Wednesday, 26 August 2026

Market regime

The week's rate move came out of the long end: the 10Y gave up 5bps to 4.64% against 2bps at the 2Y, flattening the 10Y–2Y spread 3bps to 0.47%. The front end barely participated, so this reads as a duration bid and not a repricing of the policy path. The composition matters more than the size — 3bps of the 5 came out of the real 10Y and 2bps out of the breakeven, which is the friendlier decomposition for long-duration equity, since it eases the discount rate instead of marking down expected nominal growth. The nominal long-end yield now splits evenly, 2.32% real against 2.32% of inflation compensation. Credit moved the same direction over the same week, with high yield tightening 5bps to 2.70%, so neither cross-asset channel is currently pulling against the equity inputs. Initial claims fell 6K to 206K, but that print carries a 15 August 2026 stamp and is the only labor observation on the sheet younger than eight weeks — Fed Funds, unemployment, payrolls and housing starts all date to 1 July 2026. FRED market prints are as of 25 August 2026.

Key numbers

Regime assessment

Rates and credit both eased across the week, and the equity inputs sit inside the bands the prior reads healthy. When every classified input agrees, the read carries no information about which input the model is actually leaning on. That only becomes observable when one of them leaves its band, and nothing in this block came close to doing so.

Participation is where the fragility sits. A few percentage points of deterioration — an ordinary rotation, not a shock — is enough to make breadth the first input to fail without help from volatility or credit. Anyone judging the durability of this classification is really judging that number.

The stale block is the second constraint, and it bites in a specific way. If growth turns before the next monthly release, the only real-time evidence on this sheet is one weekly claims series and the credit spread. The rate move eased the discounting side of the equation for duration-sensitive equity; it says nothing about the cash-flow side, and the data that would speak to cash flows is roughly eight weeks behind the market prints sitting beside it.

What would change the call

• Breadth at 59.0% (576/977) retains the model's unconfirmed flag. A fall out of the healthy band makes participation the first equity input to break unaided, and is the cleanest refutation this block offers.
• VIX at 15.21 sits inside the calm band. A sustained move out of that band removes the volatility condition the classification rests on.
• SPY at 765.94 is +7.4% above its 200-EMA of 713.08. Sustained compression of that gap alongside weakening breadth pressures the read on equity inputs alone, with no contribution needed from rates or credit.
• HY at 2.70% tightened 5bps. A widening of similar or larger size across consecutive prints re-establishes credit as a dissenting input.
• The real 10Y at 2.32% fell 3bps while the breakeven at 2.32% fell 2bps. A reversal into rising nominal yields led by the real component works against duration-sensitive equity and is the rate-channel condition to watch.
• The 10Y–2Y spread at 0.47% flattened 3bps. Continued flattening toward inversion turns the curve into a dissenting input, which one week of parallel-ish easing does not establish.

What to watch

• Next weekly claims print. The 206K figure stamped 15 August 2026 fell 6K on the week and is the only high-frequency labor read available. One week is a data point; a reversal of direction across the next two or three is the earliest labor evidence that could contradict the growth side of this read.
• Refresh of the 1 July 2026 block. Fed Funds 3.63%, unemployment 4.1%, payrolls 158.9M and housing starts 1,239K all update on monthly cadence. Until they do, none of them can arbitrate a current growth question, and any claim made from them here would be stale by construction.
• The next real-versus-breakeven split. This week 3 of the 5bps of long-end decline came from the real rate. Whether that persists or reverts to breakeven-led moves determines whether the rate channel keeps working with the equity tape or against it.
• Breadth confirmation. The model's qualifier resolves in one of two directions: sustained improvement removes the unconfirmed flag, sustained deterioration makes participation the failing input. No other input on this sheet is as close to its threshold.

Status

Risk-on since 30 July 2026; consecutive read #16 in the journal.