What (One-liner)

The Taylor (1993) rule-implied policy rate, computed from real FRED data (core PCE inflation, unemployment, fed funds), plus the gap between the actual rate and the rule — so you can see at a glance whether policy is tighter or easier than the canonical benchmark.

How to use

Drag taylor_rule_v1 into Canvas and run — it self-fetches core PCE (PCEPILFE), unemployment (UNRATE), and the fed funds rate (FEDFUNDS) from FRED. No input parameters. Wire it upstream of an FOMC institution_composite to give the policy-rate seats their canonical benchmark, or read it standalone.

Core formula

inflation = (core_PCE[-1] / core_PCE[-13] − 1) × 100        # YoY
implied   = max(0, 2.0 + 1.5·(inflation − 2.0) + 0.5·(4.0 − unemployment))   # ZLB
gap       = fed_funds − implied
signal    = above-rule (gap > 0.25) / below-rule (gap < −0.25) / at-rule

Assumptions & applicability

Assumptions: US constants (r*=2%, target=2%, u*=4%), ZLB=0, core PCE 13-month YoY, Taylor's 1.5/0.5 coefficients.

Applicable: US rate-path context; short-to-medium-term benchmark.

Not applicable: non-US central banks, negative-rate / YCC regimes, forecasting the committee's actual decision.

Input / Output contract

No inputs (self-fetching). Output: {taylor_implied_rate, fed_funds, gap, signal, inflation, unemployment, as_of}. On FRED unavailability (or <13 monthly core-PCE points) it returns {status: error, _error_kind: fetch_failed} — it never fabricates a number.

Known limitations

1. Three hardcoded US constants (r*, target, u*) not exposed.
2. A point-in-time rule, not a rate-path forecast or the committee's reaction function.
3. Distinct from central_bank_v1 (a TypeScript stance/probability operator) — same Taylor core, different inputs/outputs.
4. Needs ≥13 monthly core-PCE observations; otherwise an honest fetch_failed error.