Best 3 Tech ETF Picks for the 2nd Half of 2026

Tech is rebounding sharply after a late-March drawdown, with the sector now among the top S&P 500 performers in 1H 2026 following a Q2 resurgence tied to strong corporate earnings and easing geopolitical tensions. The article highlights continued AI infrastructure investment at “hundreds of billions” with improving AI-linked revenues/earnings and comparatively reasonable valuations, but flags near-term semiconductor volatility (chip stocks down ~13% in the past three weeks) and depressed software sentiment (IGV down 17% over the past year vs VGT up 39%). Overall, it frames VGT (broad tech), SMH (semiconductors/AI build-out), and IGV (software rebound setup) as ETF setups to continue riding the rally.

Analysis

The better expression here is not “own tech,” but own the subset where incremental AI spend still converts into pricing power and backlog: semis, networking, and the picks-and-shovels around inference. That favors NVDA and SMH over broad vehicles like VGT, because broad tech dilutes the capex beneficiaries with mature software and platform names that face slower monetization. The second-order risk is that once hyperscaler buildouts normalize, the earnings leverage that powered the first leg higher can reverse quickly; concentrated semiconductor baskets usually de-rate faster than broad tech when the market starts to price a digestion phase.

The near-term setup is flow-driven: momentum, dealer hedging, and underowned tech exposure can extend the rally for days to weeks even without new fundamentals. The 1-3 month catalyst path is earnings and capex commentary from large cloud and AI buyers; if they keep raising spend, semis should outperform software by a wide margin. If guidance merely holds flat, that is enough to keep the trade alive, but any sign of order normalization or inventory build would be the first falsifier.

The contrarian angle is that the market is still treating AI as a linear adoption curve, when in reality most of the economic rent may accrue to a handful of infrastructure suppliers while software faces margin compression from higher compute costs and faster feature commoditization. That makes IGV more of a “wait for proof” value trap than an immediate recovery trade. Over 6-18 months, the key watch item is whether software vendors can show measurable monetization per seat or per workflow; without that, the relative-value spread versus semis likely keeps widening.

Key Decisions for Investors

• Buy SMH on a 3%-5% pullback over the next 1-2 weeks; target 10%-15% upside over 3 months, with the thesis invalidated by any hyperscaler capex guide-down or a new inventory build signal.
• Pair trade: long SMH / short IGV for 1-3 months to express the view that AI capex still benefits hardware more than application software; exit if software earnings revisions turn positive or semis lose relative strength.
• Add NVDA only on post-earnings weakness or on a break above recent highs with volume; this is a momentum continuation trade, not a valuation anchor, and should be reduced if gross margin guidance softens.
• For broad tech exposure, prefer VGT over QQQ only if the goal is a cleaner tech beta with less index overlap; otherwise stay tactical and use it as a hedge against a short software or industrials pair.
• Set an alert on the next hyperscaler earnings cycle: if cloud capex is still accelerating, add to semis; if capex flattens, take profits in SMH and rotate into defensives.