CIO believes the tech sector’s correction so far in 2025 can be attributed to three major reasons: fundamentals-related noise, geopolitics, and tariff uncertainty. (UBS)
For investors in the tech sector, it’s an apt reminder that not only the concerns that prevailed during the 2018-19 around tariffs, geopolitics, excess build-outs (back then it was cloud and today it's AI) or China competition (back then it was Huawei and today it's DeepSeek) are rhyming; but also, the performance is repeating with strength (during 2015-17 and 2023-2024) followed by a period of extreme volatility (2018 and 2025 performance tracking so far).
However, if history is any guide, the period of extreme volatility in 2018 was also followed by a strong rebound as the market eventually looked beyond near-term uncertainty and focused instead on the tech sector’s strong fundamentals and superior balance sheets/free cash flow generation. In this blog, we analyze the lessons investors can learn from previous tech sector corrections and how to navigate near-term volatility.
We believe the tech sector’s correction so far in 2025 can be attributed to three major reasons: fundamentals-related noise, geopolitics, and tariff uncertainty.
First, let's look at fundamentals-related noise. Despite healthy fundamentals and strong quarterly results, the emergence of low-cost models like DeepSeek, market talk around data center spending, and concerns around excess build-outs have led to significant volatility year-to-date. While some of the concerns around low-cost models seem to have been largely priced in given the strong rebound in tech stocks till 20 February, there is no dearth of rumors or noise even today although the impact is marginal and diminishing. The good news is that quarterly results suggest AI fundamentals are intact, and that we should expect more innovation at the GTC 2025 event next week focusing on AI compute roadmap. Still, such noise is not new for the tech sector, and typically we witness one or two such 10% corrections every year in its shares. We have been highlighting that barring the reset in 2022, which was driven by macroeconomic events (rising rates and Russia-Ukraine escalations), over the past decade, tech indices have rebounded strongly over the next 12 months after a 10% fundamentals-related reset as investors have rewarded the sector eventually for its strong growth. Against this backdrop and with fundamentals still intact, we believe—based on lessons from the past—that the fundamentals-related noise has largely been priced in.
The second reason for the recent tech sector correction is geopolitical escalations amidst the proposed AI diffusion regulations. In an increasingly bipolar tech world, we have been reminding regularly that investors need to be prepared for a potential spike in volatility as governments and major corporations regularly reassess their plans.
We witnessed three such volatility spikes in 2019 (related to Huawei), 2022 (semi-cap restrictions), and 2023 (GPU restrictions), with the more sensitive semiconductor index (SOX) down on average 17.5%. While risks around geopolitics will likely remain for the long term, with the SOX index already down 18% from its peak in 2025 and tracking the average of previous corrections, we believe geopolitical risks have also been somewhat priced in as they were during the previous three volatility spikes; and we note that semiconductor stocks rebounded strongly after the big drawdowns.
The third reason, and what we now believe is the major reason behind the current weakness in the tech sector, is tariff-related uncertainty. CIO revised tariff scenarios last week (see our recent report “Global risk radar: Revised Tariff Scenarios” published on 5 March). The rise in US trade tariffs on Chinese goods to an effective rate of 30% from 10% is well within our base case scenario. However, the additional 25% tariff on goods from Mexico and Canada falls under our highly aggressive tariff downside scenario. With many tech supply chains relocating to Mexico, including AI servers' supply chain, tariff-related uncertainty will likely continue to weigh on tech stocks. While the EPS risk based on effective tariffs so far in 2025 is manageable with low-single-digit earnings impact, the lack of clarity on tariffs on goods from Mexico due next month and likely reciprocal tariffs could lead to significant tariff uncertainty in the near term.
In our base case, the earnings impact for tech is manageable as we still see mid-teen EPS growth for the overall global tech sector in 2025. However, in a highly aggressive tariff or downside scenario, we see further downside for tech P/E multiples and EPS. We believe such a downside scenario has not been priced in yet, and there is a risk that tech stocks may decline another 10%, tracking the 20%-plus correction witnessed in global tech stocks in 2018 amidst a significant tariff escalation.
While there are downside risks in a highly aggressive scenario, we believe investors should not lose hope; they should instead look to the past to see how the stocks performed once the uncertainty was fully priced in. After the 20%-plus correction in 2018 due to a significant tariff escalation, tech stocks rebounded by around 30% over the next 12 months as investors started to focus back on tech’s strong earnings growth and superior balance sheets/free cash flow generation.
Interestingly, the market response so far during the AI-driven rally is tracking the cloud-driven rally witnessed during 2016-19. As a background, cloud remained a key theme for global tech during 2016-19, and despite many headwinds, including concerns around geopolitics, tariffs, and excess build-outs, growth-based investors were fully rewarded for their patience during turbulent periods (and also during the next two years in 2020-21 driven by another theme around COVID-19 pandemic-driven spending).
In summary, our analysis suggests that risks around fundamentals-related noise and geopolitics seem to be largely priced in for tech, whereas risks around highly aggressive tariffs in the case of a further escalation have not been fully priced in yet. But with AI fundamentals still intact, investors should not lose sight of the long-term risk-reward and should take advantage of extreme volatility through buying-the-dip in quality AI stocks and structured strategies.
Original report: How low can tech go? Lessons from the previous corrections, 11 March 2025.
