Artificial intelligence has been a dominant theme for investors during 2023, as data centers have installed the equipment and software needed to uphold corporate clients that are rolling out their own AI technology.

But when Microsoft Corp.
MSFT,
-1.43%

CEO Satya Nadella and Nvidia Corp.
NVDA,
-2.68%

CEO Jensen Huang spoke together at Microsoft’s Ignite 2023 conference on Nov. 16, they agreed that industrial companies will benefit massively from AI over the long term.

Looking beyond the current build-out of data centers’ AI capability that has fueled a tripling of Nvidia’s share price this year, four money managers discussed how investors should think about the coming industrial-AI wave.

Nadella, Huang and industrial AI

When asked by Nadella about the “arc … of AI innovation going forward,” Huang called generative AI “the most significant platform transition in computing history. It’s bigger than PC, it’s bigger than mobile, it’s going to be bigger than internet,” he said.

Huang described three “TAM expansions,” with TAM standing for the total addressable market for computing services:

  • The first wave includes startups, such as OpenAI, which made its ChatGPT function available to the public a year ago. This wave also includes the creation of what Huang calls “AI factories,” which are data centers dedicated to “running AI models and generating intelligence.”

  • Huang called the second wave of the AI rollout the “enterprise generation,” which includes AI-enabled services, which he said have been kicked off by Microsoft Copilot.

  • The third wave, he said, “is going to be the largest wave of all,” referring to heavy industry. “This is where Nvidia’s Omniverse and generative AI [are] going to come together to help heavy industries digitalize and benefit from generative AI,” he said.

“{W]e are barely in the first wave, starting the second wave,” Huang said. Nadella agreed.

For evidence that the first wave is ongoing, consider that Nvidia’s stock has tripled in price this year and that its sales for the fiscal quarter ending Oct. 29 were triple those of the year-earlier quarter. The company’s graphics processing units continue to dominate this equipment category, which is critical for data centers’ AI rollout. Analysts expect Nvidia to continue to direct the semiconductor industry for the pace of revenue growth through 2025.

The comments from Huang and Nadella raise the question of how investors can take part in the great expansion that Huang expects for the computing-services market, with a long-term focus on the third wave.

Money managers chime in

In interviews with MarketWatch, four professional investors discussed Huang’s description of the three waves of AI technology, the timing of the waves and where investors who want to take advantage of the trends should be focused right now.

Art Amador of EquBot

Art Amador is the chief operating officer and co-founder of EquBot, which is based in San Francisco and serves as subadviser for the AI Powered Equity exchange-traded fund
AIEQ.
He didn’t offer an opinion on Huang’s statement about the three waves of AI deployment, but he did name three industrial companies that scored high in an analysis of companies based on AI concepts, as described below.

AEIQ’s objective is long-term capital growth, and it invests across industries and company sizes. The fund’s investment-selection process begins with about 6,000 U.S. companies, which are narrowed down through an automated, data-driven process that incorporates AI and machine learning. The process makes use of four deep learning models to score companies by their financial health, news events, quality of management and macroeconomic conditions.

All the data is combined in an effort “to forecast prices,” Amador said, adding that the fund “generally invests in mid- and small-cap companies.”

So the fund makes use of AI to narrow U.S.-listed stocks to a portfolio of 30 to 200 companies, but it is not designed specifically to be an AI play.

Amador said that separately from its selection of stocks for AIEQ, EquBot analyzed 50,000 companies globally to acknowledge “150 with greatest exposure to generative AI, based on scoring of more than 60 AI-related concepts.” Among the 150, he said 16 were industrial companies.

He named three examples, none of which are held by AEIQ at the moment:

  • General Dynamics Corp.
    GD,
    +1.01%
    ,
    a conglomerate that makes business jets, aircraft parts, nuclear submarines and other ships for the U.S. Navy and various other weapons systems. The company is headquartered in Reston, Va. It has a market capitalization of $68.1 billion and 106,500 employees, according to FactSet.

