Prototype v0.2 · Not for trading decisions
XS1 · Weekly Decision Packet

Swing Brief

Market regime
Trending and extended. SPY +10% over its 200-day, volatility low, breadth flat — leadership neither broadening nor narrowing.
Biggest change
First Brief — no prior week to compare against.
Why these three?
MRNA · BB · LBRT — three deep pullbacks inside strong 12-month uptrends. The only names where the setup is still forming rather than already run.
Biggest risk
Not a position — a blind spot. Robinhood state is unverified; a resting stop from the July ABT/MMM trades may still exist. Nothing else is held.
Question of the week
Is energy leadership real, or a one-month bounce inside a three-month downtrend? XLE leads on one month and lags badly on three. This single question governs one of the three themes below.
01

What changed

First Brief — nothing to compare. From next week this section leads, and carries only changes: stops moved, theses broken, concentration shifts, claims resolved. Bad news sorts to the top.
02

Scoreboard

Empty by necessity — this is the first Brief to make claims. From next week: what we said, what happened, and whether your confidence was calibrated.
03

Your book

Fact

Nothing to manage. No open XS1 positions.

BookState
Robinhood — XS1 realUnverified — neither flat nor open is claimed
Ledger A — agentic0 open · ABT +0.50R, MMM +0.06R, closed 7/30
Your discretionary bookFlat since 7/25
Interactive BrokersIBIT, VBIL, VOO — passive, out of scope
Qullamaggie paper test19 open — separate cohort, own rules

One action: get Robinhood verified. It is the only factual gap that could hide a live order.

04

Concentration

Fact

8.3 effective N of 11 nominal  ·  mean pairwise correlation 0.06

This list is genuinely diversified — it is not secretly one trade. Only two pairs move together at all: ADBE/RNG 0.54, both software, and HOOD/MRNA 0.51, shared risk appetite rather than shared business.

05

Market environment

Context
MeasureLevelRead
SPY vs 200-day+10.1%trending, and extended
50-day slope, 20 sessions+1.0%still rising
VIX15.81 · 22nd pctilecalm
Breadth, equal vs cap weight+0.3%flat — no rotation into or out of the average stock
06

Sectors

Context

Descriptive, not predictive.

Leading XLE +1.9% · XLK +0.8% · XLF +0.5%   Lagging XLU −7.2% · XLV −3.0% · XLC −3.0%

Only two sectors are worth a second look — the two whose one- and three-month signs disagree.
Sector1 month3 monthWhat it means
XLE+1.9%−8.9%leading now, badly lagging on the quarter — a bounce, not a trend
XLV−3.0%+6.4%the mirror image — quarterly leader now cooling

Everything else says the same thing on both horizons and needs no interpretation.

07

Consensus

Unvalidated
Participants treat the uptrend as intact — SPY extended, VIX subdued. Flow favours energy, tech and financials on growth-and-liquidity assumptions, while other cyclicals lag. That pattern implies selective risk appetite rather than broad re-acceleration.
NarrativeHeld byWhat would break it
Soft-landing bull intactlong-only, CTA trendgrowth scare or sticky-inflation reprice
Energy can lead without an oil spikesector rotators, commodity desksdemand destruction or inventory build
Tech multiples supported by earningsgrowth and momentum fundsguidance cuts or a real-yield jump

Priced in — compressed volatility · no near-term recession or credit event · continued energy and mega-tech bid · an orderly rates path for financials.

08

Differentiated questions

Unvalidated
QuestionWhy it mattersHow to check
Is energy leadership volume- or multiple-driven?separates a durable cash-flow bid from a squeezeXLE vs oil; capex guides
Are lagging industrials and materials a warning?tests how broad the growth assumption really isXLI/XLB vs ISM
How much soft-landing is priced into financials?leaves little room for credit or NII missescharge-offs, the curve
Does low VIX mask clustered event risk?complacency reverses on one catalystskew, term structure, catalyst calendar
The first question is already half-answered by §06 — XLE's one- and three-month disagreement is exactly the tension being asked about.
09

Watchlist

Attention

Deserving attention. Not expected to outperform. No ranking implied.

