AI is an exciting growth area, but the investing opportunities are not always obvious. Here’s how to integrate this theme into your portfolio.
One of the most exciting themes in science, technology and economics today is the transformative potential of artificial intelligence (AI) and automation, but finding the best way to invest in this theme is not always obvious.
The emergence of intelligent machines, sometimes known as “the fourth industrial revolution,” has the power to disrupt many aspects of the business world. While investment in AI and machine learning has come in and out of vogue since first discussed in the 1950s, the current mix of economic forces could unleash a wave of spending.
I believe we are at the start of a multi-year transformation in business and this theme will likely outlive any short-run business or market cycle. For now, investors can choose between companies adopting AI and those that are not. However, that may be short-lived as firms that invest in these technologies realize business efficiencies while those that ignore these tools may struggle to remain in business.
Higher wage costs are driving businesses to increase capital spending to improve efficiency in the competitive global economy and companies have access to investment capital following the 2017 tax legislation. Another long-term motivation is demographic in nature. The developed world has a unique challenge, with many countries facing labor shortfalls. By 2050, the U.S. alone will likely face an 18-million worker shortfall.
If the AI industry grows at a compound annual rate of 15.4% (my current estimate), it could reach nearly $1 trillion in revenues by 2050, based on automation replacing the projected shortfall of 18 million U.S. workers. With Europe, Japan and China facing similar demographic deficits, that growth estimate is likely conservative.
No company or set of companies has advanced a single dominant AI technology. Many firms have focused on more narrowly defined task-oriented elements of machine learning, rather than on developing general artificial intelligence. Instead of a single AI application for all purposes, investors should expect multiple AIs, or algorithms, blended for specific cognitive and physical tasks.
Commercializing AI technologies is still in the early stages. While many investors might begin their search among the companies that provide the actual automation services, they may find greater opportunities in firms that supply raw inputs required by AI algorithms or in the companies that use the core technologies to improve their primary business.
My advice for investors: Gain exposure to the technology across the AI ecosystem—upstream, core and downstream. Below is additional information on each segment:
Upstream: Companies supplying the raw material for core AI technologies include providers of processing power used in supercomputers and cloud data centers, as well as those with access to vast pools of data. Upstream opportunities also likely exist with companies that have expertise in data structuring (not just collecting data, but organizing it), as well as those with expertise in training the machines. AI also requires advanced sensors and control systems.
Not all companies in these sectors will benefit equally. Technology that could make one semiconductor company a leader in smart phones might not have the processing power for AI’s high-capacity needs.
Core: This group consists of the companies developing the actual artificial intelligence applications. Right now, a few leaders and many small start-ups are exploring new technologies, with a few firms advertising AI capabilities and products on the market, but many are still experimenting with applications. As is typical in technology, the first to market with a new product doesn’t always become the dominant player.
Large companies may have the upper hand, since they have already invested in extensive processing and cloud computing services that can help them deliver AI applications to existing customers. AI is a smaller component of total revenue and earnings for these players, many of which serve other markets. Smaller companies may turn out to be fast-growers if they possess winning technologies, but the potential uses for AI are so wide that the market can likely accommodate a number of players.
Downstream: The companies that make best use of the advances created by AI are also beneficiaries of the new technologies. Predicting which companies will successfully seize the opportunity is hard. They have to invest in new technologies and then execute on strategies effectively.
Companies that have high labor costs tied to simple repetitive tasks (think fast food) are probably the most obvious beneficiaries of AI. Retail and consumer goods firms are likely to benefit from improved profit margins, as labor costs shrink. Industries with high customer-service demands could also benefit, as natural language processing and machine learning improve. However, integrating these technologies efficiently into operations will likely prove difficult for some large firms.
Bottom Line: While AI and automation promise to be some of the most exciting growth areas in the market, this industry is still in its early days. Investors who focus on this theme should keep well in mind the basic investing tenet of diversification—both in their broader portfolio and as it applies to the emerging AI sector.
This article was derived from the February 27 issue of AlphaCurrents, a monthly Wealth Management Investment Resources publication which covers thematic investing. For more information on artificial intelligence and automation, ask your financial advisor for a copy.