  • Booz Allen Hamilton Holding Corp.
    BAH,
    +2.39%
    ,
    which is based in McLean, Va., and provides management and technology consulting and other services to governments and companies globally. The company has a market cap of $16.7 billion and has 33,100 employees.

  • TaskUs Inc.
    TASK,
    +3.23%

    of New Braunfels, Texas, which provides digital outsourcing services worldwide to various industries, including food delivery, ride-sharing, gaming and e-commerce. The company has a market cap of $1.1 billion and has about 49,500 employees in 13 countries, but mainly in the Philippines.

Matt Moberg of Franklin Templeton

Matt Moberg manages the Franklin Intelligent Machines ETF
IQM,
which holds 55 stocks of companies that uphold machine learning and that use automated processes. The fund’s top holding is Nvidia, which made up 9.9% of the portfolio as of Nov. 30.

Moberg said he believed Huang was “correct” in his description of three waves of AI deployment. He said that there were three groups of companies involved in the development, production and deployment of intelligent machines:

  • About $2 trillion in market capitalization for computer-hardware companies, which is “50 companies, [including] a lot of semiconductor companies, with Nvidia [accounting for] half.” Nvidia’s market cap is $1.16 trillion, according to FactSet.

  • He calls the second group “infrastructure software,” which is dominated by “the big three”: Alphabet Inc.
    GOOGL,
    -1.96%
    ,
    Amazon.com Inc.
    AMZN,
    -1.49%

    and Microsoft. Together, those three companies plus Alibaba Group Holding Ltd.
    BABA,
    -1.30%

    have a market cap of $6 trillion.

  • The third group, which he says represents “another $2 trillion up for grabs” in market cap, includes over 100 companies that are “expected to add AI to their product sets and are able to boost their feature sets, or competitive advantages, by using AI on top, and hopefully monetizing it,” he said.

“Over the next decade you will begin to see it applied in the physical world. Automated driving will be part of it. We will see it more in robotics,” Moberg said.

Having Nvidia as the top position in IQM emphasizes Moberg’s belief that we are still in the first phase of the AI rollout. But he also believes that AI isn’t necessarily required for companies to make good use of automation and robotics.

For an example of “the general concept of bringing more robotics and data into our daily lives” without needing AI, he pointed to Axon Enterprise Inc.
AXON,
+2.06%
,
which makes body cameras used in law enforcement.

Another of the fund’s top holdings is Intuitive Surgical Inc.
ISRG,
-1.07%
,
which Moberg emphasized is using robotics in its surgical devices, but not AI. “Without a human mind behind it, it is nothing,” he said.

Looking ahead, Moberg said: “We will see large language modeling in the future combined with the robotics we have today — just as we don’t have fully autonomous driving today.”

Moberg said that the Franklin Intelligent Machines ETF’s portfolio reflects his belief that “you need to own companies providing the infrastructure and hardware that is needed” for the AI rollout. Here are the top 10 holdings of IQM, which made up 48% of the portfolio as of Nov. 30:

Axon Enterprise Inc.

Ticker

% of IQM

Nvidia Corp.

NVDA,
-2.68%
9.9%

Synopsys Inc.

SNPS,
-2.70%
5.4%

Cadence Design Systems Inc.

CDNS,
-2.54%
5.2%

Tesla Inc.

TSLA,
-1.36%
4.9%

Apple Inc.

AAPL,
-0.95%
4.9%

Intuitive Surgical Inc.

ISRG,
-1.07%
4.7%

ASML Holding N.V. ADR

ASML,
-0.27%
4.5%

Axon Enterprise Inc.

AXON,
+2.06%
3.0%

ASM International N.V. ADR

ASMIY,
-3.80%
3.0%

Taiwan Semiconductor Manufacturing Co. ADR

TSM,
-1.56%
2.9%

Source: Franklin Templeton

Rene Reyna of Invesco

Rene Reyna is the head of thematic and specialty product strategy for Invesco ETFs.

