Saturday, August 8, 2026

The Week in Review

Market Regime

RISK-ON the regime engine's latest close-of-day read, carried into the weekend (markets closed).

The regime engine's latest read is RISK-ON, unchanged from the prior read. Nothing in the underlying inputs argues with it. VIX closed at 15.81 calm. Breadth is 65.3% of the tracked universe above its 200-EMA (637 of 975), which is healthy participation rather than a narrow handful of names carrying an index. SPY closed at 768.64 against a 200-EMA of 704.34, +9.1% above trend.

The bond tape did the confirming. The 10Y sits at 4.69% (as of 2026-08-06), down 6bps on the week; the 2Y at 4.25%, down 3bps. The 10Y-2Y spread printed 0.46% on 2026-08-07, down 1bp week-over-week a curve that moved essentially not at all while both ends drifted lower. Breakevens at 2.25% came off 3bps, and the real 10Y at 2.44% came off 3bps. That combination lower nominals, lower breakevens, flat curve shape is a market repricing the term premium modestly rather than repricing growth or inflation.

The most direct risk signal was credit. High-yield spreads closed at 2.71% on 2026-08-06, tighter by 14bps on the week. That is the largest single weekly move in this dataset, and it moved in the direction that corroborates the equity read. When credit tightens into calm vol and broad participation, the three are telling the same story; the interesting weeks are the ones where they diverge, and this was not one of them.

Labor data gave nothing to work with either way. Initial claims at 199K (week ending 2026-08-01) rose 1K noise. Unemployment 4.1% and nonfarm payrolls 158.9M are both July 1 stamps and stale by the standards of a weekly review. Fed funds 3.63% as of July 1. Housing starts 1,427K as of June 1 the oldest series in the set and the one that should carry the least weight in any current read.

What is measured: every figure above. What is inferred: that the credit-tightening plus calm-vol plus healthy-breadth combination describes a market not currently pricing stress. What would invalidate that: HY spreads widening back through the level they came from, VIX pushing out of the calm band, or breadth falling materially below the 65.3% reading while SPY holds its premium to the 200-EMA that last divergence being the one that historically precedes a regime flip rather than following it.

Themes in Motion

Ten themes are tracked. Six carry an ACCELERATING tag, three MATURING, one SATURATED.

Accelerating. Space economy (10 names) and cyclical industrials (10 names) are the two broadest accelerating baskets, which is consistent with the breadth reading accelerating tags concentrated in wide, capital-intensive groups rather than in a narrow tech sleeve. AI enterprise software (6), fintech & consumer credit (6), and travel & leisure (4) fill in the rest. Fintech and consumer credit accelerating alongside a 14bp HY tightening is a coherent pair; if credit spreads reverse, that theme tag is the first one that should be re-examined.

Maturing. Industrial power & grid is the largest theme in the book at 11 names but carries a MATURING tag the model reading a crowded, well-understood story rather than an early one. Mega-cap AI platforms (8) and quantum computing (4) sit in the same tier. M&A & special situations (3) is also MATURING, and three names is too small a set to draw any conclusion from beyond the tag itself.

Saturated. AI chips & memory (8 names) is the only theme flagged SATURATED. The distinction that matters between the mega-cap platforms at MATURING and the chip complex at SATURATED is where the model thinks the marginal dollar of narrative is going, and the chip tag says: not here. ASML and ARM both sit inside that complex, and ASML reports Monday a saturated theme with a live catalyst is where a tag gets tested.

Invalidation for the theme frame generally: a theme that stays ACCELERATING through two consecutive weeks of deteriorating credit is a tag that is lagging its own inputs, not leading them.

Under the Lens

265 dossiers were deep-refreshed this week. The named portion of that refresh runs alphabetically from AAOI through BTDR, and a few clusters are worth calling out with the caveat that a refresh is analytical coverage, not a signal.

Names with a catalyst inside the refresh window. ASML, ASTS, ATRC, BMNR and BTDR were all refreshed this week and report on 2026-08-10. That overlap is the useful part: fresh analysis meeting a dated event two days out.

• ASML (semiconductor lithography) refreshed into a Monday print, inside the only theme flagged SATURATED. The read: a saturated tag means the bar is set by positioning, not by the number.
• ASTS (AST SpaceMobile) sits in space economy, the theme tagged ACCELERATING with the widest name count. Monday's print is the first hard datapoint against that tag.
• BTDR and BMNR both digital-asset-infrastructure names, both reporting Monday. Two names is not a sector read, but they resolve on the same day and will be read together whether or not that is warranted.
• ATRC (AtriCure) medical device, refreshed with a Monday catalyst; unconnected to any of the ten tracked themes, which makes it a cleaner single-name test.

Structural clusters in the refresh. The semiconductor and AI-hardware block is heavily represented: AMD, ARM, ASML, AVGO, AEHR, AXTI, AAOI. The power-and-grid adjacency shows up through AMSC. Enterprise software coverage runs through ADBE, ADSK, ACN, AGYS. Consumer and restaurant names BJRI, BLMN, BRBR and healthcare ABT, BMY, AVTR, ALHC, ATRC, BFLY, AMLX, BIOA, BOLD round out the breadth.

On conviction. The honest statement is that the refresh list as given carries no scores, no rankings and no per-name outputs. Naming a "highest-conviction setup" from a coverage list would be inventing a ranking that the data does not contain. What the data does support: the five names above are where fresh analysis and a dated binary intersect within 48 hours of the open, and that intersection is where the week's work is most immediately testable.

The Week Ahead

Fifteen catalysts land inside the next 14 days, and fourteen of them land on a single day.

Monday, August 9 (T-1d): LGN.

Tuesday, August 10 (T-2d): ALTO, ASML, ASTS, ATRC, BHVN, BMNR, BTDR, CEVA, CODI, CRTO, DDD, DFTX, DJT, DRTS.

That is the structure of the week: one name Monday, then a fourteen-name wall Tuesday. Concentration like that has a mechanical consequence single-name dispersion spikes on one session, and index-level calm can coexist with violent moves underneath it. The 15.81 VIX print is a statement about the index, not about any of these fourteen.

What is being watched into it. First, whether ASML's print moves the SATURATED tag on AI chips & memory in either direction a beat that fails to hold a bid is more informative about the tag than a miss would be. Second, whether ASTS validates the ACCELERATING tag on the widest theme in the book. Third and this is the macro item, not the single-name one whether HY credit holds the 2.71% level after a 14bp weekly tightening, because a credit reversal would put the ACCELERATING tag on fintech & consumer credit under immediate pressure.

What is not on the calendar as given. No Fed meeting, no CPI, no payrolls print appears in the supplied catalyst list. The next scheduled read on labor is the weekly claims series, currently 199K. Treat the absence of macro prints in this window as a statement about the supplied calendar, not a claim that none exist.

Invalidation for the week-ahead frame: if the fourteen Tuesday prints resolve without moving breadth off 65.3% or credit off 2.71%, then earnings dispersion was idiosyncratic and the regime read carries forward untouched. If either aggregate moves, the single-name explanation was the wrong one.

--