October 2026 · monthly · growth × inflation regime · axes + transitions walk-forward validated · slowest input through Aug '26
NO DOMINANT SEASON · DEFLATION 31% / GOLDILOCKS 27%
No dominant season: split between Deflation and Goldilocks (growth flat, inflation falling). Watch which axis breaks first; liquidity neutral.
FINANCING CONDITIONS · 59 · NEUTRAL
Regime Compass
Current position and the 12-month trail
← GROWTH FALLING· · · 12-mo trailGROWTH RISING →
Current odds · soft classification, uncalibrated
DEFLATION
31%
GOLDILOCKS
27%
STAGFLATION
22%
REFLATION
20%
WHERE IT'S HEADING3 MO · VALIDATED
No dominant season, so here is the outlook from each side of the line.
If deflation (31% now)
stays deflation44%
→ stagflation31%
If goldilocks (27% now)
stays goldilocks60%
→ reflation24%
One dot per month. Each input's latest print carries forward until the next one lands, so the current dot uses every input and moves with the daily market ones in between. The ellipse just widens as the top season's share falls; it is a visual cue, not a confidence interval. The faded trail is the last 12 months.
Pro-risk seasons 47% vs defensive 53%, so the split mostly cancels out. Then sized by stress, confirmation and financing.
LAYER 1
Economic Season
Growth and inflation identify the season: level, direction, breadth
Growth Pulse
Is real activity strengthening or weakening?
46/100
◆ FLAT
WEAKNEUTRALSTRONG
BREADTH: 0/6 COMPONENTS AGREE
Building permitsFRED PERMIT · leads 6-9 mo · data through Sep '2617
Yield curve (10y-3m)FRED T10Y3M · level · leads ~12 mo · data through Oct '2636
Temp-help employmentFRED TEMPHELPS · turns before payrolls · data through Sep '2643
Initial jobless claimsFRED ICSA · inverted · leads 3-6 mo · data through Oct '2629
National activity (CFNAI)FRED CFNAI · coincident anchor · down-weighted · data through Sep '2659
Copper/gold ratioYahoo · 6-mo % change · growth proxy · down-weighted · data through Oct '2699
Inflation Pulse
Direction of pressure, not last month's CPI print
41/100
▼ FALLING
DISINFLATIONNEUTRALPRESSURE
BREADTH: 2/5 COMPONENTS AGREE
Sticky-price CPI trendAtlanta Fed sticky CPI (YoY) · 3-mo change · data through Sep '2644
Trimmed-mean PCEDallas Fed trimmed mean (YoY) · 3-mo change · data through Aug '265
Breakeven momentum (10y)FRED T10YIE · 3-mo change · market-implied · data through Oct '2660
Wage growthAvg hourly earnings · 12-mo % growth · data through Sep '2647
Flexible-price CPIAtlanta Fed flexible CPI (YoY) · 3-mo change · data through Sep '2670
LAYER 2
Financing Conditions
Real-economy financing decides whether the season is investable. Distinct from Market Liquidity (BTC lead)
Level vs. Impulse
Two readings, always together
Level
59/100
NEUTRAL
3-month impulse
+8 ▴
Points the level has moved over the last 3 months. A falling level can matter more than a low one that's rising: the level says stay invested, the impulse says whether to keep adding.
Four Channels of Financing
Weighted contributions, not a net-liquidity formula
Stress z = 0.6 × spread-change z + 0.4 × NFCI z, each against the last 10 years. Elevated above 0.2 · High above 0.8 · Crisis above 1.5; caps Low 100% · Elevated 75% · High 50% · Crisis 25%. HIGH and CRISIS also need the spread above its median since 1986 or NFCI above 0, so a fast widening from record-tight spreads tops out at ELEVATED. All of these were set before the backtest ran. Backtested as layer M3 (2000-09-30..2026-09-30): it cut risk in 84 months, 36 of them at HIGH or CRISIS (2000-10..2001-02, 2001-08..2001-10, 2002-07..2002-10, 2007-08..2009-03, 2018-12, 2020-03..2020-05, 2022-06..2022-10), and moved the Sharpe ratio 0.69 → 0.77 and max drawdown -14% → -12%.
Market Confirmation · THE THROTTLE
Are prices agreeing with the macro read?
Trend (S&P vs 200d + momentum)POSITIVE
Momentum6mo 7% · 12mo 12%
ConfirmationPOSITIVE → THROTTLE 85%
Positive 85% · Mixed 65% · Negative 40%. Prices get a veto over the macro story.
FINAL POSITION=NET SEASON ODDS -6+LIQUIDITY ADJ×STRESS CAP 100%×CONFIRM 85%→NEUTRAL
The season odds set the direction (pro-risk minus defensive seasons), liquidity sets conviction, stress caps it, confirmation sizes it.
Asset Class Board
Structural baseline + weather tilt = net posture, not the tilt alone
BASELINE · STAGE TWO: SEPARATION (75)Separation-or-above: overweight hard assets & gold, express duration through TIPS + T-bills, avoid long nominal bonds. The weather tilts around that.
Asset
Baseline
+ Weather
= Net
Conviction
Evidence
#1
Gold
22% Stagflation, the hedge case
Already overweight per the Stage Two baseline. The weather isn't adding here, so hold the structural weight and don't add on the tilt (and don't read the muted tilt as 'don't own').
OVERWEIGHT
SLIGHT OW
OVERWEIGHT
n≈3 · LOW
#2
Commodities · hard assets
needs the 20% Reflation path
Already overweight per the Stage Two baseline. The weather isn't adding here, so hold the structural weight and don't add on the tilt (and don't read the muted tilt as 'don't own').
