Where the risk is, today.
Nine structural axes — volatility, sentiment, liquidity, credit, recession risk, the fundamental cycle, inflation, real yields and monetary policy — each modelled independently and fused into one composite reading, published every trading day.
On Sep 23, 2026, the composite US market regime reads moderate risk on and has held for 1 trading days; 3 of the nine axes sit on the side of more risk.
Since Sep 23, 2026 · 1 days · further out on the radar = more risk
The system's own forecast of the regime
Macro state as of Sep 23, 2026 · model forecasts of the regime, never of a price
daily switch chance
range 5.2%–15.9%
implied, not forecast
Point forecast with the model's interval
What moved
The close before the week and its five sessions, on the risk ladder
State, confidence and sector tilt
| Axis | State | Confidence | Sector tilt (bps) |
|---|---|---|---|
| Volatility structure | Low | 52% | Consumer Staples +3.8Information Technology -44.6Materials -26.8 |
| Market sentiment | Bearish | 56% | Information Technology -59.1Materials -40.1 |
| Market liquidity | Ample | 60% | Communication Services +10.2Materials +6.8Energy -15.0Industrials -1.1 |
| Credit conditions | Normal | 47% | Consumer Staples +2.1Information Technology -40.1Materials -20.0 |
| Recession risk | Low risk | 58% | Consumer Discretionary +9.3Consumer Staples +7.1Information Technology -35.2Energy -32.3 |
| Fundamental cycle | Expansion | 49% | Consumer Staples +2.1Information Technology -39.0Energy -14.5 |
| Inflation environment | Elevated | 30% | Information Technology +17.3Energy +14.4Consumer Staples -1.2Consumer Discretionary -0.5 |
| Real interest rates | Positive | 45% | Energy +39.7Information Technology +15.5Consumer Discretionary -18.7Real Estate -14.9 |
| Monetary policy | Restrictive | 31% | Information Technology +38.8Energy +31.9Real Estate -2.8Utilities -2.1 |
How US sectors behaved in a moderate risk on market
The composite has read moderate risk on on 22.3% of trading days since Jan 3, 1983, across 156 episodes.
Across those sessions since Dec 22, 1998, Energy did best at +2.9% a year against the S&P 500 and Utilities did worst at -4.7%.
Share of trading days each year, 1983–2026
Annualised excess return vs the S&P 500, next session
Each axis's reading today, in history
| Axis | Today | Share of days | Best sector in this state | Worst sector in this state | page → |
|---|---|---|---|---|---|
| Volatility structure | Low | 35.3% | Information Technology +7.7% | Materials -5.4% | page → |
| Market sentiment | Bearish | 15.1% | Energy +12.3% | Financials -13.7% | page → |
| Market liquidity | Ample | 44.6% | Energy +3.5% | Consumer Staples -3.4% | page → |
| Credit conditions | Normal | 17.2% | Health Care +6.5% | Consumer Discretionary -6.2% | page → |
| Recession risk | Low risk | 40.9% | Information Technology +7.8% | Consumer Staples -4.5% | page → |
| Fundamental cycle | Expansion | 47.5% | Information Technology +5% | Materials -3.8% | page → |
| Inflation environment | Elevated | 40.3% | Information Technology +7.4% | Consumer Staples -9% | page → |
| Real interest rates | Positive | 20.6% | Energy +7.4% | Utilities -9.8% | page → |
| Monetary policy | Restrictive | 27.9% | Energy +4.9% | Utilities -2.1% | page → |
Annualised excess return vs the S&P 500 by regime and sector · today's states outlined
Nine axes, one composite, sector tilts
Nine axes, each read on its own
Each axis has its own indicators, and its state comes from where today's readings sit in their own history, against percentile thresholds set for that axis. Economic releases are placed on the day they were published, so no reading uses data that was not yet out. A hidden Markov model and a gradient-boosted classifier are trained per axis as cross-checks; today's published states and confidences come from the percentile classifier alone.
One composite
The nine readings are fused by a hidden semi-Markov model. Unlike a plain Markov chain it models how long a regime tends to last, and it switches only when the evidence for a new state clears a margin, so the composite does not flip on one noisy day.
Sector tilts
For each axis and each GICS sector, the tilt is the sector's historical sensitivity to that axis: the slope of the sector's forward return on the axis score, estimated over all history up to the day, with bootstrap checks on its significance. It is shown in basis points and says what a move in the axis has meant for the sector, not what it will mean.