret, self._regime_window)
        return self._rank(mom * self._sign(spread_trend))

Related Pages

• Data Sources Overview — All supported data types
• Onchain Data — DeFiLlama and RPC collectors
• AlphaOperator DSL — Building alphas with macro inputs