AMD set for 90% crash on this date, according to expert

Technical outlook warns AMD could fall ~90% to around $53 by early Dec 2029 if a historical channel pattern repeats, versus ~$577 currently. It also flags bearish RSI divergence and prior drawdowns of ~91% (1984–85), ~94% (2000–02), and ~96% (2006–08). Offsetting that risk, AMD reported record Q2’26 revenue of $11.5B (+50% YoY), with Data Center up 107% to $6.7B (58% of sales), and expects Q3’26 revenue of about $13B.

Analysis

The chart-only bear case is only actionable if the market starts treating AMD as a cyclical semis name again rather than an AI compounder. The real downside mechanism is multiple compression: if data-center growth decelerates and margin expansion stalls, a premium revenue multiple can compress quickly even without a disaster in the business. That is a 6-18 month risk, not a days-only event; near term, momentum and positioning can keep the stock elevated until guidance or order visibility cracks.

Second-order competitive impact matters more than the historical channel. If AMD keeps taking share, the pressure lands first on pricing power across the accelerator stack, which is a relative win for hyperscalers like META and MSFT that can buy more compute per dollar. The cleaner competitive loser in a slowdown would be NVDA only if AMD forces broad price cuts; otherwise NVDA remains the higher-quality beneficiary if the market decides AMD's wins are tactical rather than structural.

Contrarian view: the consensus is overweighting pattern recognition and underweighting regime change. A 90% drawdown thesis is effectively a late-cycle liquidation scenario; the more realistic falsifier is continued data-center growth with stable or rising gross margin over the next two earnings cycles. If that persists, the bearish analogue is likely a false positive; if growth slips and the market starts questioning AI payback, AMD can still fall sharply long before any 2029-style outcome.

Key Decisions for Investors

• Do not short AMD outright on this headline alone; wait for the next earnings/guidance reset. If the stock loses momentum and management tone softens, use a 6-12 month bear put spread to express a valuation-compression view with defined risk.
• Relative-value trade: long NVDA / short AMD on rallies if AMD starts to underperform semis peers. Thesis: AMD is more exposed to pricing pressure as it tries to defend AI share, while NVDA still owns the scarcity premium.
• If already long SMH or SOXX, hedge with AMD puts rather than index puts. AMD is the cleaner early-warning indicator for AI multiple compression if the market rotates from scarcity to commoditization.
• Watch item, not a trade yet: if AMD data-center growth falls below ~30% year over year or gross margin fails to improve for two consecutive quarters, upgrade the bearish view from technical caution to fundamental short.
• Pair for a 1-3 month horizon: short AMD vs long META or MSFT only if evidence emerges that hyperscalers are using AMD competition to reduce AI infrastructure cost. Otherwise keep this as a monitoring basket, not a conviction spread.