If the Market Crashes, I'm Buying These 3 Stocks Before the Dust Settles

The article highlights three AI- and tech-related names as buy-the-dip candidates: TSMC, Roku, and Arm Holdings. TSMC continues to control more than two-thirds of the global foundry market, Roku posted Q1 revenue growth of 22% and gross profit growth of 27%, and Arm has multiple recent licensing deals that have yet to fully show up in reported revenue. Overall tone is constructive but mainly long-term and valuation-focused rather than a near-term catalyst.

Analysis

The main takeaway is not that TSMC, Roku, or Arm are all “good businesses,” but that each sits in a different part of the same capital cycle. TSMC is the cleanest expression of AI capex durability: if hyperscaler spend stays elevated, foundry utilization and pricing power remain resilient, and the real risk is not demand but sentiment-driven multiple compression. Second-order, every attempt by customers to diversify away from TSMC likely raises capex and time-to-market for rivals, which can actually reinforce TSMC’s moat over a 12-24 month horizon.

Roku is the more interesting marginal inflection. The market still tends to treat it as a cyclical ad-tech/streaming proxy, but the embedded installed base gives it quasi-consumer-platform characteristics; that matters because engagement and monetization can improve even in a slower ad market. The underappreciated benefit is that stronger platform usage should increase negotiation leverage with content partners and reduce churn in the device ecosystem, making the earnings stream less tied to pure streaming spend than consensus assumes.

Arm is the highest-beta setup and the most vulnerable to disappointment. The market is already discounting a multi-year royalty expansion before the reported revenue shows up, so the stock is effectively priced on the assumption that design wins convert cleanly into monetization with minimal slippage. Any delay in revenue recognition, weaker handset/PC demand, or slower enterprise AI adoption would hit this name harder than the others because the multiple leaves little room for execution error.

Contrarianly, the best risk/reward may be in waiting rather than chasing. TSMC and Arm both look structurally strong, but their current setups argue for buying only on market-wide drawdowns; Roku, meanwhile, may be the most mispriced because improving fundamentals are still not fully embedded in expectations. The near-term catalyst path is earnings and guidance, but the real test is whether AI spend broadens beyond a few hyperscalers and whether streaming ad budgets stabilize rather than accelerate.

Key Decisions for Investors

• Buy TSMC on a 3-8% market pullback; favorable 12-month risk/reward if AI capex remains intact, but avoid paying peak multiple for quality.
• Accumulate ROKU on dips and consider a 3-6 month long position into the next earnings cycle; upside comes from operating leverage if engagement continues to improve, while downside is limited by an already depressed market narrative.
• Avoid initiating new long ARM at current levels; if desired, use a staged entry only after a 10-15% reset, since the multiple implies near-perfect execution over the next 12 months.
• Pair trade: long TSM / short a basket of weaker AI infrastructure laggards on any broad selloff, to express quality-within-AI rather than outright beta.
• For more aggressive traders, use call spreads on ROKU into earnings rather than outright equity; the implied upside from improving monetization is meaningful, but the stock can gap sharply on macro ad-spend headlines.