Bank of America Names Five Tech Stocks It Calls Top Buys for Second Half of 2026

Bank of America put out a list of five tech stocks it calls top ideas for the second half of 2026, according to CNBC. The names: Nvidia, Meta, Snowflake, Dynatrace, and Sandisk. Each comes with a buy rating and a specific price target from a named analyst.

Sandisk: Up 820%, and Bank of America Thinks There Is Still Room

Analyst Wamsi Mohan made the most striking call. Sandisk shares have already climbed more than 820% in 2026, and Bank of America still rates it a buy, raising its price target to $2,100 per share from $1,550.

Mohan's reasoning centers on Sandisk's shift away from traditional spot-market NAND pricing toward multiyear contract-based revenue. His argument: locked-in supply agreements give customers cost predictability and give Sandisk revenue stability, reducing the cyclical volatility that has historically punished memory chip companies.

"We see the NBMs as win-win as they lock in committed supply for customers, and committed financials for Sandisk," Mohan wrote, as quoted by CNBC. He also cited the company's joint venture partnership, expected share gains, and long-term industry consolidation potential.

A stock up 820% in under six months is already pricing in a great deal of good news. Mohan's case rests almost entirely on whether Sandisk's pricing power holds in a business notorious for brutal commodity swings.

Dynatrace: Down 4% Year-to-Date, Analyst Calls It a Second-Half Setup

Analyst Koji Ikeda raised Bank of America's price target on Dynatrace to $50 per share from $48, calling the AI-powered observability software company a "great" idea for the second half.

Ikeda's core thesis is that Dynatrace is positioned to capture larger enterprise deals as companies invest in AI deployment infrastructure that requires robust monitoring and security. He pointed specifically to net-new annual recurring revenue in constant currency as the metric to watch.

"We believe Dynatrace is poised for multiple years of high growth in a large and expanding market," Ikeda wrote, per CNBC.

Shares are down 4% year-to-date as of the most recent data in the source, which makes it the only name on this list where Bank of America is buying actual underperformance rather than momentum.

Meta: Down 13%, Bank of America Calls It a Dip Worth Buying

Analyst Justin Post is bullish on Meta despite shares falling roughly 13% year-to-date. His primary near-term catalyst is Meta's rollout of an AI-powered search feature.

"The 'search' opportunity remains large for Meta, if the company can drive product adoption," Post said, according to CNBC.

Post also listed a wave of upcoming catalysts: consumer agentic AI products, more advanced large language models, Meta's Connect conference scheduled for September 2026, and expanding enterprise AI opportunities.

Meta is a company with a history of regulatory pressure in both the U.S. and Europe, ongoing antitrust scrutiny, and a core advertising business that is cyclically sensitive to economic slowdowns. Investors who bought Meta as a "dip" during prior drawdowns sometimes waited years to see those thesis payoffs materialize. Post's case is coherent, but it depends heavily on Meta executing on AI monetization at a scale that justifies the valuation.

Nvidia: The AI Infrastructure Anchor

Bank of America has a $350 price objective on Nvidia, based on roughly 26x its estimated calendar-year 2027 price-to-earnings ratio excluding cash. The firm says that multiple is within Nvidia's historical 25x–56x forward-year PE range and is justified by Nvidia's dominant position in AI compute and networking markets.

Bank of America also flagged risks explicitly: lumpy global AI project timelines, a cyclical gaming GPU market, and concerns around access to power. Those are real risks worth taking seriously, not boilerplate disclaimers.

Snowflake: Cloud Data Warehouse With AI Traction

Bank of America rates Snowflake a buy, citing what it calls compelling competitive advantages: first-mover status as a data warehouse in the cloud, native interoperability with major public clouds, the ability to run multiple workloads including data warehousing and data science, a large installed base of enterprise customers, and traction in the AI software market.

Analyst Buy Lists and Structural Bias

Bank of America's research carries weight. The firm's analysts have named access to company management and sector depth that individual investors generally do not. These are not random predictions.

But analyst buy lists have a structural bias problem worth acknowledging: sell-side firms have business relationships with the companies they cover. That does not make any individual call wrong. It does mean investors should weigh buy ratings with that context in mind.

Five of five picks being buys is, by design, how these lists work.

The Open Question

The most genuinely unresolved issue here is Sandisk. A stock up 820% year-to-date that still carries a buy rating and a price target raised well above the prior level is either a legitimate paradigm shift in how memory companies get valued, or it is a momentum-driven overshoot. Mohan's multiyear-contract thesis is testable: Sandisk's next earnings report, and specifically whether new business model contract volumes are growing as projected, will be the clearest signal of whether that thesis holds.