Nvidia, AMD, and Intel: Wall Street Says to Buy 2 and Avoid 1. I Disagree.

The article argues Nvidia is the best buy among the AI chip trio based on valuation versus analyst targets: NVDA at ~$220 vs a ~$303 average 1-year target (+38%), compared with AMD at ~$484 vs ~$613 (+27%) and Intel at ~$97 vs ~$115 (+19%). It highlights that 32 of 48 Intel analysts rate it a “hold,” reflecting concerns that turnaround/revival expectations are already priced in. Overall, the piece is constructive on Nvidia’s upside relative to AMD and Intel, driven by perceived “reasonable valuation” while Intel and AMD are viewed as more expectation-dependent.

Analysis

This is less a fundamental upgrade than a positioning signal: the easiest capital to attract is incremental long-only flow that screens for upside to analyst targets. That favors NVDA because it has the cleanest earnings visibility and the least binary execution risk versus turnaround names; AMD and INTC both require investors to underwrite future share gains that are still not fully proven. The second-order effect is that relative underperformance in AMD/INTC can persist even if their absolute fundamentals are fine, because crowded AI capital tends to concentrate in the perceived toll collector rather than the challengers.

The main risk is that the market is already looking through these targets, so the near-term trade is mostly about whether NVDA can keep beating the consensus growth curve, not whether analysts think it is undervalued. Over 1-3 months, the key catalyst is next earnings/guidance: if NVDA only meets expectations or margins soften, the "cheap versus peers" argument fades quickly. For AMD, the thesis breaks if datacenter share gains slow; for INTC, the recovery case needs real external foundry volume, otherwise the stock risks becoming a story stock with balance-sheet drag.

Contrarian view: analyst target dispersion is usually backward-looking, and the market often prices the best-ranked name long before the upgrades arrive. So the more interesting setup may be to own NVDA on dips rather than chase it here, while fading the idea that AMD/INTC can keep rerating purely on hope. If NVDA remains the default AI exposure, the winners underneath are the adjacent suppliers with direct AI capex leverage; the losers are the names that need multiple expansion before the operating turnaround shows up.

Key Decisions for Investors

• Buy NVDA on a 3-5% pullback for a 1-3 month momentum/earnings continuation trade; thesis fails if next guidance shows gross margin compression or datacenter growth decelerates materially versus consensus.
• Pair trade: long NVDA / short INTC over the next 2-4 months to express "proven AI monetization vs. turnaround optionality"; target modest relative outperformance, but keep size small because INTC can gap on foundry/government headlines.
• If already long AMD, trim into strength and re-enter only after the next datacenter/share-gain proof point; the asymmetry is worse than NVDA because the stock needs repeated beats to justify the current multiple.
• Treat AMD as a tactical trading vehicle, not a core long, until the next earnings cycle confirms sustained acceleration; if guidance misses by even low-single digits, expect multiple compression rather than just a small de-rate.