Files
OpenFin/backend/services/fred_bootstrap.py
2026-07-01 16:16:51 +02:00

414 lines
15 KiB
Python

"""
FRED historical bootstrap — fetches monthly/quarterly/weekly release data
from FRED's public CSV endpoint (no API key required) and stores it in
the economic_events table with rolling z-score surprises.
"""
import logging
from datetime import date
from typing import Any, Dict, List, Optional, Tuple
import httpx
import pandas as pd
logger = logging.getLogger(__name__)
# ── Series catalog ────────────────────────────────────────────────────────────
#
# transform options:
# None — store raw FRED value as-is
# "yoy_pct" — compute year-over-year %: (val / val_12m_ago - 1) * 100
# "qoq_annualized"— compute annualized QoQ growth: ((val/val_prev)^4 - 1)*100
# "div1000" — divide by 1000 (FRED gives raw count, display in K)
#
# delta_absolute: if True, surprise_pct = val - prev (absolute pp change)
# if False, surprise_pct = (val-prev)/|prev| * 100 (% change)
FRED_SERIES: Dict[str, Dict[str, Any]] = {
"PAYEMS": {
"name": "Non-Farm Payrolls",
"unit": "K",
"freq": "monthly",
"category": "employment",
"ff_currency": "USD",
"higher_is_bullish": True,
"assets": ["SPY", "QQQ", "EURUSD=X", "TLT"],
"zscore_window": 12,
"transform": None,
"delta_absolute": False,
},
"ADP_NFP": {
"name": "ADP Non-Farm Employment",
"unit": "K",
"freq": "monthly",
"category": "employment",
"ff_currency": "USD",
"higher_is_bullish": True,
"assets": ["SPY", "QQQ"],
"zscore_window": 12,
"transform": None,
"delta_absolute": False,
"ff_only": True,
},
"UNRATE": {
"name": "Unemployment Rate",
"unit": "%",
"freq": "monthly",
"category": "employment",
"ff_currency": "USD",
"higher_is_bullish": False,
"assets": ["SPY", "QQQ", "EURUSD=X"],
"zscore_window": 12,
"transform": None,
"delta_absolute": True,
},
"CPIAUCSL": {
"name": "CPI All Items (YoY %)",
"unit": "%",
"freq": "monthly",
"category": "inflation",
"ff_currency": "USD",
"higher_is_bullish": False,
"assets": ["TLT", "GLD", "EURUSD=X", "SPY"],
"zscore_window": 12,
"transform": "yoy_pct",
"delta_absolute": True,
},
"CPILFESL": {
"name": "Core CPI (YoY %)",
"unit": "%",
"freq": "monthly",
"category": "inflation",
"ff_currency": "USD",
"higher_is_bullish": False,
"assets": ["TLT", "GLD", "EURUSD=X"],
"zscore_window": 12,
"transform": "yoy_pct",
"delta_absolute": True,
},
"PCEPILFE": {
"name": "Core PCE (YoY %)",
"unit": "%",
"freq": "monthly",
"category": "inflation",
"ff_currency": "USD",
"higher_is_bullish": False,
"assets": ["TLT", "GLD", "SPY"],
"zscore_window": 12,
"transform": "yoy_pct",
"delta_absolute": True,
},
"FEDFUNDS": {
"name": "Fed Funds Rate (FOMC)",
"unit": "%",
"freq": "monthly",
"category": "monetary",
"ff_currency": "USD",
"higher_is_bullish": False,
"assets": ["TLT", "SPY", "EURUSD=X", "GLD"],
"zscore_window": 12,
"transform": None,
"delta_absolute": True,
},
"ICSA": {
"name": "Initial Jobless Claims",
"unit": "K",
"freq": "weekly",
"category": "employment",
"ff_currency": "USD",
"higher_is_bullish": False,
"assets": ["SPY", "QQQ"],
"zscore_window": 52,
"transform": "div1000",
"delta_absolute": False,
},
"GDPC1": {
"name": "GDP Growth (QoQ Ann. %)",
"unit": "%",
"freq": "quarterly",
"category": "growth",
"ff_currency": "USD",
"higher_is_bullish": True,
"assets": ["SPY", "QQQ", "EURUSD=X", "TLT"],
"zscore_window": 8,
"transform": "qoq_annualized",
"delta_absolute": True,
},
"BAMLH0A0HYM2": {
"name": "HY Credit Spread (OAS)",
"unit": "pp",
"freq": "weekly",
"category": "credit",
"higher_is_bullish": False,
"assets": ["HYG", "LQD", "SPY"],
"zscore_window": 52,
"transform": None,
"delta_absolute": True,
},
"T10Y2Y": {
"name": "Yield Spread 10Y-2Y",
"unit": "pp",
"freq": "weekly",
"category": "rates",
"higher_is_bullish": True,
"assets": ["TLT", "IEF", "SPY", "HYG"],
"zscore_window": 52,
"transform": None,
"delta_absolute": True,
},
"T10Y3M": {
"name": "Yield Spread 10Y-3M",
"unit": "pp",
