Short Screen
Two-axis short screen across the 48-ticker LOGOS universe (13 deep dives + 35 L1 reviews) plus a 25-name extended market scan · 2026-09-14
Framework: Two Types of Shorts
| Type | Dominant dims | Root cause | Target | Duration |
|---|---|---|---|---|
| Cyclical short | M1 + M3 | Industry supply/demand | Industry / sector | Months ~ 2 years |
| Structural short | M2 + M4 + M5 | Company competitiveness | Single stock | Multi-year |
The core test: peers are thriving while the target is deteriorating — that is a structural short. A cyclical short fades an industry overheated at a point in time; a structural short fades a company compounding its own mistakes.
Cyclical Short Candidates (M1+M3)
Thesis: copper at historic highs, ~37x P/E (peak multiple on peak earnings), energy-transition narrative pricing a cyclical like a growth stock. You are shorting the copper cycle, not the company.
Trigger: copper -15~20% or 3 consecutive months of global manufacturing PMI contraction. Invalidation: tightening supply gap + copper making new highs.
Thesis: a levered mirror of a overheated BTC complex — TTM earnings are a function of the trailing four quarters of BTC price. Short the overheating; MSTR is the vehicle.
Status: 9% of 52-week range (-61% from high) — the left side already played out; poor odds here. Second window: BTC approaching prior highs with mNAV premium re-expanding.
Thesis: IT distribution at an inventory-cycle top — the 2026 AI device refresh is the overheating source. With ~1.6% net margins and heavy working capital, downside elasticity is just as large when the cycle turns.
Trigger: channel inventory build / order momentum fading. Invalidation: AI refresh cycle extending beyond expectations.
Structural Short Candidates (M2+M4+M5)
Thesis: the hottest industry (NVDA/AVGO thriving) vs a company losing $451M/quarter with $180M quarterly revenue carrying a $45.6B market cap (~18x P/S). Peers thriving, company deteriorating — the textbook structural short.
Trigger: sovereign-order delivery misses / widening losses. Risk: self-reinforcing AI narrative, squeezes.
Thesis: the LNG industry in harvest mode (Cheniere thriving) vs VG selling long-term contracts below peer pricing, mired in arbitration with Shell/BP, governed by A/B shares. Share bought with price is not a moat.
Trigger: adverse arbitration ruling / new contracts still priced below peers. Invalidation: favorable settlement + contract repricing.
Thesis: hot warehouse-automation industry vs SYM with Walmart customer concentration, a 2024 accounting restatement, dual-class governance, and a 0.5% net margin. At 9% of range the move already happened — wait for a bounce to short.
Thesis: thriving e-commerce peers (AMZN/Sea) vs CPNG suddenly loss-making — positive through 3Q25, -$840M cumulative in 1H26, with questionable disclosure on the expansion. At 3% of range the left side is gone; short the bounce.
Thesis: the CTV gateway is being squeezed by Netflix/Amazon advertising (peers thriving), ~6% net margin buffer, ad revenue sensitive to the ad cycle. Hybrid cyclical/structural; shorting odds are good at 94% of range.
Thesis: shorting an overheated AI narrative (cyclical) plus a structural profitability defect — 7 years without GAAP profit, 7-8%/yr SBC dilution — while peers CRWD/PANW are profitable. Narrative unwind and structural flaw reinforce each other.
Batch 4: Mid-Cap $5-15B Scan + Data-Driven Structural Screen (2026-09-14)
Extending into mid-caps (550 names) with a data-driven screen: 52-week position deviating ≥25pp below the sector median finds "peers strong / company weak" without prior lists. Sector heat (median position): HC 76% / Fin 64% / Energy 61% / Tech 51% / RE 49% / Cons Disc 49% / Util 41% / Ind 37% / Staples 23%.
New Cyclical Candidate
A trans-Pacific container-freight top (the heavier-asset sister of EXPD): tariff pull-forward and restocking have pushed rates to a peak; the asset-heavy model has no cushion on the way down (Jones Act domestic is a stabilizer — discount the elasticity). Shares triggers with EXPD.
New Structural Candidates
XBRL confirms it: revenue flat at $0.63B for six straight quarters (zero growth) plus five consecutive quarters of declining net income ($150M→$96M), with the stock at 99% of range supported only by buybacks and the dividend. AI-native workflows disintermediating file-collab tools is the structural headwind. A cleaner fake-strength sample than BBY.
Tinder payers have shrunk for years and revenue has rolled over from the 2025Q3 peak ($0.91B→$0.85B); the NI "improvement" ($118M→$171M) is entirely cost cuts; 99.6% of range prices a recovery that does not exist. Take-private rumors are the main event risk (verify borrow).
