Wedbush cautions that Meta may not be showing a clear payoff from its AI-driven spending

Meta’s push deeper into artificial intelligence and related infrastructure is running into a familiar question from Wall Street: when does heavy spending start to show up in returns? In remarks tied to the company’s most recent earnings, Ygal Arounian, Managing Director of Internet Equity Research at Wedbush, said he is “not seeing” the payoff from Meta’s capital expenditures, referring to the link investors typically expect between spending and measurable performance.

Arounian’s concern was framed less as a debate about whether Meta is investing in AI, and more about whether that investment is yet translating into improved efficiency. The key return yardstick he pointed to is ROIC, or return on invested capital, a metric that measures how much profit a company generates relative to the capital it has put to work. Investors often use ROIC to judge whether a business is turning spending into durable economic gains rather than simply increasing costs.

The Wedbush executive also suggested that Meta’s “forays” into the AI space could be complicating that story. Meta has been broadening its AI efforts across products and systems, which can require substantial engineering time, data-center capacity, and ongoing compute expenses. The timing of those investments can make early results harder to see, particularly if new capabilities are still being scaled or if near-term costs rise faster than business benefits.

The earnings-related discussion also matters for how Meta’s narrative is evaluated by investors. During periods when capex is elevated, markets frequently look for signs that spending is being absorbed into operating performance. Those signs can include improving margins, faster revenue growth from new features, or more efficient infrastructure utilization that supports ROIC. In Arounian’s view, those indicates were not sufficiently visible in the results he reviewed.

Arounian’s comments underscore a broader tension in the technology sector’s current AI cycle, where companies must balance aggressive buildout against the risk of longer payback periods. For Meta, the company’s large scale and diversified ad and engagement platforms create upside potential, but the return profile can still be questioned if infrastructure spending accelerates while monetization takes time. That is especially relevant to ROIC, because ROIC declines when companies invest heavily before profits rise proportionally.

Meta did not provide additional details in the Wedbush-linked post beyond the context of earnings commentary. No specific ROIC figures, capex totals, or investor guidance were included in the material associated with the remarks. As a result, it is not possible from the cited commentary alone to determine whether any improvements are already occurring in the underlying economics, how quickly they are expected to appear, or what internal targets Meta may be using to measure AI-related investment progress.

Investors are likely to keep watching for clearer evidence of how Meta’s spending is flowing through to returns in subsequent reporting periods. The next quarter’s disclosures could be closely read for trends in capital spending, profitability, and any metrics management highlights as it connects AI infrastructure to product performance and monetization. For now, Wedbush’s stance is a reminder that even when AI investment is strategically necessary, markets still demand a credible timeline for financial payoff.