AMZU

Current Thesis

AMZU is not a standalone narrative — it is the 2x daily-reset leveraged wrapper on AMZN. The only reason to own AMZU (vs. AMZN common) is to express a short-duration directional bet into a specific catalyst. That catalyst is AMZN Q1 2026 earnings, expected ~2026-04-30 (roughly 10 trading days out). The narrative being leveraged is the AI-hyperscaler capex + AWS re-acceleration + Trainium/Anthropic story — it's mega-cap-AI-infra, not M&A/activism/special-sits as the discovery engine mislabeled. Operator takeaway: dossier exists, theme is wrong, retag and treat as a binary-catalyst tool.

Bull Case

• AWS re-acceleration narrative intact: prior quarters have shown AWS YoY growth stepping up on AI workload pull-through; any print showing AWS >20% YoY with CapEx guide-up is a +6-10% AMZN gap → +12-20% AMZU gap (2x daily).
• Trainium/Anthropic deal is an underwritten kicker: Anthropic's Project Rainier $11B+ Trainium build = visible silicon royalty into 2026-27, gives AMZN a credible "we have our own AI chip" story vs. Google TPU + MS Maia.
• Relative catch-up trade: AMZN has lagged MSFT/GOOGL on the 2025-26 AI re-rating; a strong print closes that discount fast.
• Leveraged ETF AUM flows: AMZU + AMZN-bull ETF AUM has grown meaningfully since the 2024 launch — retail/degen bid into earnings adds gamma.
• Clean option: AMZU is the lazy 2x expression — no margin required, no options Greeks to manage, pure directional.

Bear Case

• Structural decay is the boss-level bear: daily rebalancing of 2x leveraged ETFs causes volatility drag. A flat-to-choppy AMZN over 2-3 weeks can bleed AMZU 3-6% even with AMZN unchanged. Anything >5 trading day hold is fighting the product's design.
• Q1 binary risk is symmetric: AWS deceleration back toward high-teens YoY or a CapEx-guide-hold would print a -5% AMZN move → -10% AMZU gap. Earnings within 3 trading days = rule-engine DEFER by playbook.
• AWS competitive pressure is real: Azure OpenAI + GCP Vertex have been taking net-new AI workload share; Bedrock adoption data matters more than revenue headline.
• Retail/margin squeeze tax: consumer segment remains the drag vs. pure-play AI names — AMZN is not a clean AI story, it's a conglomerate.
• Liquidity risk in the wrapper itself: AMZU ADV is meaningfully lower than AMZN; fills on a fast-tape earnings gap can slip 30-80bps.

Setup & Price Structure

• No price feed attached to this dossier (pipeline didn't pass OHLC), so structural calls deferred pending data.
• Reference asset AMZN: need weekly 20-EMA + weekly 50-EMA to validate trend — check next refresh.
• For AMZU specifically, any trade needs: (a) AMZN holding above its weekly 20-EMA, (b) AMZU ADV >$20M (liquidity gate), (c) realized vol regime on AMZN <35% (otherwise decay tax > edge).
• Pattern watch: AMZU gap-and-go on a beat = rideable 1-3 days; AMZU gap-and-fade = cut same session, no hero hold.

Catalyst Calendar (next 30 days)

• ~2026-04-30 (est.) — AMZN Q1 2026 earnings. Confirm exact date via IR. This is THE binary for AMZU over the next month. Consensus watchpoints: AWS YoY growth, CapEx guide, retail operating margin.
• ~2026-05-01 (post-earnings) — sell-side revision wave (24-48h post-print). If AWS beats, expect cluster of PT hikes feeding narrative acceleration.
• ~2026-05-15 (est.) — AWS re:Inforce or similar event window; check for Trainium/Bedrock product drops that extend narrative.
• 2026-05-13 (approx.) — 13F filings deadline; watch whale AMZN positioning changes.

What Would Change Our Mind

• AMZN Q1 prints AWS YoY growth <18% → cut, narrative cracks.
• CapEx 2026 guide flat or down vs. 2025 → AI-infra narrative breaks, AMZU has no reason to exist in the book.
• Weekly close of AMZN below its 20-EMA → trend invalidation on the reference asset → AMZU stop.
• AMZU daily volume drops below $20M for 5 consecutive sessions → liquidity degraded, wrapper unusable regardless of thesis.
• Discovery of active Anthropic/Trainium contract renegotiation OR AWS losing a named hyperscale customer → negative surprise risk too high to hold leveraged.
• Realized vol on AMZN >40% → decay math flips against the holder even if direction is right.

Correlation Notes

• 1.0 with AMZN on intraday basis (that's the product); slight decay from daily reset mechanics on multi-day holds.
• High correlation to MSFT / GOOGL / META on AI-hyperscaler macro days — a broad AI-infra sell-off drags AMZU disproportionately (2x).
• Sensitive to SOXX / SMH as a second-order read — semis weakness = capex fear = mega-cap hyperscaler repricing.
• Inversely correlated to 10Y yield spikes on duration-trade days (mega-cap AI = long-duration).
• Do not pair-trade AMZU vs AMZN for "arb" — the tracking error isn't exploitable at retail fill quality; it's a wrapper, not a mispriced security.
• Pair risk: if already long AMZN common + AMZU, you're triple-long AMZN. Size the book accordingly — this is the #1 beginner trap with leveraged ETF wrappers.