Monday, 31 August 2026

Market Regime

The 2Y added 10bps on the week to 4.34% while the 10Y gave back 1bp, putting the entire week's rate move at the front end and taking 11bps out of the 10Y–2Y spread. At 0.39% the curve is still positive, but the flattening came from the short end selling off while duration sat still. The 10Y's 1bp decline also conceals a decomposition working against the discount rate: the real leg rose 2bps to 2.42% and the breakeven fell 3bps to 2.31%, so what eased on the week was inflation compensation while the real cost of duration ticked up. Credit ran the other way — high yield tightened 10bps to 2.60%, the second-largest move on the sheet and the only one that argues against the front end. The 2Y continues to print above the 3.63% Fed Funds level stamped 2026-07-01, so this is a repricing of the expected policy path with no realized policy move behind it. One timing caveat: equity inputs are read as of 2026-08-31 while every rate and credit observation carries a 2026-08-28 stamp, so the cross-asset comparison here is assembled across a weekend. The only fresh growth observation is initial claims at 203K, down 4K, stamped 2026-08-22; unemployment, payrolls, Fed Funds and housing starts all still carry 2026-07-01 dates.

Key Macro Reads

Regime Assessment

Measured: volatility sits inside the calm band, credit tightened, and participation carries the model's mixed qualifier. Inferred: the block now contains a disagreement it did not have to resolve while every channel eased together. Participation and the front end lean one way; volatility and high yield lean the other. A middle classification is what falls out of inputs that no longer line up, and the more useful question is which side breaks first.

The equity trend input carries less weight here than its size suggests. SPY standing 7.6% above its 200-EMA is a slow average of the same tape that breadth already samples directly, so it cannot independently confirm participation — it lags it. That leaves breadth and the curve doing the work, and both can deteriorate further without volatility or credit registering anything.

The fundamental side cannot arbitrate. Labor and housing are frozen at their 2026-07-01 stamps and will not update before their next monthly cadence, leaving one weekly claims series as the sole high-frequency growth observation. Between now and those releases, any change in this read will be driven by market-priced inputs, which is a statement about what is observable, not about what is true.

What Would Invalidate

• Breadth returning to the healthy band. 54.3% (531/978) carries the mixed marker and is the input most directly responsible for the current classification. A sustained move back into the band the model treats as healthy — held across consecutive prints, not one session — removes the clearest equity-side support for a middle read.
• The front-end move retracing. The 2Y rose 10bps to 4.34%. A retreat back toward the 3.63% Fed Funds level, restoring the 10Y–2Y spread above where it started the week, would strip out the largest single input change in this block.
• Continued flattening from 0.39%. Another week of the front end selling off at this pace pushes the curve toward inversion and converts it into a clearly dissenting input. One more 11bp week would do it; a 1–2bp drift would not.
• HY widening at comparable speed. 2.60% tightened 10bps. A widening of similar magnitude across two consecutive prints removes the strongest observation currently arguing against the front end, and would push the read toward the risk-off side.
• VIX leaving the calm band. 14.51 sits inside it. A sustained exit — not a single spike — removes the volatility condition and would make the disagreement in this block resolve downward.
• Real-led yield decompression. The real 10Y rose 2bps while the breakeven fell 3bps. If the real leg continues to carry the move while nominals rise, the discount-rate channel works against equities without any change in growth compensation.

Forward Catalysts

• Next weekly claims print. 203K, stamped 2026-08-22 and down 4K, is the only growth observation that can update before the monthly block refreshes. A directional reversal sustained across two or three prints is the earliest available evidence on the labor side.
• Refresh of the 2026-07-01 block. Unemployment (4.1%), nonfarm payrolls (158.9M), housing starts (1,239K) and Fed Funds (3.63%) all carry that stamp. Four of the sheet's inputs are structurally stale until then, and their release is the next point at which fundamentals can contradict a market-priced read.
• Next weekly rate stamps. Whether the 2Y's 10bp move extends or retraces determines whether 0.39% keeps compressing, and it is the single highest-information observation due.
• Whether HY holds near 2.60%. Credit is currently the dissenting input; if it stops tightening, the disagreement described above collapses and the read stops being balanced between opposing channels.

Status

NEUTRAL as of 2026-08-31; flip from the RISK-ON read published 2026-08-28.

--