Jeff Kilburg is adding Fortinet, Cloudflare, and Arista Networks to his Q4 portfolio, pitching the trio on CNBC's The Exchange as a way to gain exposure to AI spending without buying the hyperscalers that dominate the trade.
Two of the three picks already trade above analyst price targets, and one of them—Cloudflare—doesn't make money, a point Kilburg conceded on air. The third name, Arista, sits roughly 20% below its analyst target.
Why Kilburg is avoiding the hyperscalers
The logic is straightforward: rather than own the handful of mega-caps that have driven the AI rally, Kilburg wants the companies selling picks and shovels to the data-center buildout. Arista sells data center networking gear, Fortinet sells cybersecurity software, and Cloudflare routes and secures internet traffic. All three sit in the path of AI infrastructure spending without being the spenders themselves.
Arista is up about 58% this year and runs operating margins near 50%. Fortinet is up 130% this year and trades at roughly 53 times projected earnings. Cloudflare trades above a $350 target, with a forward P/E ratio near 300, according to Kilburg.
The valuation pushback
Host Kelly Evans didn't let the Cloudflare math slide. She cited the forward P/E as 246 rather than 300 and compared the multiple to Netflix's early years—a reference that cuts both ways, given how that story eventually played out.
Kilburg's answer was essentially that he's willing to pay up for the growth. He already holds CrowdStrike and Palo Alto Networks, so Fortinet adds diversification within cybersecurity rather than a new theme.
The broader earnings backdrop is what makes the valuation question relevant. FactSet projects S&P 500 earnings up about 29% in the third quarter—a third straight reading above 25%, against an 8% average over the past decade. That kind of growth can justify premium multiples, at least for now.
Borrowing costs complicate the math
The 10-year Treasury yield hit its highest level since 2002 and matches iCapital's raised yield forecast of 5.3%. When the risk-free rate is that high, the discount applied to future earnings gets steeper, and stocks trading at triple-digit multiples have the most to lose if growth doesn't materialize.
Schwab's Kevin Gordon has warned that a single mega-cap capital spending miss could disrupt the AI-driven market. That's the risk sitting underneath Kilburg's picks: if the hyperscalers slow their buildout, the networking and security vendors selling into it feel it first.
What the Q3 reports will show
Third-quarter reports, due over the coming weeks, may show whether profit growth can keep carrying premium multiples with borrowing costs at multi-decade highs. For Kilburg's three picks, the earnings calendar is the next real test. Fortinet, Cloudflare, and Arista all report in the weeks ahead, and each will have to show that AI-driven demand is still translating into revenue—and, in Cloudflare's case, into profit.
Kilburg's pitch rests on a simple bet: the AI trade doesn't have to be about the biggest companies. Whether the market agrees depends on what those three earnings reports say.
