The Stagflation Signal
When the Fed can't cut: tariffs, recession risk, and the policy trap
Signal Snapshot — as of 2026-03-18
Core CPI YoY
2.5%
Unemployment
4.4%
Policy Space
0.5
Quadrant
Stagflation
10Y Breakeven
2.4%
The Policy Trap
The Federal Reserve faces a dual constraint: core CPI at 2.5% year-over-year — modestly above the 2% target — while the unemployment rate has risen to 4.4% and our composite recession probability model reads 31%. The effective tariff rate has increased from 2.1% to 11.7% since January 2025, though the extent to which this has passed through to consumer prices is debated (see inflation decomposition below).
The Fed cut three times in late 2025, bringing the target range to 3.50-3.75%. Then it stopped. The January 2026 FOMC held steady. Goldman pushed its expected 2026 cuts to September and December. The market is pricing a pause that could extend indefinitely.
This is a policy bind — not unprecedented in severity, but structurally familiar. Growth is slowing (rising recession probability, weakening labor market), but the inflation constraint limits the standard easing response. The Fed is reluctant to cut into above-target inflation without risking credibility and de-anchoring expectations. The bind is milder than the 1970s-80s stagflation episodes (when inflation exceeded 10%), but it is real enough to constrain the pace and timing of rate cuts in a way that pure recession scenarios would not.
Stagflation Quadrant — Recession Prob vs Core Inflation
Dots colored by subsequent 12m SPY returnCurrent Quadrant
Stagflation
Core Inflation
2.7%
Recession Prob
30.7%
Avg Return
Stagflation: +9.1%
Recession: +9.7%
Goldilocks: +13.0%
Reflation: -1.4%
Stagflation Quadrant Analysis
We construct a two-dimensional classification of the macro environment using recession probability (X-axis) and core CPI inflation (Y-axis). The quadrants are:
- Goldilocks (low inflation, low recession risk): the benign state where policy has maximum flexibility. The economy spends nearly half its time here (47.5% of observations since 2000).
- Reflation (high inflation, low recession risk): growth is strong enough to absorb price increases. The Fed can tighten without triggering recession. Relatively rare (6.0%).
- Recession (high recession risk, low inflation): the straightforward easing case. The Fed cuts, growth recovers, inflation is not a constraint.
- Stagflation (high inflation AND high recession risk): the trap. Cutting risks inflation; holding risks deeper contraction.
The scatter plot maps every monthly observation since 2001 into this space, colored by subsequent 12-month SPY return. Average forward return by quadrant (observations through February 2026; forward returns included only where observable by publication): Goldilocks +13.0%, Recession +9.7%, Stagflation +9.1%, Reflation −1.4%. The benign quadrant is also the best-performing one. Reflation — high inflation with low recession risk, the classic setup for tightening — is the only quadrant with negative average forward returns.
Correction (2026-11-17): the paragraph above replaces one that reported Stagflation as the best-performing quadrant (+11.7%) and Goldilocks the worst (+0.1%), with an explanation about peak pessimism and depressed valuations. That finding was an artifact. The forward-return column in the frozen data was exactly zero for 217 of 301 months (January 2001 – January 2019) because the platform's cached SPY history began in 2019 when the data was frozen. The quadrant labels were correct; only the returns were wrong. The chart and the figures here now use corrected data. See the postmortem at the end of this post.
Policy Space — Fed Funds Rate vs Core Inflation
Monthly since 2000Policy Space
0.50%
Real Rate
1.17%
Fed Funds
3.64%
Core Inflation
2.47%
PS Percentile
81th
Policy Space
We define a heuristic policy space indicator:
Policy Space = (Fed Funds - Core Inflation) - Recession Probability x 2
This is a constructed metric, not an empirically derived model. The coefficient on recession probability (2x) is chosen to produce a scale where values near zero indicate tension. The indicator is useful for directional monitoring, not for precise thresholds.
When policy space is positive, the real rate is positive and recession risk is modest — the Fed can hold without urgency. When policy space approaches zero or turns negative, the combination of low real rates and rising recession risk signals a squeeze.
Current policy space is approximately 0.5: the real rate is positive (3.75% - 2.5% = 1.25%), but the recession probability drag (-0.62) compresses the composite. Further deterioration in the labor market — for example, the Sahm Rule crossing 0.40 — would push policy space toward zero, indicating the bind is intensifying.
Inflation Decomposition — Goods vs Services
Tariff-sensitive breakdown since 2018Goods CPI YoY
1.50%
tariff-sensitive
Services CPI YoY
2.90%
demand-driven
Goods-Services Spread
-1.40%
demand pressure
10Y Breakeven
2.37%
inflation expectations
Tariff-Sensitive vs Demand-Driven Inflation
The inflation decomposition reveals the composition of current price pressures:
- Goods CPI (tariff-sensitive) is running at 1.5% YoY — actually below services inflation and below the 2% target. This suggests that whatever tariff pass-through occurred in 2025 has already been absorbed or is fading from year-over-year calculations. The claim that tariffs are currently adding 70+ basis points to core inflation — widely cited in financial media — may overstate the ongoing effect, even if the level adjustment was real.