The textbook structural pairing from the data screen: record ad industry (META/GOOGL/AMZN) vs TTD's declining revenue and compressed NI ($116M→$64M). The break is ongoing (Amazon CTV + agentic buying), but 4.3% of range makes this a late short — waiting for a recovery bounce offers better odds.
A growth stock turning loss-making for the first time: NI swung +$49M→-$92M/-$86M (GLP-1 compounding crackdown + customer-acquisition burn) while revenue still grows. The story degrades from "disrupting healthcare" to "buying revenue," and 30% of range is often not the bottom. Headwinds: thick cash and squeeze-rally risk.
Mid-Cap Crash-Zone Post-mortem (Reverse List)
- Housing/building complex: WMS 1.6% / QXO 1.9% / AAON 2.6% / BLDR 3.7% / IBP 4.3% / POOL 0.7% — the other side of the rate/housing cycle (same source as the large-cap LEN/DHI unwind).
- AI-adjacent narrative crash, wave 2: OKLO 1.4% / BEPC 1.2% / JOBY 1.1% / IONQ 21% / QBTS 13% / RGTI 6% — nuclear, renewables, eVTOL, and quantum all back to earth.
- Consumer-growth crashes: DECK 2.0% / BROS 2.2% / WYNN 1.5% / NCLH 3.1% / CAVA 21%; China ADRs: TME 1.8% / NIO 3.8%. Mostly long material or watchlist — no shorts.
- Screen-noise caveat: deviation screening mixes "sector-wide crash with a worse stock" against true structural pairings — names added above were NI-verified via XBRL; the rest are noted only.
Batch 3: Full-Universe 52-Week Scan (2026-09-14)
No heuristics this time: all 558 US ≥$15B names scanned for 52-week position (9/14 snapshot). 22 entries ≥95% (16 commons after dropping preferreds/notes/MLPs), 21 ≤4% (13 commons); key names cross-checked via SEC XBRL frames.
New Cyclical Candidates
CVX 95.1% / $424B · COP 96.6% / $166B · APA 97.5% · PR 97.8% · midstream MLPs: SUN 98.3% / PAA 97.1% (noted only)
The entire sector sits in the 52-week-high zone — a textbook oil-cycle top (overheating M1 + sector-wide peak margins M3). Short oil, not companies: CVX/COP provide the liquidity. Trigger: OPEC+ increases, inventory builds, oil breaking its prior floor. Invalidation: oil making new highs while energy capex discipline holds.
The shadow play of a global trade-cycle top: freight forwarders earn in sync with trade volumes and freight rates, and XBRL confirms a hot-but-unbroken cycle. 97.8% of range prices continued expansion. Trigger: air/ocean rates rolling over, trade volumes weakening.
CRWD 99.1% / $242.5B · OKTA 99.8% / $32B · FTNT 95.4% / $123B · GEN 97.3% · NET 97.0% (batch-1 candidate, position still rising)
A sentiment-cycle short (valuation dimension): XBRL shows CRWD's GAAP net income just turned positive (~$100M/quarter scale) — $242.5B of market cap implies a four-digit GAAP P/E pricing the "AI security is the next platform" narrative. Demand is real; the price is narrative. Trigger: ARR/NDR deceleration, delayed mega-deals. Invalidation: earnings catch-up compressing multiples back into historical range.
New Structural Candidate
XBRL confirms the "fake strength": latest quarterly revenue $8.77B vs $9.45B a year ago (roughly -7% y/y) with quarterly net income stuck at $0.14-0.32B — fundamentals sliding while the stock prints 99.8%. Thriving peers (WMT/COST/AMZN) plus declining revenue is a textbook structural pairing, with odds near the top of the whole universe. Trigger: negative holiday same-store sales, guidance cuts. Invalidation: revenue returning to growth.
Batch-3 Watchlist & Exclusions
- CNC: 97.2% of range, but XBRL shows quarterly NI back at pre-crisis levels ($1.09-1.54B vs -$6.6B in 2025Q3) — the bounce prices a real recovery; peers are not thriving either, so it fails the structural definition → watchlist.
- VZ / STT / CRBG (95-98%): rates/value positioning trades with defensive character — no short thesis → excluded.
- Crash-zone post-mortem (reverse list, no shorts): TDG 2.3% / ROL 2.4% / IDXX 2.5% / LHX 2.5% / LOW 3.1% / TJX 3.6% / XYL 3.9% — quality compounders collectively at lows (long material or shorts already paid); the utilities cluster (PPL 0% / PEG 0.6% / CMS 1.0% / EXC 1.0% / NRG 1.2%) confirms a second wave of the AI-power unwind.
