AI investment doubts trigger a global chip stock selloff — short the semiconductor slide

Verdict: compelling setup on SMH, but fighting cash-flow titans

This short thesis carries a live, actionable signal on SMH — the ETF sits $4.22 below its lower Bollinger Band at $587.88 with ADX at 35.1, well above the 25 threshold — and the idea cites genuine contagion news: per the Yahoo Finance piece on July 16, Marvell dropped 8% on AI capex fears, with selling spreading to Broadcom, AMD, and Intel. The strongest support is that the backtest produced a 311% cumulative return across 189 trades over 60 months, showing the rule set can capture sector-wide panics when they occur. The strongest headwind is fundamental: AMD grew revenue 34.3% year-over-year to $34.6B and Broadcom grew 23.9% to $63.9B, both generating record free cash flow — shorting companies with this balance-sheet firepower means betting against businesses that can absorb a demand shock without breaking. The 47.1% win rate across the full backtest sample and the 29.4% maximum drawdown mean the strategy bleeds frequently, leaning on a small number of large wins during genuine meltdowns. No robust parameter sensitivity setup was established, so no adjusted configuration is recommended — the published thresholds stand as-is. **Conviction breakdown:** - **Thesis support (60):** Real contagion news supports the narrative, but AMD and AVGO entry conditions remain untriggered — AMD's ADX at 15.6 sits 9.4 points below threshold, suggesting the panic may be isolated to the ETF level. - **Trade readiness (75):** SMH has both core conditions met right now; AMD and AVGO are waiting for their entry conditions, with AVGO closest at an ADX of 23.3, just 1.7 points shy. - **Risk quality (50):** The 2.4% hard stop caps individual losses, but the strategy's 29.4% historical drawdown and sub-50% win rate mean sustained drawdowns are a structural feature. - **Backtest evidence (55):** The 311% return is impressive but built on a 47.1% win rate with only 23 trades in the tighter 24-month window — exit fills were approximated on daily bars, which may overstate realized quality. - **Fundamentals trend (35):** Both target companies show accelerating revenue growth, peer-leading margins — AVGO's operating margin at the 98th percentile — and record free cash flow, making them inherently dangerous shorts.

Trade now: SMH entry is live, AMD and AVGO are close

**SMH is the only candidate with a live signal right now.** The semiconductor ETF closed at $583.66, already $4.22 below its lower Bollinger Band at $587.88, and ADX (14) sits at 35.1 — well above the 25 threshold the strategy requires. Both core entry conditions are met for SMH on the daily timeframe. If you are following this setup, SMH is the actionable name today. AMD and AVGO are not there yet. AMD last closed at $544.82 against a lower Bollinger Band at $529.85 — the price needs to drop roughly $14.97 (about 2.7%) to trigger the band condition, and ADX sits at just 15.6, well below 25. AVGO is tighter but still short: last close $394.29 versus a lower band at $386.64 (a $7.65 gap, roughly 1.9%), with ADX at 23.3 — close to the threshold but not yet through it. "Wait" for these two means placing alert orders at their respective band levels and reassessing only if ADX pushes above 25 on the same day the price breaches. On risk management, the strategy applies a hard stop at a 2.4% loss and a take-profit at 4.7% for every position, translating to an effective reward-to-risk ratio of roughly 2:1. The 60-month backtest on AMD delivered a 311% cumulative return across 189 trades with a 47.1% win rate and a 29.4% maximum drawdown — so expect frequent small losers and position sizing accordingly. The strategy caps any single position at 25% of portfolio equity with a minimum $100 trade size. No robust parameter sensitivity setup was established, as the optimization evaluation exceeded its time budget without producing a nearby-parameter recommendation. Treat the current published thresholds as the operative configuration.

Why the short-bias thesis has real fuel

The thesis rests on the idea that panic selling in semiconductors, once triggered, feeds on itself. The cited news supports the setup: per the Yahoo Finance piece on July 16, Marvell dropped 8% on AI capex slowdown fears and the selling spread to Broadcom, AMD, and Intel. The Bloomberg market wrap from the same evening confirms the contagion was going global, with Asian markets preparing to track the U.S. slide. That is precisely the environment where a momentum-breakdown strategy should fire — a coordinated de-risking event across the sector rather than an isolated single-name problem. The backtested evidence gives this idea a concrete backbone. Over the 60-month evaluation window on AMD, the rule set produced 189 trades and a cumulative…

Scores

• Conviction score breakdown: 55
• Thesis support: 60
• Trade readiness: 75
• Risk quality: 50
• Backtest evidence: 55
• Fundamentals trend: 35

Watch items

• SMH — Price vs Lower Bollinger Band (20, 2)
• SMH — ADX (14)
• AMD — Price vs Lower Bollinger Band (20, 2)
• AMD — ADX (14)
• AVGO — Price vs Lower Bollinger Band (20, 2)
• AVGO — ADX (14)
• SMH — Nearest Resistance
• AMD — RSI (14)
• AMD — Price below Bollinger (20)
• AMD — ADX (14) above 25