Investing.com -- UBS is urging investors to take advantage of the recent market correction and buy quality AI stocks, emphasizing that the structural AI growth story remains intact despite short-term volatility.
“With long-term fundamentals intact, we recommend taking advantage of extreme volatility through structured strategies or by buying the dip in quality AI stocks, including big tech,” UBS strategists led by Sundeep Gantori said in a note.
The firm remains confident in AI’s long-term trajectory, projecting strong double-digit annualized returns over the next five years. The latest earnings reports from major tech companies reinforce this outlook, as UBS maintains its estimate for global tech profit growth at 18% for 2025 and Big Tech profits at 21%.
AI stocks have come under pressure recently, partly due to concerns about the emergence of low-cost models like DeepSeek, which some fear could disrupt AI compute economics. However, UBS remains confident in the industry's outlook, pointing to robust capital expenditure trends from major technology players as a key driver of future growth.
“With a strong AI outlook from both Microsoft (NASDAQ:MSFT) and Meta (NASDAQ:META), we keep our estimates intact despite the recent uncertainty created by initial success of low-cost models like DeepSeek,” strategists noted.
Investor sentiment has also been rattled by reports of potential US export restrictions on NVIDIA’s H20 chips sold to China. While such measures could pose a short-term challenge, UBS downplays the broader impact, estimating that China accounts for just 6% of NVIDIA’s AI chip revenue.
“With strong demand continuing for NVIDIA (NASDAQ:NVDA) from the Big 4 and other customers, we believe the proposed controls are more manageable today than in October 2023, when such export controls were first introduced,” UBS’s team continued.
Thus, despite recent volatility, UBS believes the AI trade remains compelling for long-term investors. The firm notes that technology stocks have historically rebounded strongly after similar corrections.
In its report, UBS outlined both bullish and bearish scenarios for AI compute, with its base case projecting steady long-term growth.
In a bear case, the bank said that the proliferation of low-cost models like DeepSeek could drive down training and inference costs, leading to lower margins across the industry. However, UBS still expects robust AI training demand from enterprises and emerging markets, projecting a five-year compound annual growth rate (CAGR) of 20% for AI compute revenues and 12% for profits.
While pricing power may weaken, the firm sees manageable downside risks, forecasting low-teen annualized returns even in this scenario.
In its base case, UBS expects AI compute revenues to grow at a 30% CAGR through 2029, supported by steady demand for both AI training and inference. While margins may decline from 50% in 2024 to 40% in 2029, the firm still sees strong profit growth of 25% CAGR, making AI an attractive long-term investment opportunity at current valuations.
In a more optimistic bull case, UBS references long-term guidance from the foundry industry, suggesting AI compute revenues could grow at a 36% CAGR over the next five years. Profit growth could also remain strong, with margins declining only modestly from current levels.
Under this scenario, UBS expects AI investments to deliver annualized returns above 20%, with the enabling layer of AI infrastructure driving outsized gains.