Still forming

The highest-value charts this week.

TickerWhy this chart deserves your attention1-month percentile12-1
MRNAFormer leadership now attempting a fresh base after the deepest pullback on the list.
p0
+190%chart
BBA very large 12-month run, now deeply retraced — a base candidate if it stops going down.
p2
+197%chart
LBRTEnergy-services theme worth monitoring, but the medium-term sector evidence is mixed.
p3
+106%chart
HOODSharp pullback in a name with only middling 12-month trend — the base has more to prove.
p3
+7%chart
ADBECounter-trend bounce rather than a classical base — a strong month inside a broken year.
p91
−35%chart
ACIWeak on both horizons with no momentum shape at all — lowest-priority chart here.
p7
−25%chart
Already underway

Management, not entry.

TickerWhy this chart deserves your attention1-month percentile12-1
RNGStrongest one-month move on the list — already extended, so the question is management.
p100
+73%chart
CLFSharp one-month surge with no 12-month trend behind it — momentum without a foundation.
p100
−1%chart
THCQuietly strong on both horizons — the least controversial name on the list.
p98
+30%chart
GMTrending with the lowest volatility here — the calmest chart on the list.
p93
+46%chart
LQDAAn extraordinary 12-month run now cooling — consolidation or distribution?
p75
+305%chart
Read the bars, not the table. Every underway name sits at the top of one-month momentum; every still-forming name except ADBE sits at the bottom while carrying strong 12-month trend. That is the big-move-then-quiet-base shape — and it tells you which charts are entries and which are exits.
Themes — each with its falsifier
ThemeNamesFalsified if
Energy services strengthLBRTXLE's three-month lag persists, or oil rolls over. The one-month lead is currently the only support — and §06 already disputes it.
Software / platform multiplesADBE · RNG · BBXLK leadership narrows to mega-cap, leaving mid-cap software behind.
Selective fintech betaHOODXLF strength proves rate-driven rather than risk-appetite-driven — HOOD needs the latter.
10

What to ignore

Fact
11

Open questions

  1. Robinhood is unverified — a resting stop may exist. Highest-priority gap.
  2. §07 and §08 are unvalidated model output — no independence test has been run yet.
  3. Concentration is correlation-only — no sector, factor, or catalyst decomposition yet.
12

Your decisions

Sealed here at read time, before any outcome exists — take / pass / defer per name, with confidence and one line of reasoning. Not yet built. Until it is, this Brief can be read but not scored.
A

Methodology

Nothing below is needed to use the Brief.

A1 · Why §06 says descriptive, not predictive. Sector rotation tested 1999–2026, ~320 monthly cross-sectional observations, three lookbacks. Spearman IC of trailing-relative-return rank against next-month relative return, orthogonalised against the one-month move. Every lookback returned |t| < 1.0 and residual IC between −0.015 and +0.021 — all below the pre-registered 0.03 threshold.

A2 · Why §10 calls the curated list a source, not a signal. 1,986 nominations, March 2022 – July 2026, modelled against the eligible universe on the same dates, out-of-sample via temporal holdout. Within the eligible universe, three features — one-month return, 12-1 momentum, ATR% — separate nominated from non-nominated names with AUC 0.79 long and 0.76 short, dominated by the one-month return.

A3 · Percentiles. Comparison universe is the 150 most liquid names (price ≥ $10, 20-day dollar volume ≥ $25M). Characteristics recomputed live 6 August 2026.

A4 · Effective N. Eigenvalue concentration of the six-month daily-return correlation matrix, (Σλ)²/Σλ².

A5 · Consensus provenance. grok-4.5, single call, temperature 0.3, cost $0.094, raw response preserved. Single model and single retrieval path — the independence test is approved but unrun, so §07 and §08 carry no validated authority.

A6 · What is not yet measurable. Calibration and decision quality require sealed decision capture (§12), which does not exist.

A7 · Chart links point to Finviz. TradingView deliberately not used.