This is the weakest cohort on the desk's board this week, and the weakness runs deeper than this week's RSI numbers: $WMT cratered into the oversold band (a 26-point single-scan drop) and $XLU relapsed hard right back toward it, but the real story is how many names in this 15-ticker cut are trading below their own 200-day average with a negative year-over-year return — $WMT, $HD, $MCD, $LOW, and $XLU all fail that test outright. Only $TGT, $SBUX, and $VZ read as genuinely healthy long trends. $PPC's 15.6% weekly rally is a bounce inside a still-broken longer trend (down 32% over the year, below its 200-day), not a new uptrend.

Data as-of the 2026-08-21 close (Friday's settled session) — every name in this cut carries a fresh bar this scan.

Source: defensive.json. Prior scan: 2026-08-14-defensive-scan.

$WMT was this scan's clearest casualty: RSI fell from 55.7 to 29.6 — a 26-point single-scan collapse — dropping into the oversold band for the first time this cut has shown, with price down 10.0% for the week and now 12.5% below its 200-day average. Its year-over-year return is barely positive (+5.9%) against a badly negative alpha (-16.0), which is the tell for a name that's actually broken rather than just having a rough week — the underlying data agrees, tagging its regime "pullback" but with the long-trend numbers pointing the other way from that label.

$XLU told a similar story a scan late: it staged a 12-point recovery two scans ago, and this scan gave essentially all of it back — RSI 48.2 to 32.6, price down 3.5% for the week, now 4.4% below its 200-day with a flat year-over-year return (+2.1%) and a badly negative alpha (-19.7). Both $WMT and $XLU are now trading below their own 200-day average with a return that fails to keep pace with the broader tape — that is a genuine long-trend problem, not a one-week wobble.

The weakness is broader than those two names. $HD, $MCD, and $LOW are all below their 200-day average with negative year-over-year returns (-15.6%, -13.5%, and -15.5% respectively) and alpha readings between -35 and -37 — badly lagging the tape over the past year. $HRL, $COST, and $TSN sit closer to their 200-day but still carry negative or near-flat year-over-year returns with weak alpha. Even $T, whose RSI (67.6) looks strong on the surface, is barely above its 200-day (+0.7%) with a negative year-over-year return (-14.1%) and a badly negative alpha — flat-or-down on the year with lagging alpha is the same tell $WMT and $XLU show, just dressed up in a healthier-looking RSI this week.

The exceptions are real, though: $TGT is up 70% over the year and 39% above its 200-day, still extending at RSI 76.0 — the one unambiguously healthy long trend in this cohort. $SBUX (RSI 55.0, up 21% on the year, 11% above its 200-day) and $VZ (RSI 67.1, up 10% on the year, 9% above its 200-day) both clear the test as well, if less dramatically. $PPC's 15.6% weekly rally (RSI 50.1 to 69.7) reads well on the surface but sits inside a trend that's still down 32% over the past year and below its 200-day — a bounce, not a reversal, until it clears that longer-term ceiling.

• $WMT crashed into the oversold band. RSI 55.7 → 29.6 (-26.1), price $115.27 → $103.70 (-10.0%) — the cut's steepest single-scan move by a wide margin.
• $XLU relapsed almost all the way back. RSI 48.2 → 32.6 (-15.6), price $44.31 → $42.77 (-3.5%) — undoing most of the recovery flagged two scans ago.
• $PPC extended its rally into overbought. RSI 50.1 → 69.7 (+19.6), price $27.90 → $32.25 (+15.6%) — still down 32% over the past year and below its 200-day.
• $TGT pushed further into overbought. RSI 72.3 → 76.0 (+3.7), price up 7.1% for the week, now 39% above its 200-day.
• $HRL cooled further. RSI 46.4 → 38.7 (-7.7), price down 2.7% for the week.
• $COST and $LOW both eased. COST: 55.0 → 49.0 (-6.0). LOW: 53.6 → 50.0 (-3.6).
• $T and $VZ both held up. T: 66.2 → 67.6 (+1.4). VZ: 65.6 → 67.1 (+1.5).
• The oversold band went from empty to holding three names. $WMT (29.6), $XLU (32.6), and $HRL (38.7, close to the line) — versus zero last scan.
• No stale bars this scan — every name carries a fresh 2026-08-21 close.
• Universe unchanged. Same 15 tickers as the prior scan.

RSI measures how stretched a stock is over the past 14 sessions — under 30 signals oversold, over 70 signals overbought. vs SMA20 is the price gap to the 20-day average. 1Y% and vs SMA200 are the numbers that separate a genuine breakdown from a rough week: a name below its 200-day with a flat-to-negative year and a badly negative alpha — like $WMT, $XLU, $HD, $MCD, $LOW, and even the higher-RSI $T — is structurally weak regardless of how this week's RSI reads. This scan is a clean example of why the RSI number alone can mislead: $T's RSI (67.6) looks like one of the cut's healthier prints, but its 1-year return and alpha are both worse than $WMT's, which just crashed into oversold.

None.

$WMT (29.6), $XLU (32.6), $HRL (38.7) — three names, versus zero two scans ago. $WMT and $XLU are both genuinely below their 200-day average; $HRL is technically just above its 200-day but carries a negative year-over-year return.

• Genuine breakdown (below 200-day, negative 1-year): $HD, $MCD, $LOW — three clean breakdowns among the higher-RSI names.
• Flat/negative 1-year despite decent RSI (weak alpha, watch closely): $T, $PPC, $ECL, $TSN, $XLP, $COST — six names where the year-over-year number and alpha undercut a more constructive-looking RSI.
• Genuinely healthy long trend: $SBUX, $VZ, $TGT — the only three names in the cut that clear both the 200-day and the 1-year alpha test cleanly.

• $WMT just crashed 26 RSI points into the oversold band and sits below its 200-day — does it find a floor near its shelf ($102.87), or does the decline deepen?
• $XLU relapsed almost all the way back toward last scan's low — is the utilities weakness a rate story, or something specific to the name?
• $PPC's rally is a bounce inside a trend still down 32% over the year — does it clear its 200-day, or fade back into the range?
• Five of fifteen names in this cut are genuine long-trend breakdowns (below 200-day, negative year) — does the defensive complex broadly stabilize, or does more of the cohort join that group next scan?

• Tape data (price/RSI/SMA/returns): full-scan pre-computed summaries as of the 2026-08-21 close — defensive.
• Prior scan for comparison: 2026-08-14-defensive-scan.