�۶���Ƶ

Podcast: (5 min)
Podcast:JumpStart: Tariff threats on Russian oil, semis, pharma, and Swiss gold (5 min)
Podcast: Investors Club: Semis tariffs, margin pressures, and tech's growth outlook (13 min)
Video: The AI Show: CIO's Sundeep Gantori on tech earnings (3 min)

If tech momentum holds on Monday, investors could build on the Nasdaq’s two-day winning streak of record-high closes from last week. Rising hopes for a more dovish Federal Reserve and robust second quarter tech earnings have encouraged investors to look past macro challenges around employment and US tariffs. Large cap tech earnings meanwhile have confirmed strong cloud revenue growth and yet another rise in AI capital expenditure. We now forecast global AI capex will rise to USD 500bn in 2026.

However, other headlines are also raising some new uncertainties. Leading US AI chipmakers NVIDIA and AMD have agreed to share 15% of future China AI chip sales revenue with the US government, according to sources cited by Bloomberg and the Financial Times, as a prerequisite of sorts for securing US tech export licenses for their China-focused AI chips. Chinese state media meanwhile criticized US AI chips as potentially insecure on national security grounds, and insufficiently powerful for the needs of China’s domestic AI players. Separately, OpenAI’s highly anticipated ChatGPT 5.0 release has drawn a lukewarm response from its public user base, undershooting lofty expectations and its strong benchmark test results.

We remain constructive on the long-term AI opportunity set, but these latest developments point to a more mixed near-term outlook:

Two major AI catalysts came with surprise negative caveats. Both the anticipated release of GPT-5 and the widely expected resumption of US AI chip sales to China generated excitement and anticipation ahead of the fact. While chip stocks rallied on news of imminent export license approvals tied to US-China trade talks, the reported 15% revenue levy by the Trump administration raises new questions for investors. Likewise, GPT-5 initially received positive influencer feedback for its advanced capabilities, but negative public reviews soon followed over a lack of user control and other teething issues. Despite impressive benchmark scores, the consensus appears it is more evolutionary than revolutionary, especially given the sector’s elevated capex.

Margin pressure is emerging as a broader risk. Recent tech earnings have highlighted rising margin pressure across the sector. If these pressures extend to chipmakers, who may need to absorb this new 15% payment on anticipated China revenue, investors can expect more volatility. While details of the agreement are still pending, we think the impact could resemble the earnings drag seen from tariffs on other sectors, or regulatory costs more common in tech.

AI monetization is still a work in progress. Despite improving monetization, AI revenue growth has yet to fully match the sector’s aggressive capex. Our analysis of the Big 4 tech firms shows incremental capex of USD 115bn over the past year, versus incremental revenues of USD 170bn, or a capex intensity (capex divided by sales) of 67%. Looking ahead, we expect capex intensity to moderate toward 50%. For context, pre-generative AI capex intensity was just 10-15%, accentuating the need for further improvements on this as AI adoption broadens.

So, we keep our optimistic view on AI, but note that more recent developments do warrant some near-term caution. We continue to recommend a balanced approach within our AI portfolios, diversifying across semiconductors, software, and internet, and selectively adding exposure to AI laggards. Elevated valuations and challenging near-term headlines could see volatility rise in the coming weeks. Should a market pullback materialize, we believe long-term investors underallocated to our Transformational Innovation Opportunities (including AI) should take advantage of these dips to build exposure to these critical structural growth themes.