Meta Platforms Inc. (NASDAQ:META) has something new to show investors after years of pouring billions into artificial intelligence: consumers appear to want the product. Muse has raced up app charts since its Sept. 8 launch, helping push Meta shares more than 20% higher. But Jensen Investment Management recently sold its Meta position over concerns about the returns on the company’s escalating AI investment — putting that decision under a new spotlight.
"META validated an important part of our original thesis," Allen Bond, Managing Director and Portfolio Manager at Jensen Investment Management, told Benzinga in an exclusive email interview. But he said Meta’s subsequent acceleration in capital spending and R&D was "greater than we anticipated," while the returns on much of that additional investment remain uncertain. "Our sale reflected this change in incremental ROIC," Bond said, "rather than deterioration in META’s high-quality core business."
Muse Changes the AI Debate
The early traction has quickly made Muse the centerpiece of Meta’s AI story.
Early adoption has given investors plenty to celebrate. Apptopia estimates Muse generated 2.8 million global installs during its first 12 days, while its U.S. and Canada iOS downloads outpaced ChatGPT’s comparable early mobile launch, reported TechCrunch. Muse has also reached No. 1 on the U.S. App Store.
Meta shares have responded accordingly. The stock gained more than 20% since Muse launched, with the stock’s rally adding about $200 billion in market value as investors began viewing the agent as a potential new revenue engine.
But Bond’s assessment goes beyond whether Meta has built a hit. The issue for Jensen is what happens to the economics as Meta’s AI investment gets larger. While AI has already demonstrated benefits for "user experience, ad targeting and monetization," the acceleration in capex and R&D means Meta now has to generate enough additional economic value to justify the expanding investment.
Investors Need More Than Downloads
For Bond, the evidence of sufficient AI returns would have to show up in the financial statements.
"Evidence of sufficient AI returns would include sustained revenue benefits," he said, followed ultimately by "a return toward historical free-cash-flow and margin profiles" across the major hyperscalers as investment growth normalizes.
And Bond sees several variables that make the return equation difficult to forecast. He points to "the magnitude and duration of the capex cycle," AI-model monetization, political pushback and "rapid improvements in chip efficiency and technology."
The ROI Test
But Bond’s framework shifts the investor question from "Will people use Muse?" to "Will the economics justify what Meta is spending to build it?"
For META, downloads may have fueled the rally; sustained revenue, cash flow and margins will determine whether the AI investment can ultimately earn its keep.
Photo: JRdes / Shutterstock – ejk
Login to comment