- Services CPI (demand-driven) is running at 2.9% YoY — the dominant source of above-target inflation. This is the component the Fed can influence through interest rate policy, and it remains sticky.
The goods-services spread is currently negative (-1.4pp), meaning services inflation exceeds goods inflation. This is the opposite of what a tariff-dominated inflation environment would look like. The inflation problem is primarily domestic and demand-driven, not supply-side — which means monetary policy is the relevant tool, but also means the Fed's hesitation to cut is well-founded.
The 10-year breakeven inflation rate at 2.37% suggests the bond market expects inflation to normalize near target — consistent with tariff effects being transitory at the YoY level, but not offering the Fed immediate relief on current readings.
Implications
The stagflation quadrant position has consequences for our broader analytical frameworks:
- Liquidity regime: the Fed's reluctance to ease means net liquidity remains under pressure. The liquidity z-score at -0.64 is unlikely to improve absent a policy shift. This keeps the Tightening regime probability elevated in our Monte Carlo model.
- Recession probability: if the labor market continues to weaken (Sahm Rule at 0.27, approaching the 0.40 early warning threshold), the composite probability will rise toward 40-50% — deepening the policy bind rather than resolving it.
- Momentum: gold as the sole momentum leader while everything else clusters near zero is consistent with markets unable to price the resolution. (A second sentence here, which read the corrected-away quadrant artifact as evidence that pessimism preceded gains, was removed on 2026-11-17.)
What Resolves the Bind
Three potential exits, in approximate order of likelihood:
- Inflation normalizes mechanically — tariff cost pass-through is a one-time level shift that fades from YoY calculations over 12-18 months. If goods inflation remains at 1.5% and services inflation gradually declines as the labor market softens, core CPI drops below 2.5% and the Fed can resume cutting. This requires patience but no policy change.
- Labor market deterioration forces the Fed's hand — recession probability crosses 50%, the Fed cuts despite above-target inflation, accepting a temporary credibility cost. This is the 2001 playbook.
- Tariff rollback or exemption expansion — reduces goods inflation directly, potentially accelerating the timeline for rate cuts. Politically uncertain but economically the cleanest resolution.
The stagflation signal is not a prediction — it is a description of the current constraint set. The quadrant analysis, policy space heuristic, and inflation decomposition together provide a framework for monitoring which exit materializes.
Postmortem (added 2026-11-17)
Written eight months after publication. Nothing above has been edited except the quadrant paragraph (corrected and marked) and one sentence in Implication 3 that rested on the same error.
Data error disclosed. The frozen quadrant data shipped with 217 of 301 forward-return values equal to exactly zero — every month from January 2001 to January 2019. The platform's cached SPY series began in 2019 when the data was frozen, and the pipeline coerced the unobservable returns to 0.0 rather than null. Averaged in, those zeros dragged Goldilocks (47% of all months, nearly all of them pre-2019) to +0.1% and left Stagflation (rarer, more recent) at +11.7%. The "counterintuitive finding" and its rationale were written on top of that artifact. It passed review because verification checked that the text matched the API, not that the API was right. The root cause — a cache check that let one fresh series mask stale or truncated siblings — was fixed on 2026-11-17 and is documented in the platform review.
Thesis vs. outcome. The March argument was that the Fed was trapped: above-target inflation blocking cuts into a weakening labor market, with three possible exits. None occurred. The Fed hiked.
| March 18, 2026 | November 17, 2026 | |
|---|---|---|
| Fed funds target | 3.50–3.75%, on hold | 3.75–4.00% — raised |
| Unemployment | 4.4% | 4.1% (October) |
| Recession probability | 31% | 25% |
| Sahm Rule | 0.27, "approaching 0.40" | 0.00 |
| Core CPI (YoY) | 2.5% | 2.57% (Sep/Sep; October 2025 is missing from FRED after the shutdown, so no clean Oct/Oct figure exists) |
| Headline CPI (YoY) | — | 3.82% |
| Goods / services CPI | 1.5% / 2.9% (spread −1.4 pp) | 4.76% / 3.31% (spread +1.45 pp) |
| Liquidity z-score | −0.64, Tightening | +0.53, Easing (since Nov 11) |
| Gold | sole momentum leader | −16.7% since April 30; composite Sell |
| Quadrant, by the post's own rule | Stagflation | Reflation |
The inflation-composition argument — that goods inflation sat below services, so price pressure was domestic and tariff pass-through had faded — inverted completely: goods inflation now exceeds services by 1.45 points. The labor-market deterioration the Sahm Rule was meant to flag did not arrive; the rule went to zero. The liquidity regime flipped from Tightening to Easing. By the post's own classification the economy now sits in Reflation, the one quadrant with negative historical forward returns.
What held. The 10-year breakeven stayed anchored (2.37% → 2.2%), consistent with the claim that markets did not expect a permanent inflation shift. Framing the situation as a policy bind rather than 1970s stagflation was right in spirit: the Fed did not cut. It resolved the bind in the one direction the post did not consider.
Assessment. A framework post that mislabeled its own historical evidence and whose forward thesis was wrong on the Fed, the labor market, and inflation composition within eight months. It stays published, corrected, because the point of dating these pieces is to be able to say exactly that.