Disclaimer
Value Situations is NOT investment advice and the author is not an investment advisor.
All content on this website and in the newsletter, and all other communication and correspondence from its author, is for informational and educational purposes only and should not in any circumstances, whether express or implied, be considered to be advice of an investment, legal or any other nature. Please carry out your own research and due diligence.
In my Model Portfolio year-end review published last month, I remarked how I took some satisfaction in my Portfolio’s SPX-matching performance of +24% being achieved without owning any of the “Magnificent 7” names, as I explained:
My objective of achieving positive absolute returns regardless of broader market performance was achieved, as evidenced by the composition of the Model Portfolio - I effectively matched the SPX’s performance without owning the key driver of the index’s performance, i.e. the Magnificent 7 (or what I previously referred to as the Megacap 8, including Netflix); excluding the “Mag7,” the SPX’s performance in 2023 was flat, and so I take some satisfaction in the Portfolio’s market-matching performance being driven by idiosyncratic situations such as YCA, SOL, TE and DOLE, which all comfortably outperformed the broader market on their own merits and uncorrelated to the Mag7 phenomenon.
In summary, the Portfolio essentially matched the market’s performance without needing to rely on the Mag7 names, which I felt were propelled by a speculative fervour underpinned by enthusiasm around artificial intelligence (AI) and an expectation that pending interest rate cuts would drive their valuations ever higher.
At 2023 year-end, the average EV/S multiple across the Mag7 cohort was ~9x, with AI poster-boy NVDA trading at ~27x (and reaching as high as 45x at one point last July!):
In EV/EBITDA terms, the valuation picture was equally speculative, with the Mag7 ending the year at an average multiple of ~28.5x, nearly twice the SPX market multiple of ~15.5x:
Indeed, just as Reddit meme stocks were the winning market theme in 2021, followed by “old economy” energy stocks in 2022, Mag7/AI was the winning theme in 2023:
Mag7 performance in 2023
AI-related stocks performance in 2023
While I was happy to avoid the AI theme last year given the speculative, momentum-driven valuations of AI-related stocks, I have to acknowledge that this technology clearly has a role to play in the economy of the future. This has therefore prompted me to think whether there might be a less speculative, more fundamentally-driven “back door” way to play the theme.
In mulling this question over, I came across two interesting pieces from my friend Herb Greenberg in his On The Street newsletter, one relating to AI and the second to the “old economy” technology that is the PC. In the first piece, Backdoor Play on AI, Herb discussed how data and analytics company S&P Global (SPGI) could be a potential indirect play on the AI theme. This makes a lot of sense given that data is the lifeblood of AI and the large language models such as ChatGPT that generated so much excitement last year.
Herb’s second piece, What’s Old is New Again: The Coming Boom in PCs, was equally interesting in its discussion of a revival in demand for PCs driven by a new replacement cycle with much more powerful PCs equipped to run AI models and applications. Likely beneficiaries in this regard would seem to be PC manufacturers such as Dell (DELL), HP (HPQ) and Lenovo Group (992), and PC chip names such as Intel (INTC) and AMD (AMD).
These two articles prompted me to pull further on this “backdoor” AI thread, to find an overlooked business or situation that should benefit from AI, while also meeting my idiosyncratic value-with-a-catalyst criteria.
After some searching, I believe I’ve found a very interesting situation that fits both these thematic and fundamental value filters. This stock is a US-listed name and is my first watchlist idea for 2024…