Artificial intelligence (AI) software has been around for many years, but big tech just unleashed a new kind called generative AI -- services that can use massive amounts of data to aid in the creation of new content. Microsoft invested in ChatGPT parent OpenAI, which has prompted a response from Alphabet with its Google Bard system. Other companies are following suit and figuring out ways to put AI to work in their own opertions.
A new round of software investment is thus heating up and laying the groundwork for the next bull market. Businesses that are ready for the coming boom could be in store for big gains. Three Fool.com contributors think Micron Technology (MU 2.20%), Lemonade (LMND 1.89%), and Palo Alto Networks (PANW 1.43%) are a great buy right now. Here's why.
Could memory makers experience a "V-shaped" recovery?
Billy Duberstein (Micron Technology): The memory and storage industry is in its worst downturn since 2008, but as bad as the current downturn is, could the memory stocks bounce back just as hard next year, in a "V-shaped" recovery?
While that scenario may not necessarily pan out, the possibility is certainly there. And if that's the case, Micron Technology stock would look very, very cheap today.
What could make that scenario come about? Well, demand evaporated last year just as the memory producers increased production from the strong demand in 2021, leading to oversupply. Therefore, if demand comes back with a vengeance while memory producers scale back production, that could have the opposite effect, increasing prices.
On that note, on its earnings call back in December, Micron announced a dramatic 40% decrease in 2023 capital expenditures, with a more than 50% reduction in capital equipment spend, along with a 10% cut to its headcount and a 20% near-term decrease in its current wafer starts.
Meanwhile, large rival SK Hynix, which also makes DRAM and NAND and reported results on Feb. 7, also announced a large 50%-plus cut to its own capital spending cut for 2023. While the third and largest DRAM manufacturer, Samsung, announced capital expenditures would remain flat, management also indicated most would go toward building new clean rooms and node technology transitions, not capacity expansions, and that certain upgrades would actually take some wafer production offline this year.
With at least two of the large players meaningfully scaling back production, limiting supply, there is the chance that demand could rebound. Memory price declines have led to an increase in density per device in phones and servers, due to the demand elasticity from lower prices. Furthermore, China's reopening could unexpectedly increase demand this year for smartphones and PCs. Meanwhile, last Wednesday's retail sales strength also pointed to perhaps a stronger-than-expected U.S. consumer as well.
But what could really spark demand to the upside are the AI wars that have kicked off with large language models such as ChatGPT. AI applications require huge amounts of fast processing and data storage. Meanwhile, the major processor chipmakers, including Nvidia, Advanced Micro Devices, and Intel, have each recently released new high-performance processors that will likely be deployed into these AI applications this year. As those are adopted, they will require new, higher-density and more expensive memory technologies such as DDR5 and high-bandwidth memory (HBM).
A recent Feb. 14 article in Business Korea noted SK Hynix and Samsung have seen a "surge" in new orders for HBM chips since the beginning of the year. And although that was in a Korean publication, leaving Micron out, Micron also makes HBM. While HBM is a relatively small part of the overall DRAM market today, the note is still a promising indicator that the AI wars could help spark a rebound in memory demand.
With suppliers pulling back on production and the potential for demand to surprise to the upside, investors shouldn't rule out a strong rebound for memory prices. In fact, SK Hynix's management said on their Feb. 7 conference call, "we expect that not only will inventory levels normalize this year, but also a better than expected upturn may come next year."
While all memory suppliers would benefit in that scenario, Micron has impressed on the technology side, as it became the first memory maker to mass produce 1-beta DRAM and 232-layer NAND in 2022. As long as Micron keeps up its technological lead and targets the AI market, it should have a much better second half of 2023 and 2024.
When Mr. Market gives you lemons, bring some sugar
Anders Bylund (Lemonade): Let's talk about Lemonade, an insurance technology company powered by artificial intelligence and highly automated processes.
By automating much of the traditional insurance process using AI, Lemonade is able to offer affordable insurance policies to its customers. Its homegrown AI technology is similar to the language model GPT, which runs the popular ChatGPT service. Like ChatGPT, Lemonade's AI system uses natural language processing to make sense of a large amount of potentially disorganized data. The sign-up process is managed by AI-driven chatbots and insurance claims are run through the same chatbot, triggering a specialized module that specializes in analyzing risks, damage, repair costs, and so on.
The company likes to shine a spotlight on its cool AI tools. In last November's investor day presentation, for example, Lemonade's custom chatbot provided a Shakespearean sonnet about AI and insurance, followed by a computer-generated schematic of an AI-driven insurance system:
Image source: Lemonade.
The inner workings of Lemonade's AI-powered analytics and customer service systems are proprietary and unknown. Still, the company clearly makes at least some use of OpenAI's GPT (language-based modeling) and DALL-E (image generation) services -- tweaked to serve its specific insurance service needs.
The more customers Lemonade earns, the more real-world data it gets to train its systems on. As a result, its underwriting and risk management policies should become more accurate and profitable over time. So when people panic over Lemonade's less-than-perfect loss ratios and negative profits, they're treating temporary issues like permanent business flaws.
This innovative insurance disruptor has seen share prices fall more than 90% in two years while it expanded its coverage offerings both geographically and into new insurance types, with a fully nationwide push into the lucrative car insurance market still ahead. The top-line revenue is not just skyrocketing, but accelerating over time:
Data by YCharts.
Yes, the company is burning some cash in an effort to maximize its market share before a torrent of imitators come along. That's a smart long-term growth strategy in my eyes, as long as you can accept lumpy and often unprofitable financial reports along the way.
When the next bull market comes along, it'll throw gasoline on Lemonade's growth fires as consumers start buying cars and homes again. On top of that catalyst, Lemonade has captured less than about 0.01% (not a typo!) of the American insurance market so far.
This company faces a mind-bogglingly large opportunity for disruptive growth.
Cybersecurity that's ready for the AI era
Nicholas Rossolillo (Palo Alto Networks): Palo Alto Networks is the largest cybersecurity pure-play stock out there, with $5.8 billion in revenue over the last reported 12-month stretch. (Microsoft said it did $20 billion in cybersecurity sales in 2022, so it's far and away the security software leader.) Cybersecurity is a tricky industry. It requires constant innovation and necessitates playing non-stop offense to stay ahead of digital threats. Under the guidance of CEO Nikesh Arora, Palo Alto Networks has transformed itself into a technological powerhouse ready to tackle the complexities of the AI era.
During its last quarterly update, Palo Alto Networks said its "next-gen" security platform (designed for modern cloud-based IT and other advanced computing) was generating $2.1 billion on an annualized basis. With an advanced suite of services that cover all bases of a large organizations' operations, Palo Alto Networks is well positioned to capture new business in the years ahead as more of its customers migrate to the cloud and adopt AI-powered applications.
This cybersecurity leader is also in good financial shape, which gives it the ability to keep playing offense. For example, it recently acquired a company that specializes in software supply chain security. Most software today, including AI, is virtually "assembled" with smaller pieces of software, pieces that need to be protected as they travel from one developer to the next. Palo Alto's recent purchase could thus help fill a critical need as new AI systems get rapidly deployed in the coming years.
Palo Alto can afford to make these occasional bolt-on purchases of smaller peers. It generated $2.4 billion in free cash flow over the last 12-month period, more than enough to make acquisitions and fund its share repurchase program (which is currently being used to offset the effects of employee stock-based compensation).
Palo Alto Networks stock currently trades for just under 22 times trailing-12-month free cash flow. For a company that expects to grow at least 20% in the next year and sustain high levels of cash profits, this looks like a very fair price to me. I remain a buyer.
