Build a macro regime monitor

A macro regime monitor answers one question from the official record: is the US economy expanding or contracting, is inflation above or below target, and is the yield curve signaling stress? This recipe assembles that read from sovereign sources only - FRED series for growth, inflation, unemployment and the yield spread, the BIS policy-rate dataset as an independent cross-check, and the US Treasury par yield curve for the full curve shape.

What you build

Six requests and one small reduction step produce a JSON summary with three regime axes: growth (real GDP year over year, FRED GDPC1 with the pc1 transform), inflation (CPI year over year vs the 2 percent objective, with unemployment as a labor-side sanity check) and curve (the 10-year minus 2-year spread plus the full daily par curve, cross-checked against the BIS policy rate). All responses use the standard data + meta envelope described in the Quickstart. Replace YOUR_API_KEY with your key.

1. Pull real GDP growth

FRED serves level series by default; the units=pc1 transform converts to percent change from a year ago, which is the growth number a regime read needs.

Response (truncated):

{"data": {"series_id": "GDPC1", "title": "Real Gross Domestic Product",
  "frequency": "Quarterly",
  "observations": [
    {"date": "2026-01-01", "value": 2.68474},
    {"date": "2025-10-01", "value": 1.9893}],
  "transform": {"units": "pc1"}},
 "meta": {"source": "fred"}}

Real GDP is up 2.68 percent year over year in the latest quarter. Positive and near trend: the growth axis reads expansion.

2. Pull CPI inflation

The same transform turns the CPI index into a year-over-year inflation rate.

Response (truncated):

{"data": {"series_id": "CPIAUCSL", "frequency": "Monthly",
  "observations": [
    {"date": "2026-06-01", "value": 3.46353},
    {"date": "2026-05-01", "value": 4.16661}],
  "transform": {"units": "pc1"}}}

Headline CPI is running 3.46 percent year over year, decelerating from 4.17 the month before but still clearly above the 2 percent objective. The inflation axis reads above target.

3. Check the unemployment rate

UNRATE needs no transform - it is already a percent.

Response (truncated):

{"data": {"series_id": "UNRATE", "title": "Unemployment Rate",
  "observations": [
    {"date": "2026-06-01", "value": 4.2},
    {"date": "2026-05-01", "value": 4.3},
    {"date": "2026-04-01", "value": 4.3}]}}

Unemployment at 4.2 percent, down from 4.3 the month before, is consistent with the expansion read from step 1 rather than contradicting it.

4. Pull the 10y-2y yield spread

T10Y2Y is the pre-computed 10-year minus 2-year constant maturity spread, updated daily. A negative value is the classic inversion signal.

Response (truncated):

{"data": {"series_id": "T10Y2Y", "frequency": "Daily",
  "observations": [
    {"date": "2026-07-17", "value": 0.37},
    {"date": "2026-07-16", "value": 0.41}]}}

The spread is positive at 37 basis points, so the curve is not inverted. The curve axis reads normal.

5. Cross-check the policy rate against BIS

An independent institution should confirm the policy stance you infer from FRED. The BIS policy-rate dataset reports the US target rate monthly.

Response (truncated):

{"data": [
  {"REF_AREA": "US", "TIME_PERIOD": "2026-06", "value": 3.625},
  {"REF_AREA": "US", "TIME_PERIOD": "2026-05", "value": 3.625}],
 "meta": {"source": "bis"}}

The policy rate is unchanged at 3.625 percent. Against 3.46 percent CPI inflation the real policy rate is roughly +0.2, a near-neutral stance - useful context when labeling the regime.

6. Confirm the curve shape from the US Treasury

The v2 fixed-income surface returns the full daily par yield curve, letting you verify the spread from step 4 against primary tenors.

Response (truncated):

{"data": {"dataset": "par_yield", "date": "2026-07-17",
  "points": [
    {"tenor": "1M", "yield": 3.73},
    {"tenor": "2Y", "yield": 4.18},
    {"tenor": "10Y", "yield": 4.55},
    {"tenor": "30Y", "yield": 5.06}]},
 "meta": {"source": "us_treasury"}}

The curve slopes upward from 3.73 percent at 1 month to 5.06 at 30 years, and 10Y minus 2Y (4.55 - 4.18 = 0.37) matches the FRED spread exactly. Cross-check passed.

7. Assemble the regime read

Reduce the six responses to one summary. The classification rule is deliberately simple; refine thresholds to taste.

With the values above the monitor prints: growth expansion, inflation above target, curve normal, real policy rate +0.16. Re-run the six requests on a schedule and alert when any axis flips - a negative T10Y2Y print or a GDP pc1 value below zero is the signal this monitor exists to catch.

Going further

Sugra Macro direction - BIS policy rates for every country, ECB datasets, cross-country indicator panels and the rest of the official record.

Rate limits and errors - budget the six calls per refresh against your plan's daily quota.

Quickstart - the response envelope, authentication and rate-limit headers used throughout this recipe.

Build a macro regime monitor

What you build#

1. Pull real GDP growth#

2. Pull CPI inflation#

3. Check the unemployment rate#

4. Pull the 10y-2y yield spread#

5. Cross-check the policy rate against BIS#

6. Confirm the curve shape from the US Treasury#

7. Assemble the regime read#

Going further#

What you build

1. Pull real GDP growth

2. Pull CPI inflation

3. Check the unemployment rate

4. Pull the 10y-2y yield spread

5. Cross-check the policy rate against BIS

6. Confirm the curve shape from the US Treasury

7. Assemble the regime read

Going further