Eric Santa Menargues
January has certainly been a dynamic moment for global markets, setting the tone for what looks to be an eventful year. Markets kicked off the year with plenty of headlines from the inauguration of President Trump to a jolt in the tech sector following the emergence of Chinese AI company DeepSeek.
Adding to this patch of uncertainty has been the looming threat of new tariffs on global trade, the release of DeepSeek R1 model was supposedly significantly lower training costs triggered a dramatic reaction. Nvidia’s market value plummeted by nearly $600 billion in a single day, making the largest one day loss in U.S. market history as investors weighed up compute demand going forward.
The initial market reaction was aligned with a shift in value creation from AI enablers, those that provide the foundational infrastructure for AI and had until now performed within the AI trend to AI integrators, those that build software applications and services on top of that infrastructure.
One potential implication of lower training costs is an increase in LLM competition leading to further opportunities to increase AI into business models. Although this development might have tilted the growth opportunities from enablers to integraters, we believe there are plenty of opportunities throughout the AI value chain and access to compute will remain a competitive advantage. We dive deeper into these trends and the mock implications in our latest monthly available on our webpage.
