AMD Slips Another 12% But Analysts Keep Raising Their Targets

AMD closed Tuesday at $534.74, down 11.7% on the week, yet the analyst community spent the same stretch lifting price targets rather than trimming them.

That gap between what the Street is willing to pay on paper and where the stock is actually trading defines the AMD setup heading into November 3 earnings. The consensus mean target has now moved to $637.61, roughly $103 above Tuesday's close. Susquehanna raised its target to $620 from $550 on Wednesday, keeping a Positive rating. Wedbush reiterated Outperform at $600 the day before. Citi took its target all the way to $800 from $575 earlier in the month, and BofA Securities raised to $720 from $620 in late September. The direction of travel is unmistakable: across a dozen actions in the past four weeks, not one firm has cut a rating or a target. The Street is treating the drawdown as an entry question, not a thesis question. The bull case rests on data centre revenue of $6.72B in Q2, up 107% year over year, and the pending World Labs acquisition bringing Dr. Fei-Fei Li aboard as Chief Scientist. The bear case centres on the all-stock deal structure, gaming revenue down 31% year over year to $779M, and management's own guidance for modest sequential gaming declines in the second half of 2026.

Positioning in the lending market offers no pushback to the selloff whatsoever. Short interest in AMD is effectively inert: 2.39% of the free float, barely changed over the past week and down roughly 8% over the past month. Borrow costs 0.32% annually, one of the lowest readings in months after peaking near 0.71% in early September. Availability is at its maximum reportable reading, meaning the lending pool is almost entirely unused. There is no short-side pressure building. Options tell a similar story of calm rather than alarm: the put/call ratio is 1.10, fractionally below its 20-day average of 1.11 and sitting one standard deviation below that mean. That is not the profile of a market hedging aggressively into the print. The stock is falling on sellers, not on a crowded short base.

The peer group is providing no refuge. LSCC fell 11.3% on the week, LRCX dropped 12.5%, and MKSI shed 14.7%. INTC was off 10.8%. MU and CEVA each lost roughly 6%. AMD's 11.7% decline tracks almost precisely with the peer median of 11.1%, confirming the pattern from last week's note: this is a sector drawdown, not an AMD-specific deterioration. The stock's one-month decline of 4.5% is actually shallower than much of the group, partly because AMD arrived at this pullback from a stronger position after September's run. On valuation, the trailing P/E has compressed to 37.8 from over 46 a month ago, and EV/EBITDA has moved down to 32.4 from around 35. That compression is meaningful for a growth name, though the EV/EBIT factor score ranks in just the 6th percentile of the universe, a persistent reminder that AMD's multiples remain stretched in absolute terms even after the selloff.

Institutional ownership data as of September 30 shows BlackRock holding 9.2% and adding 4.7 million shares in the quarter. T. Rowe Price added 3.4 million shares and JP Morgan Asset Management added 12.8 million. FMR added 10.5 million shares. These are not trimming moves. The insider picture is more nuanced: EVP Forrest Norrod sold roughly $8.7M of stock on October 15, all under a pre-arranged 10b5-1 plan and therefore a mechanical disposal rather than a discretionary call. Net insider value over 90 days is minus $76.4M, but the plan-driven nature of the October sales limits their signal value.

Retail attention, according to Wikipedia page-view data tracked by ORTEX, has dropped sharply below AMD's own 90-day average, with a z-score of minus 3.3. That is colour rather than a directional indicator, but it fits the picture of a stock losing mindshare during a sector rout even as institutional holders hold their ground.

The specific thing to watch is November 3. The last earnings print produced a one-day move of minus 0.5% and a five-day move of minus 2.1%, both small. The question heading into Q3 is whether 107% data centre growth and any update on Instinct GPU demand can close the $103 gap between price and analyst consensus, or whether the Street will finally start trimming targets to meet the stock.