The emergence of DeepSeek, a Chinese artificial intelligence startup rattled the markets on Monday. However, as AI demand continues to rise many described the reaction to be "overdramatic" and suggested to "buy the dip". Here is a list of top semiconductor exchange-traded funds, with the lowest expense ratios that investors could consider.
What Happened: As DeepSeek boasts of high performance at a significantly lower cost, semiconductor stocks which are the backbone of AI development, bore the brunt of the selloff.
"To me, it's an overreaction," Fundstrat Global Advisors‘ head of research Tom Lee told CNBC's "Closing Bell," comparing the drop to Nvidia Corp.’s NVDA March 2020 decline that proved to be a significant entry point for investors.
Similarly, Subho Moulik, the founder and CEO of Appreciate said, "The sudden emergence of DeepSeek is a Sputnik-like moment for the U.S. and potentially undermines U.S. AI exceptionalism, but we see yesterday's panic selling to be overdramatic."
As Nvidia shares plunged 17% on Monday, eroding $593 billion in market capitalization, Moulik said, "It's not just Nvidia, the other semicaps that took a battering in the markets – Broadcom Inc. AVGO and Micron Technology Inc. MU – can be an interesting buy on dip."
Investors could consider buying ETFs with a low expense ratio to "buy on dip." An expense ratio is a percentage that represents the annual cost of owning a fund or an ETF which is calculated by dividing a fund’s operating expenses by its net assets.
According to Charles Schwab, “ETFs with lower expense ratios and lower trading costs typically offer better value to investors, because higher expense ratios and/or steeper trading costs may significantly erode returns over time.”
| ETF Name | AUM (In Billion) | Expense Ratio | Six Month Returns |
| SPDR S&P Semiconductor ETF XSD | $1.28 | 0.35% | -0.29% |
| VanEck Semiconductor ETF SMH | $23.60 | 0.35% | -1.17% |
| iShares Semiconductor ETF SOXX | $13.86 | 0.35% | -7.14% |
| First Trust Nasdaq Semiconductor ETF FTXL | $1.32 | 0.60% | -9.11% |
| ProShares Ultra Semiconductors USD | $1.24 | 0.95% | -7.38% |
See Also: Microsoft’s AI Revenue Set To Surpass $10 Billion: Piper Sandler Maintains Overweight Rating
Why It Matters: According to Moulik, DeepSeek’s claim of lower training costs of $6 million vs ChatGPT’s $100 million plus, remains unverified. The Chinese firm's reliance on less-advanced Nvidia chips, due to U.S. chip export restrictions, could also hinder its ability to maintain a competitive edge in AI development.
Even if DeepSeek’s cost claims hold, it’s still positive for Big Tech, said Moulik, as AI extends beyond Large Language Models. Cost-effective LLM training allows U.S. companies to prioritize LLM inferencing – the next stage. Inferencing involves applying pre-trained LLMs for real-world tasks, optimizing AI investments by shifting focus from costly training to scalable, practical applications, ultimately driving long-term growth.
"For industry leaders like Microsoft Corp. and Alphabet Inc., this means redirecting massive spending from training infrastructure to real-world applications such as search, customer service, and automation," added Moulik. Cloud servers like Amazon.com Inc.’s Amazon Web Service could benefit too.
Although the short-term will be bumpy, the long-term runway appears stable, added Moulik, "Unless DeepSeek drops another Chinese New Year gift to the world later this year.”
Read Next:
Photo courtesy: Shutterstock
© 2025 Benzinga.com. Benzinga does not provide investment advice. All rights reserved.
Trade confidently with insights and alerts from analyst ratings, free reports and breaking news that affects the stocks you care about.