OVERWEIGHT
SLIGHT OW
OVERWEIGHT
UNTESTED
#3
Cash · T-bills
low stress = cash drag
Already overweight per the Stage Two baseline. The weather isn't adding here, so hold the structural weight and don't add on the tilt (and don't read the muted tilt as 'don't own').
OVERWEIGHT
NEUTRAL
OVERWEIGHT
HIGH
#4
Equities · quality & growth-sensitive
27% Goldilocks + 20% Reflation
NEUTRAL
SLIGHT OW
SLIGHT OW
n≈14 · MED
#5
Duration · TIPS + T-bills (not long nominal)
31% Deflation tail keeps a hedge slot
Structural veto: long nominal bonds are out at Separation. Any duration need expresses through TIPS + T-bills, never long nominal, whatever the weather says.
AVOID LONG NOMINAL
SLIGHT OW
HEDGE · TIPS + T-BILLS
n≈10 · MED
#6
Corporate credit · IG over HY
low stress, carry
SLIGHT UW
SLIGHT OW
SLIGHT UW
UNTESTED
Baseline from Stage Two (structural, unbacktested). The weather tilts are hand-set from the season odds. They are not the weights the backtest below tested, and credit and commodities were never in that backtest at all. Where baseline and tilt disagree, we show both: a muted weather tilt on a baseline-overweight asset means don't add, never don't own.
n is independent episodes, not months. Stagflation covers ~2-3 of them in the postwar data, which is why gold reads low-confidence no matter how good the story sounds. Per-asset confidence intervals are not computed yet.
How the score works
24 series · equal-ish weights · point-in-time
Every input runs the same pipeline: economically relevant rate of change → z-score vs its own history → map to 0-100 → weighted combine. The growth and inflation axes use a robust z (median and MAD over 15 years, so one 2020-sized month can't mute an input for a decade), carry each input's latest print forward until the next one lands (up to 3 months), and smooth over 2 months. Weights are equal or theory-motivated, never fitted to past returns; that's how frameworks die out of sample. Shared series (NFCI, HY OAS, net liquidity) feed Financing Conditions and stress only, never the growth axis, so nothing is double-counted.
Validation status: The backtest builds the axes from the same recipe as this page and runs the pre-registered gates walk-forward, each input counted only after its release date. The leading growth axis beats a coincident CFNAI baseline at its documented 6-12 month lead (rank IC 0.42 vs 0.33 at 12 months). The 3-month transition forecast behind "where it's heading" beats both persistence and the base rate out of sample. The current-regime odds are a different thing: a soft classification with a hand-set temperature, never calibrated against outcomes, so read 31% as “leans this way”, not as a probability. The allocation ladder beat static 60/40: Sharpe 0.46 → 0.80, max drawdown -31% → -9%, with the uplift's 90% bootstrap CI clearing zero. That ladder includes the stress cap (layer M3). The backtest uses revised data, not the figures as first published.
Validation ladder
Each layer has to earn its place out of sample. Real backtest results.
T3 · leading axis vs CFNAI (12mo)
PASS 0.42 > 0.33
Transition Brier · model
0.636
vs persistence (T1)
BEATS 0.93
vs base rate (T2)
BEATS 0.75
M4 vs static 60/40 · Sharpe
0.80 vs 0.46
Sharpe uplift · 90% CI
+0.34 [+0.08, +0.55]
M3 · stress cap · Sharpe
0.69 → 0.77
Current-regime odds
UNCALIBRATED
First-release data test
NOT TESTED
Walk-forward, 313 months (2000-09-30..2026-09-30), each input available only after its release (period end plus publication lag), frozen a-priori weights. The leading growth axis beats a coincident CFNAI baseline at its documented 6-12 month lead (rank IC 0.42 vs 0.33 at 12 months). The 3-month transition forecast behind "where it's heading" beats both persistence and the base rate out of sample. The allocation ladder beat static 60/40: Sharpe 0.46 → 0.80, max drawdown -31% → -9%, with the uplift's 90% bootstrap CI clearing zero. Stress cap tested on Baa-Treasury spread momentum plus NFCI with the level gate, inputs in 312 of 312 months. It capped risk in 84 months (HIGH or CRISIS in 36), and the gate held 6 would-be HIGH/CRISIS months at ELEVATED. Adding it moved Sharpe 0.69 to 0.77 and max drawdown -14% to -12%. Not yet tested on first-release data: this test uses final data with publication lags, so release-timing look-ahead is removed but revision look-ahead is not. The exposure is small by construction: the growth axis leans on the Treasury curve and copper/gold (market data, never revised) and on claims and permits (minimal revision), while the one heavily revised input, CFNAI, is 10% of the axis. Equity is a price-return proxy applied the same way across the ladder, so the relative comparison is fair.
Read this before acting
This gauge classifies the environment; it does not forecast prices. The postwar sample holds only a handful of independent recessions, inflation shocks, and liquidity crises. Every statistic here is drawn from that small set, and the quadrant playbooks are hypotheses under test, not laws. Data arrives late and gets revised.
What would make this wrong: a regime where the link from measured liquidity to asset prices breaks (fiscal dominance, capital controls, a structural money-demand shift; the AI-capex cycle is the live example), or a market driven by a factor these 24 series can't see. The leading axis and the transition forecast pass their pre-registered tests out of sample. Still untested: the current-regime odds, the asset-board tilts, and results on first-release data.
1 monthly reading logged since 2026-10-10. Graded against the season the economy is in three months later. The odds are a soft classification, not calibrated probabilities, so they are logged and not graded as forecasts.