3 Tech Stocks Poised for Comebacks

The article argues AI is a tailwind for software/learning and semiconductors: Adobe reported 13% YoY revenue growth in fiscal 2026 Q2 and said AI-first annual recurring revenue more than tripled to over $500M, helping explain its post–AI-fear 37% YTD drawdown. Duolingo also saw AI concerns misread, with Q1 revenue up 27% YoY as shares rebounded more than 40% from April lows despite ~30% YTD decline. Broadcom’s AI semiconductor revenue surged 143% YoY (nearing half of total revenue), and management expects AI semiconductor revenue to more than triple YoY next quarter, supported by a multi-year Apple partnership valued at $30B+.

Analysis

The key mechanism here is not “AI helps software,” but that AI can widen the moat for platforms that already own workflow, distribution, and proprietary data. That is most favorable for ADBE and, to a lesser extent, DUOL: the market is still pricing them like single-product SaaS stories when the more durable setup is higher usage intensity and better retention. The second-order risk is that AI-native point solutions compress the middle of the software stack, pressuring smaller niche vendors and creating a barbell where incumbents with embedded workflows win while undifferentiated apps get commoditized.

AVGO is a different animal: the investable story is custom silicon becoming a capex-efficiency tool for hyperscalers, not just “AI semis up.” That supports margins and visibility because ASIC wins tend to be sticky once designed in, and it also creates a downstream beneficiary set in the ecosystem, including AAPL, where silicon integration strengthens supplier lock-in. The flip side is that this is more schedule-dependent than the market assumes; revenue can lag design wins by quarters, and any pause in hyperscaler spend would hit sentiment before fundamentals.

Contrarian view: the market may be too quick to extrapolate “AI tailwind” into a durable rerating without demanding proof on monetization quality. ADBE is only compelling if AI-driven ARR converts into sustained operating leverage, while DUOL still trades on a narrative that can unwind if free AI substitutes improve faster than engagement. For AVGO, the near-term catalyst is earnings guidance; if AI revenue merely meets the very high bar rather than blows through it, the stock can de-rate even with strong absolute growth.

Key Decisions for Investors

• Buy ADBE on weakness over the next 2-6 weeks; target a 3-6 month hold into the next earnings cycle. Risk/reward is favorable if AI ARR keeps compounding and the market starts to re-rate the stock from a low-teens growth multiple toward a quality software multiple; invalidate if management signals AI is mostly offsetting legacy pricing pressure or growth slips back to low teens.
• Initiate a modest long AVGO / short NVDA pair trade over 6-12 months. Thesis: AVGO should benefit more from hyperscaler custom-ASIC adoption and capex efficiency, while NVDA carries more multiple risk if inference economics normalize. Keep sizing small; cover if NVDA growth reaccelerates or AVGO AI guidance disappoints.
• Treat DUOL as a watchlist name rather than an outright buy unless retention and paid conversion confirm the AI engagement thesis in the next print. If you want exposure, use a 3-6 month call spread instead of stock to limit valuation risk; stop if user growth slows or subscription churn rises.
• Use AAPL as a lower-beta way to express the second-order beneficiary angle from custom silicon and supplier stickiness, but only on broad-market weakness. This is a slower-moving catalyst than AVGO and should be treated as a multi-quarter accumulation idea, not a near-term trade.