Williams AO/AC Signals W39 2026 — 24 Signals Across 7 Sectors

Edition 23 · 2026-09-25

W38 Scorecard

🔴 MODEL OUTPUT — Published candidates from last week. No edits. No omissions.

Performance of all Category A candidates named in W38. Prices measured from the last weekly close of W38 (2026-09-18) to the last weekly close of W39 (2026-09-25). Polygon weekly bars — last trading day of the week.

Result key: ✓ escalation · ✗ signal lost · — holding/no position. Entry Reference = last weekly close of W38 (2026-09-18). W39 Close = last weekly close 2026-09-25. Source: radar.db weekly_bars.

Scorecard summary: 0 escalated · 5 lost signal · 8 holding · 4 of 13 positive · Avg Δ: -1.4% · SPY W38→W39: +1.3%

Exit Analysis — W38 Signals

🔴 ANALYSIS — Did the 5 names that lost signal exit on real appreciation or noise?

Summary: 0 real appreciation (0%) · 1 range (20%) · 4 no price support / deterioration (80%)

Notable: TTD (-9.5%) did not lose its signal because the setup resolved — it lost it because the deterioration deepened. Worth monitoring in W39 as a continuation of negative momentum.

Portfolio Tracker

🟢 LIVE POSITIONS — Actual entries with real or paper capital. Updated every week.

Mode: Paper

The Number of the Week

🔴 MODEL OUTPUT

24 active signals across 366 tickers analyzed — 6.6% of the universe.

The headline count nearly doubled from last week — W38 ran just 13 active signals at 3.6% density, the lowest count since W21 (13) and W20 (12). W39's 6.6% is a rebound off that trough, but the composition is what deserves attention: all 24 are S1 setups, with zero S2 degraded and zero pure S2 confirmations. This is the widest S1 field since W30 (26), yet not a single name has printed a full confirmation. In other words, the market re-populated the early-signal stage this week without pushing anything through to conviction. Of the 24, ten sit at structural lows (≤p30), so the fresh interest is concentrated at the depressed end of the two-year (104-week) range rather than in names that have already recovered. A universe expanding back to 6.6% with every signal stuck at the S1 stage is one where setups are re-forming quickly but the follow-through has not arrived.

Follow-Up — W38 Key Names

🔴 MODEL OUTPUT

W39 Candidates — Category A

🔴 MODEL OUTPUT — Algorithm-generated. Not editorial picks.

The 24 tickers the model flagged as priority for W40 monitoring. Ordered by price percentile (lower = more depressed relative to the 104-week (2-year) range).

Decision week: W40.

S2 Signals — Week 39

🔴 MODEL OUTPUT

Pure S2 — Full Confirmation

None this week.

S2 Degraded — 0 Tickers

None this week.

Pre-Radar — Approaching the Signal

🔴 MODEL OUTPUT

53 tickers at structural lows (≤p15) with no active signal yet. Names to watch heading into W40.

The Universe

366 tickers · 13 sectors Sectors covered: XLU, XLI, XLP, XLE, XLF, XLV, XLB, XLY, XLK, XLC, XLRE, IBB, XBI Market reference: SPY

Active signals by type:

• S1 active: 24
• S2 degraded: 0
• S2 pure: 0
• Signal tickers at structural lows (≤p30): 10

The Ticker of the Week — Deep Dive

🟡 ANALYST COMMENTARY — Editorial interpretation. Not model output.

TTD — The Trade Desk (XLC): a signal lost the wrong way.

TTD carried an S2 degraded flag into W39 and came out of the week without a signal at all — but the exit tells the opposite story from what a dropped signal usually implies. The price fell −9.5% over the week, from $13.92 to $12.60, and the model now reads it at p0%, the absolute floor of its two-year (104-week) range. When a name loses its signal, the constructive version is a stock climbing out of the setup on real appreciation. This was the destructive version: the S2D label expired because it only fires on the week the AC crosses zero, and with the AC still positive no S1 could replace it — so the name left the list while price kept falling, not because the setup resolved.

The oscillator confirms the same read. TTD's AC remains positive at +0.97 — it has not turned back negative, so there is no accelerating downside momentum signature to point to from the AC alone. The damage this week came from price, not from a fresh AC reversal: a stock already at structural lows shedding another tenth of its value while the oscillator holds a positive reading. That combination — positive AC, price at p0%, a −9.5% weekly print — is why TTD lands in the pre-radar set for W40 rather than on any candidate list. It is a name where the model has stepped back and is watching for either a genuine base to form at these lows or the AC to roll over and rebuild a signal from scratch. Until one of those happens, there is nothing here to act on — only a name worth watching as one of the communications names at p0% in a pre-radar field led by staples (12) and real estate (10), with communications third (8).

Manager Note — W39

🟡 ANALYST COMMENTARY — Portfolio manager's macro read. Not model output.

Two things happened this week that point in different directions, and squaring them is the whole read. The signal field widened — 24 active names against W38's 13 — while the names we were already carrying went the wrong way. The W38 scorecard closed at 4 of 13 positive, an average of −1.4%, against SPY's +1.3%. The book lagged the tape, and it lagged it while the tape was up. That is the uncomfortable combination: a broadening opportunity set sitting on top of last week's underperformance.

The exit analysis sharpens the point. Five names lost their signal, and none exited on real appreciation. One (LIN) ranged out on the filter; the other four — NNN, LNT, NRG, TTD — broke down. That is 80% of exits leaving on deterioration rather than resolution, TTD the worst of them at −9.5% and now sitting at p0%. When signals are lost this way, the model is not booking wins, it is trimming names that failed to hold. The discipline is working as designed, but the raw material it is trimming is weak.

Where the constructive case sits is at the depressed end. Ten of the 24 active signals are at structural lows (≤p30) — OLN and XYL at p0%, NKE at p1%, NCLH at p3%. Behind those active names sits a much larger pool: the pre-radar set has swelled to 53 names at ≤p15, none of which has triggered a signal. That pre-radar field is heavily tilted toward the defensives and rate-sensitives: 12 in staples (XLP), 10 in real estate (XLRE), 8 in communications (XLC). This is where the pressure is concentrated, and it is where the next wave of S1s will most likely come from if the market keeps rotating away from these groups.

So the posture into W40 is patient, not aggressive. The field is wider than it has been since W30, but it is entirely S1 — no name has confirmed, and the book that ran ahead of this week lagged. There is no reason to force positions into a set that has produced no full confirmations and whose recent exits have been breakdowns. We stay in paper mode, keep the ten structural-low holdings on watch, and let the staples-and-real-estate cluster in pre-radar tell us whether this rebuild in signal count turns into anything that clears the confirmation bar.

Published: 2026-09-25