Americans are renowned for their inventive and entrepreneurial spirit. This underpinning of society has birthed US technological advancements within numerous industries across the globe over the past century. These leadership positions within industries are not guaranteed to continue, as we are not the only ones seeking to win the proverbial race.
The US government can implement export controls on certain technology and products to minimize economic competition and ensure military superiority. One of these restricts the sale of the most advanced AI chips produced by Nvidia to protect the US position as the AI leader going forward.
These export controls have been beneficial, and the US stock market has experienced a strong couple of years driven primarily by significant AI ecosystem investments and buildout. This investment has impacted multiple industries including technology hardware, power generation, and data centers. The increased revenue forecasts for 2024 and 2025 within the Russell 3000 are shown below:
While global military rivals of the US lack access to the newest and fastest processors, their desire to compete and develop AI technology has not been deterred. Necessity is the mother of invention.
China recently announced the creation of a new language learning model called DeepSeek to compete with the US based OpenAI. Notably, Chinese researchers utilized older Nvidia chips and completed the project at an allegedly 95% cost discount compared with US designs. The new AI model is unique in that they utilized OpenAI’s inference data and improved on the test trial times via reinforced learning (reward driven optimization). This is a novel approach, counter to the pre-training and post-training scaling used in other language learning models to date.
The announcement of DeepSeek had a significant impact on the stocks of major players in the US AI ecosystem. It raised the possibility that expenditures may have been excessive if novel approaches are cheaper, easier to implement, and require less power. On the day of the announcement, NVDA experienced the largest market capitalization decline in stock market history. The market worried that scalability may not require as much investment, the barriers to entry may be lower than previously anticipated, and the US may have lost the technological edge in the space.
The chart below depicts the market capitalization (stock price multiplied by number of shares outstanding) losses experienced by some of the past year’s highest-flying stocks on specific dates. DeepSeek caught the market’s attention on the weekend prior to 1/27/25. As shown by the unprecedented single-day loss, there was a marked level of worry about the impact on NVDA’s sales:
Most of the affected securities have stabilized as of mid-February, albeit at a lower pricing base. Notably, NVDA’s shares have not yet recovered to levels seen prior to the announcement of DeepSeek. Numerous companies continue to comment on the changes within the AI space, yet there is no consensus on structural impacts to AI cost deflation.
We at Foster Victor are monitoring developments carefully. Given that AI investment has driven the market for two years, a valuation reset within segments of equity markets may be upon us. Time will tell if the economics of AI advancement are truly decelerating, which could lead to a reduction in associated industry revenue forecasts.
Investment Commentary: Necessity is the Mother of Invention
By Andrew R. Duncan, CFA®
Americans are renowned for their inventive and entrepreneurial spirit. This underpinning of society has birthed US technological advancements within numerous industries across the globe over the past century. These leadership positions within industries are not guaranteed to continue, as we are not the only ones seeking to win the proverbial race.
The US government can implement export controls on certain technology and products to minimize economic competition and ensure military superiority. One of these restricts the sale of the most advanced AI chips produced by Nvidia to protect the US position as the AI leader going forward.
These export controls have been beneficial, and the US stock market has experienced a strong couple of years driven primarily by significant AI ecosystem investments and buildout. This investment has impacted multiple industries including technology hardware, power generation, and data centers. The increased revenue forecasts for 2024 and 2025 within the Russell 3000 are shown below:
While global military rivals of the US lack access to the newest and fastest processors, their desire to compete and develop AI technology has not been deterred. Necessity is the mother of invention.
China recently announced the creation of a new language learning model called DeepSeek to compete with the US based OpenAI. Notably, Chinese researchers utilized older Nvidia chips and completed the project at an allegedly 95% cost discount compared with US designs. The new AI model is unique in that they utilized OpenAI’s inference data and improved on the test trial times via reinforced learning (reward driven optimization). This is a novel approach, counter to the pre-training and post-training scaling used in other language learning models to date.
The announcement of DeepSeek had a significant impact on the stocks of major players in the US AI ecosystem. It raised the possibility that expenditures may have been excessive if novel approaches are cheaper, easier to implement, and require less power. On the day of the announcement, NVDA experienced the largest market capitalization decline in stock market history. The market worried that scalability may not require as much investment, the barriers to entry may be lower than previously anticipated, and the US may have lost the technological edge in the space.
The chart below depicts the market capitalization (stock price multiplied by number of shares outstanding) losses experienced by some of the past year’s highest-flying stocks on specific dates. DeepSeek caught the market’s attention on the weekend prior to 1/27/25. As shown by the unprecedented single-day loss, there was a marked level of worry about the impact on NVDA’s sales:
Most of the affected securities have stabilized as of mid-February, albeit at a lower pricing base. Notably, NVDA’s shares have not yet recovered to levels seen prior to the announcement of DeepSeek. Numerous companies continue to comment on the changes within the AI space, yet there is no consensus on structural impacts to AI cost deflation.
We at Foster Victor are monitoring developments carefully. Given that AI investment has driven the market for two years, a valuation reset within segments of equity markets may be upon us. Time will tell if the economics of AI advancement are truly decelerating, which could lead to a reduction in associated industry revenue forecasts.
Shannon DermodyTEST