Reference
Analysis & playbooks
How pros trade specific event types — not just patterns, but catalysts like ADR arb and capacity bottlenecks.
e.g. SKHY Jul 29 — ADR ↔ Korean share conversion opens
When ADRs trade at a large premium vs local shares, arbitrageurs convert once the window opens. Premium compresses — often via ADR selling pressure, not fundamental deterioration.
Typical play
- Don't chase into the event — mechanical selling is common pre/post
- Watch premium % daily as conversion date approaches
- If thesis intact (HBM demand), scale in on post-event dip (-10–20%)
- Set limit orders below current price; avoid market orders on event day
Avoid when: Premium already compressed, or fundamental thesis broke (guidance cut, sector rotation)
Tickers: $SKHY
e.g. Kimi K3 paused subscriptions — ran out of GPU capacity
When AI labs hit compute ceilings, they must buy more hardware. Bottleneck ripples: GPUs → HBM memory → power delivery → liquid cooling → optical networking.
Typical play
- Map the bottleneck layer (memory vs cooling vs networking)
- Buy the supplier with longest contract visibility (MU > commodity memory)
- VRT benefits on cooling/power lag — often catches up after memory moves
- Hold through volatility; these are multi-year capex cycles
Avoid when: One-off PR with no order backlog confirmation, or stock already extended 50%+ on same narrative
Tickers: $MU, $VRT, $SKHY, $AAOI
e.g. VIX ≥ 25 — Jun 9-11 style correction entry
Elevated VIX = market pricing fear. Quality names with locked-in demand (HBM contracts) often recover faster than broad market in AI infra cycles.
Typical play
- Deploy per shopping list allocation when VIX ≥ 25 (watch band 25–30, allocate 30+)
- Phase in over 2-3 days — don't lump sum on day 1
- Start with highest conviction (MU), then VRT, then SKHY
- Keep 7% cash for post-arb opportunities
Avoid when: VIX spike from single-stock event unrelated to your thesis, or during FOMC/CPI binary within 24h
Tickers: $MU, $VRT, $SKHY, $SHAZ, $AAOI, $VIX
e.g. MU Q4 — HBM revenue mix and forward guidance
For infra suppliers, guidance matters more than EPS beat. Raised guidance = demand visibility. Lowered = thesis review, not automatic exit.
Typical play
- Read guidance section first, not headline EPS
- Raised HBM/capacity guidance → hold or add on post-earnings dip
- In-line + stock down 5%+ → often overreaction for quality infra
- Lowered guidance → trim position, reassess 1-sentence thesis
Avoid when: Initiating new swing 2 days before earnings unless earnings IS the thesis
Tickers: $MU, $VRT, $AAOI
e.g. Insider shares unlock — float increases
Post-IPO lock-up expiry increases tradable float. Insiders may sell, creating short-term pressure before price discovery stabilizes.
Typical play
- Wait 3-5 trading days post-expiry for selling pressure to clear
- Buy only if price holds key support + thesis intact
- Use smaller position size than core infra holds
Avoid when: Buying day-of expiry into known large insider overhang