"freq": "weekly",
"category": "rates",
"higher_is_bullish": True,
"assets": ["TLT", "IEF", "SPY"],
"zscore_window": 52,
"transform": None,
"delta_absolute": True,
},
}
# Series that were renamed/replaced — will be purged from DB on bootstrap
DEPRECATED_SERIES = ["A191RL1Q225SBEA"]
CATEGORIES = sorted({v["category"] for v in FRED_SERIES.values()})
# ── FRED fetch (official JSON API — bypasses CloudFlare) ──────────────────────
#
# The public CSV endpoint (fred.stlouisfed.org/graph/fredgraph.csv) is blocked
# by CloudFlare on server IPs. Use the official REST API instead:
# https://api.stlouisfed.org/fred/series/observations
# Free API key: https://fred.stlouisfed.org/docs/api/api_key.html
_FRED_API_URL = "https://api.stlouisfed.org/fred/series/observations"
def _fetch_fred_api(series_id: str, from_date: str, api_key: str) -> Optional[pd.DataFrame]:
params = {
"series_id": series_id,
"api_key": api_key,
"observation_start": from_date,
"file_type": "json",
"sort_order": "asc",
}
print(f"[FRED API] Fetching {series_id} from {from_date}", flush=True)
try:
resp = httpx.get(_FRED_API_URL, params=params, timeout=30, follow_redirects=True)
print(f"[FRED API] {series_id} HTTP {resp.status_code}", flush=True)
resp.raise_for_status()
data = resp.json()
if "error_message" in data:
print(f"[FRED API] {series_id} API error: {data['error_message']}", flush=True)
return None
records = []
for obs in data.get("observations", []):
val_str = obs.get("value", ".")
if val_str == ".": # FRED uses "." for missing/unreleased values
continue
try:
records.append({"date": pd.Timestamp(obs["date"]), "value": float(val_str)})
except (ValueError, KeyError):
continue
if not records:
print(f"[FRED API] {series_id}: no valid observations", flush=True)
return None
df = pd.DataFrame(records).set_index("date").sort_index()
print(f"[FRED API] {series_id}: {len(df)} rows OK", flush=True)
return df
except Exception as e:
print(f"[FRED API] ERROR {series_id}: {type(e).__name__}: {e}", flush=True)
logger.warning(f"[FRED API] Failed {series_id}: {e}")
return None
# ── Transforms ────────────────────────────────────────────────────────────────
def _apply_transform(df: pd.DataFrame, transform: str) -> pd.DataFrame:
"""Apply a series-level transform before storing. Drops NaN rows introduced."""
if transform == "yoy_pct":
# Year-over-year % change from index level
df = df.copy()
df["value"] = (df["value"] / df["value"].shift(12) - 1) * 100
df = df.dropna()
elif transform == "qoq_annualized":
# Annualized quarter-over-quarter growth rate
df = df.copy()
ratio = df["value"] / df["value"].shift(1)
df["value"] = (ratio ** 4 - 1) * 100
df = df.dropna()
elif transform == "div1000":
df = df.copy()
df["value"] = df["value"] / 1000.0
return df
# ── Z-score ───────────────────────────────────────────────────────────────────
def _compute_zscore_series(values: pd.Series, window: int) -> Tuple[pd.Series, pd.Series, pd.Series]:
"""
Z-score using previous-window only (no look-ahead).
Z = (value - rolling_mean_prev_N) / rolling_std_prev_N
"""
roll_mean = values.shift(1).rolling(window, min_periods=max(4, window // 3)).mean()
roll_std = values.shift(1).rolling(window, min_periods=max(4, window // 3)).std()
z = (values - roll_mean) / roll_std.replace(0, float("nan"))
return z.round(2), roll_mean.round(4), roll_std.round(4)
def _direction(zscore: float, higher_is_bullish: bool) -> str:
if abs(zscore) < 0.5:
return "neutral"
return "bullish" if (zscore > 0) == higher_is_bullish else "bearish"
def _resample_weekly(df: pd.DataFrame) -> pd.DataFrame:
return df.resample("W-FRI").last().dropna()
# ── Main bootstrap ────────────────────────────────────────────────────────────
def bootstrap_fred(
from_date: str = "2020-01-01",
series_ids: Optional[List[str]] = None,
force: bool = False,
api_key: Optional[str] = None,
) -> Dict[str, Any]:
"""
Fetch and store FRED data from from_date to today.