- Anomaly: HBAN (regional bank) crashed alone to 0.7% — different in nature from the clusters; added to the research list, no conclusion yet.
- Preferreds/notes/MLPs (AGNC series, Southern series, STRK, Duke notes, PPL unit, SUN/PAA) are excluded from the equity short pool.
Batch 2: Extended Market Scan (2026-09-14)
Extending beyond the 48 carded names: 558 US-listed large caps (≥$15B), heuristically prescreened by sector overheating and peer contrast (NASDAQ screener + Yahoo, 9/14 snapshot).
Cyclical Shorts Already Paid (Post-mortem)
AI-power (VST 11% / CEG 21% / TLN 4% / NRG 1%) and homebuilders (DHI 13% / LEN 4%) all sit at 52-week lows — the "electricity is the new oil" and rate-cut trades already unwound. Lesson: when overheating signals appear (top-of-range M1 + M3), the short window usually opens 2-4 quarters after the narrative peaks.
New Cyclical Candidates
The strongest new cyclical short: a memory supercycle top — $1T market cap pricing record DRAM/NAND contract prices (peak earnings + peak multiple). Memory is history's most violent cycle: the moment contract prices flatten, margins and capex both bite back.
Trigger: DRAM/NAND contract prices flattening or rolling over, supply capex accelerating. Invalidation: AI memory demand pushing prices higher for two more quarters.
Peak mining margins at record gold prices: AISC inflation lags spot (top-of-cycle M3 shape), industry-wide supply response is coming. Central-bank buying is a real long-term support — moderate odds; you are shorting gold exuberance, not the company.
Same copper-cycle short as SCCO (sister trade from batch 1): single-asset Grasberg concentration plus Indonesian smelter/export-tax friction are company-level extras. The short target is identical: top-of-range copper.
New Structural Candidates
The 2026 upgrade of the structural short: the hottest industry (NVDA/AVGO/TSM thriving) vs a foundry that still loses money with unproven 18A ramp — yet the stock is up ~4x and $509B prices the turnaround in. You are shorting the narrative's price, with the fundamental crack still open.
Trigger: 18A customer ramp delays / foundry losses not narrowing. Invalidation: binding external customer wins + a clear path to positive foundry gross margin.
Thriving retail peers (WMT/COST making new highs) vs TGT's damaged traffic and margins plus brand-controversy hangover. The 86%-of-range bounce offers good odds — a textbook peers-strong/company-weak pairing.
Thriving payments peers (V/MA/AXP at highs) vs PYPL losing branded-checkout share to Shop Pay/Apple Pay with stalled TPV growth. At ~12x earnings and 38% of range the valuation is already low — the short rides continued structural share loss; moderate odds.
Thriving aerospace peers (GE/HWM/RTX at highs) vs BA's 737/787 ramp and quality issues with ongoing cash burn. At 42% of range with a strong turnaround narrative, odds are moderate — verify borrow cost.
Extended-Batch Watchlist
- HOOD: retail trading volume at a cyclical top ($102B cap), 56% of range — a standard M1/M3 volume-cycle name; wait for heat signals.
- WBD: NFLX thriving vs split/M&A chaos at 85% of range — event-driven with hard borrow; noted only.
- CVS/NKE: real company-level cracks, but their sectors are not collectively thriving (insurance/apparel under their own pressure) — fails the structural definition; dropped.
- COIN/CCJ/AEM/NOC/LMT: 17-56% of range — insufficient exuberance or cycle not yet topped.
Valid Theses, Poor Odds (Watchlist)
- ZM: zero growth vs thriving collaboration peers — but net cash, cheap valuation and buybacks make squeeze risk exceed downside.
- CDE: peers are better positioned in the precious-metals bull market, but a 24% net margin means no company-level crack.
- RKT: rate-cycle play but TTM profitable at 8% of range — poor short odds; falling rates are a refi tailwind.
Short Discipline
- A short is a trade with a time box and defined odds — not the mechanical inverse of a research verdict.
- Every candidate needs all three: entry trigger, invalidation stop, time box (3-18 months cyclical / until invalidation structural).
- Borrow costs, dividend payments, squeezes and M&A risk are not priced into the theses above. This is a screen, not investment advice.
Method: 48 tickers normalized to per-dimension risk ratios — deep dives = yes-count / dimension total; L1 reviews = (5 - score) / 4. Cyclical axis = (M1+M3)/2, structural axis = (M2+M4+M5)/3.