AMD Drops 14% on the Week as Analysts Rush to Reprice After the Print

AMD fell 14% over the past week to close at $483.90, a sharper move than the median peer decline of roughly 5%, and the market's first verdict on a Q3 earnings report that landed into one of the most damaged sector backdrops in years.

The post-earnings analyst response has been unusually split. Morgan Stanley's Joseph Moore raised his target to $550 from $510 but held his Equal-Weight rating, a signal that the firm sees value emerging but is not ready to call a bottom. Piper Sandler and Wells Fargo both arrived at $650, though from opposite directions: Piper lifted from $600 while Wells cut from $700. Truist, DA Davidson, and others raised targets modestly, all maintaining Buy-equivalent ratings. RBC held Sector Perform at $540. The consensus mean now sits at $648, roughly 34% above Tuesday's close. That gap has widened, not narrowed, since the drawdown began. The Street's posture is consistent with what prior notes observed: analysts are treating this as a valuation reset, not a thesis break. The bull case remains server CPU demand running ahead of supply, management's guidance for more than 80% server growth in the second half of 2026, and Data Center revenue expected to more than double in 2027. The bear case is narrower but pointed: late console-cycle semi-custom weakness is driving a material sequential revenue decline in that segment, and execution on MI450 and Helios ramp timing remains unproven.

Positioning in the lending market is not telling a bearish story. Borrow availability is extremely loose, with over 1.6 billion shares available against fewer than 37.4 million short, and short interest covers just 2.3% of the free float. That figure has fallen roughly 8% over the past month and is near its lowest point in the full 2026 dataset. Cost to borrow dropped sharply to 0.25% on November 3, down 37% on the week, its lowest reading since October. There is no evidence of short sellers rebuilding into or after the print. Options positioning has eased slightly from recent averages: the put/call ratio is at 1.09, modestly below its 20-day mean of 1.11 and sitting about 1.6 standard deviations below that mean. Relative to the past year's range of 0.91 to 1.19, the PCR is comfortably in the middle. The collective read is that the market has repriced sharply but shorts are not pressing the trade.

The peer picture adds context to AMD's severity. INTC fell another 35% on the week after its own earnings collapse, extending the destruction in the legacy CPU space. MU dropped 8% and TER lost nearly 17%. AMD's 14% decline was painful but not the worst among its closest correlated peers. LSCC recovered 2% on the week and RMBS held near flat, illustrating how differently the sector is trading depending on exposure to data centre versus consumer and legacy industrial end markets. AMD's drawdown of 24% over the past month is the market pricing in both the sector contagion from Intel's collapse and the near-term headwind from semi-custom weakness, even as the data centre narrative remains structurally intact.

One institutional data point worth noting: JP Morgan Asset Management added over 12.7 million shares in the quarter ended September 30, the largest absolute addition among AMD's top holders, bringing its position to 29.5 million shares. FMR LLC added 10.5 million shares in the same period. Both moves were made at prices well above current levels, so they do not represent fresh conviction at the post-earnings price, but they do indicate that large institutional buyers were building into the drawdown before it accelerated. Insider activity from October 15 was entirely 10b5-1 plan sales by EVP Forrest Norrod, all pre-arranged and carrying a significance score of 2 out of 10. No discretionary buying appears in the recent record.

The next scheduled earnings event is February 2, 2027, ninety days away. Between now and then, the specific thing to watch is whether AMD's Q4 guidance, delivered alongside the Q3 print, frames the semi-custom decline as a one-quarter event or a multi-quarter drag, and whether any of the fifteen-plus analysts who currently sit above $600 begin cutting targets if the stock fails to close the gap to consensus.