Requires a free FRED API key (fred.stlouisfed.org/docs/api/api_key.html).
If api_key is not passed, reads from DB config key 'fred_api_key'.
"""
from services.database import get_conn, get_config
import json
# Resolve API key
key = api_key or get_config("fred_api_key") or ""
if not key:
return {
"_error": (
"Clé API FRED manquante. "
"Inscris-toi gratuitement sur fred.stlouisfed.org/docs/api/api_key.html "
"puis saisis la clé dans le panel Bootstrap."
)
}
target_series = series_ids or list(FRED_SERIES.keys())
results: Dict[str, Any] = {}
conn = get_conn()
# Purge deprecated series from DB
for dep in DEPRECATED_SERIES:
try:
conn.execute("DELETE FROM economic_events WHERE series_id=?", (dep,))
except Exception:
pass
conn.commit()
for sid in target_series:
meta = FRED_SERIES.get(sid)
if not meta:
logger.warning(f"[FRED bootstrap] Unknown series: {sid}")
continue
if meta.get("ff_only"):
results[sid] = {"skipped": "ff_calendar only — no FRED fetch"}
continue
# Warm-up period: extra years before from_date for transform + z-score
transform = meta.get("transform")
extra_years = 3 if transform == "qoq_annualized" else (2 if transform == "yoy_pct" else 1)
fetch_from = str(date(int(from_date[:4]) - extra_years, 1, 1))
df = _fetch_fred_api(sid, from_date=fetch_from, api_key=key)
if df is None or df.empty:
results[sid] = {"status": "fetch_failed", "count": 0}
continue
# Apply transform on full (warm-up included) dataset
if transform:
df = _apply_transform(df, transform)
if df.empty:
results[sid] = {"status": "transform_empty", "count": 0}
continue
# Downsample continuous daily/weekly series to weekly
if meta["freq"] == "weekly" and len(df) > 200:
df = _resample_weekly(df)
# Z-score on transformed values
window = meta["zscore_window"]
z_series, _, _ = _compute_zscore_series(df["value"], window)
delta_absolute = meta.get("delta_absolute", False)
inserted = 0
skipped = 0
for dt, row in df.iterrows():
ev_date = dt.strftime("%Y-%m-%d")
if ev_date < from_date:
continue # warm-up only, don't store
val = float(row["value"])
z = float(z_series.get(dt) or 0.0)
if pd.isna(z):
z = 0.0
prev_raw = df["value"].shift(1).get(dt)
prev = float(prev_raw) if prev_raw is not None and not pd.isna(prev_raw) else None
if delta_absolute:
# pp / absolute change (for rates, spreads, YoY series)
surprise_pct = round(val - prev, 4) if prev is not None else None
else:
# % change (for levels: NFP, ICSA in K, etc.)
surprise_pct = round((val - prev) / abs(prev) * 100, 2) if prev and prev != 0 else None
direction = _direction(z, meta["higher_is_bullish"])
try:
conn.execute(
"""INSERT INTO economic_events
(event_name, series_id, event_date, actual_value, actual_unit,
forecast_value, previous_value, surprise_pct, surprise_zscore,
surprise_direction, assets_impacted, source)
VALUES (?,?,?,?,?,?,?,?,?,?,?,?)
ON CONFLICT(series_id, event_date) DO """ + (
"UPDATE SET actual_value=excluded.actual_value, "
"previous_value=excluded.previous_value, "
"surprise_pct=excluded.surprise_pct, "
"surprise_zscore=excluded.surprise_zscore, "
"surprise_direction=excluded.surprise_direction, "
"fetched_at=datetime('now')"
if force else "NOTHING"
),
(
meta["name"], sid, ev_date, round(val, 4), meta["unit"],
None, prev, surprise_pct, z, direction,
json.dumps(meta["assets"]),
"FRED/bootstrap",
),
)
if conn.execute("SELECT changes()").fetchone()[0]:
inserted += 1
else:
skipped += 1
except Exception as e:
logger.debug(f"[FRED bootstrap] {sid} {ev_date}: {e}")
conn.commit()
results[sid] = {
"status": "ok",
"name": meta["name"],
"inserted": inserted,
"skipped": skipped,
"total_fetched": len(df[df.index >= pd.Timestamp(from_date)]),
}
logger.info(f"[FRED bootstrap] {sid}: {inserted} inserted, {skipped} skipped")
conn.close()
return results