DeepSeek’s AI model sent shockwaves through the stock market last week.
The open-sourced AI model from China could impact the outlook for a number of US stocks, according to JPMorgan.
The bank identified stocks like Amphenol and Caterpillar that could be potential losers from DeepSeek.
The market volatility shock from China’s DeepSeek AI model last month may be over for now as mega-cap technology companies commit to their massive AI spending plans.
Still, the ramifications for certain stocks could be big.
JPMorgan published a note on Wednesday highlighting the biggest potential losers from the stock market’s DeepSeek AI trade.
That revolves around the idea that Western companies could adopt new training methodologies unearthed by the open-sourced DeepSeek AI, leading to more efficient and less power-hungry large language models.
«A key theme to come out of their analyses is the potential for reduced training and inference costs, which could lead to balancing acts between (1) the decreased need for large-scale investment in current applications versus the rapid propagation of new AI applications that were previously not economically viable; and (2) a potential value shift both up to the application layer (inferencing) and down to the infrastructure layer (training),» JPMorgan analyst Caludia Hueston wrote in the note.
While the initial reaction to DeepSeek plunged Nvidia into the largest one-day market cap loss on record, the stock market’s second derivative trades are still playing out.
These are the eight potential biggest losers from the DeepSeek AI trade, according to JPMorgan.
Amphenol
Ticker: APH
Performance since DeepSeek:
JPMorgan rationale: Amphenol sells electrical and fiber optic connectors used in data centers, which JPMorgan says could be exposed to disruptions in demand.
«Content increase for connectors is tied to the increasing density of the vertically integrated solutions from Nvidia from one generation to another, and lower demand for the high density solutions could be a potential headwind to the bull case despite offsets from the increase in Inferencing demands.»
Caterpillar
Ticker: CAT
Performance since DeepSeek: -11.9%
JPMorgan rationale: «We estimate CAT’s revenue exposure to data centers is ~HSD% of total ME&T sales through backup power generators and Solar turbines that have seen strong order and sales growth in the past year. CAT, in response, is raising engine capacity by 125% vs. 2023 over the next 3-4 years. Growth in these two categories may slow if power demand from data centers decelerates.»
ME&T refers to Caterpillar’s Machinery, Energy and Transportation division, which includes sales of its earth-moving machinery that this integral to construction sites.
Custom Truck One Source
Ticker: CTOS
Performance since DeepSeek: +1.3%
JPMorgan rationale: Custom Truck One Source sells and rents out trucks and other heavy machinery, and has about 60% of its revenue tied to the transmission and distribution markets, the banks said. These markets «could see growth outlook curtailed because of efficiency and cost effectiveness attained through DeepSeek or similar models.»
Cummins
Ticker: CMI
Performance since DeepSeek: +0.03%
JPMorgan rationale: Cummins, a manufacturer of engines and generators, has significant sales exposure to data centers. The bank said sales estimates will be lower under the «pre-DeepSeek power generation models.»
Intel
Ticker: INTC
Performance since DeepSeek: -5.5%
JPMorgan rationale: «As compute complexity continues to grow, we believe there will be a continued shift to accelerated computing, which would negatively impact server CPU demand.»
CS Disco
Ticker: LAW
Performance since DeepSeek: +5.1%
JPMorgan rationale: CS Disco provides cloud- and AI-based services for the legal industry.
«We could see greater opportunity for CS Disco disintermediation as legal service customers could look to utilize efficient, low-cost AI models more frequently.»
Oracle
Ticker: ORCL
Performance since DeepSeek: -7.2%
JPMorgan rationale: «Our sense is that the emergence of DeepSeek potentially increases the likelihood of overbuilding on overly aggressive data center plans, although again the potential outcomes are many and complicated.»
Fabrinet
Ticker: FN
Performance since DeepSeek: -17.2%
JPMorgan rationale: The company provides manufacturing solutions for electrical components, with demand «tied to Nvidia’s vertically integrated solutions, which are less likely to find favor in Inferencing workloads, particularly with the rise in interest from hyperscalers in custom ASICs,» the bank wrote.
Read the original article on Business Insider