He explained that the Invesco AI & Next Gen Software ETF
IGPT
changed its strategy in August to track the STOXX World AC Next Gen Software Development Index. The index is made up of 100 companies “with significant exposure to technologies or products that contribute to future software development through direct revenue,” according to Invesco.

When deciding to change the fund’s strategy, Reyna said, “we saw a trend where hyperscalers, near term, seemed likely to win.” The company’s holdings are weighted by market capitalization. “If we look at a year in which AI has been top of mind, it stands out that the Nvidias of the world, with their market share, are leading the way,” he said.

The fund’s largest holdings include several hardware companies, such as Nvidia, Advanced Micro Devices Inc.
AMD,
-2.32%
,
Intel Corp.
INTC,
-3.18%

and Micron Technology Inc.
MU,
-2.16%
.

Reyna expects most “traditional AI” companies to be acquired and said that in addition to the hyperscalers such as Nvidia and Microsoft (which is not held by IGPT), the long-term survivors in the AI space will be companies with strong cash flow.

For the long-term adoption of AI by heavy industry, Reyna pointed to Amazon.

“They have AI robots able to distinguish certain packages by color. You begin to see, anecdotally, that these devices are probably going to revolutionize the industrial world,” he said.

Within the IGPT portfolio, Reyna highlighted companies making AI strides with software, including Meta Platforms Inc.
META,
-1.48%
,
which he said is “trying to get more involved with chatbots,” and Adobe Inc.
ADBE,
-1.29%
,
which has “rolled out generative fill-in in Photoshop,” as well as Firefly, an AI tool that can be used with various Adobe software platforms.

He also mentioned a piece of “older news,” which was Intel’s work with Volkswagen AG
VOW,
-0.75%

unit Audi “to automate and boost critical quality-control processes in their factories.”

Here are the top 10 stocks held by the Invesco AI & Next Gen Software ETF, which together make up 60% of the portfolio:

Company

Ticker

% of IGPT

Adobe Inc.

ADBE,
-1.29%
8.2%

Meta Platforms Inc. Class A

META,
-1.48%
8.1%

Nvidia Corp.

NVDA,
-2.68%
7.6%

Alphabet Inc. Class A

GOOGL,
-1.96%
7.3%

Advanced Micro Devices Inc.

AMD,
-2.32%
7.3%

Intel Corp.

INTC,
-3.18%
6.5%

Qualcomm Inc.

QCOM,
+0.22%
4.6%

Intuitive Surgical Inc.

ISRG,
-1.07%
4.1%

Keyence Corp.

6861,
-2.20%
3.3%

Micron Technology Inc.

MU,
-2.16%
3.1%

Source: Invesco

Hendi Susanto of Gabelli Funds

Hendi Susanto, a portfolio manager at Gabelli Funds, believes the pace of AI adoption will vary by industry.

“Investors and markets usually point to financial services for technology adoption,” he said, “as apposed to agriculture, where they are still talking about IoT adoption.” IoT stands for Internet of Things, which describes the network integration of physical objects, such as switches and sensors, and which is no longer a cutting-edge technological concept.

When asked if it might be too early for investors to focus on heavy-industrial companies’ adoption of AI, Susanto said yes, adding that “AI will still necessitate the basic building blocks of IT infrastructure. This includes computing, storage and networking.”

For AI implementation, there are two standards for network connectivity, Susanto said. “One is web-based, the other is InfiniBand,” which is a high-performance connectivity standard that “Nvidia owns,” he said.

He emphasized that heavy industry will need to utilize “building blocks” to bring computer networks up to standard for eventual adoption of AI.

C3.ai Inc.
AI,
-1.88%

“wants to furnish large-scale AI software platforms for different applications, including the oil and gas industry, for example,” Susanto said.

Moberg agreed, saying a “favorite example” of an AI application for the energy industry is building artificial data for use in oil extraction. “With some of the models now, you can drill fewer holes while looking for natural resources,” he said.

Don’t miss: Here’s why Nvidia is a compelling stock when compared with the rest of the ‘Magnificent Seven’

Source link